Health Glance 2017 en
Uploaded by
Guto VieiraHealth Glance 2017 en
Uploaded by
Guto VieiraHealth at a Glance 2017
OECD Indicators
Health at a Glance
2017
OECD INDICATORS
This work is published under the responsibility of the Secretary-General of the OECD. The
opinions expressed and arguments employed herein do not necessarily reflect the official views
of OECD member countries.
This document, as well as any data and any map included herein, are without prejudice to the
status of or sovereignty over any territory, to the delimitation of international frontiers and
boundaries and to the name of any territory, city or area.
Please cite this publication as:
OECD (2017), Health at a Glance 2017: OECD Indicators, OECD Publishing, Paris.
http://dx.doi.org/10.1787/health_glance-2017-en
ISBN 978-92-64-28039-7 (print)
ISBN 978-92-64-28040-3 (PDF)
ISBN 978-92-64-28088-5 (epub)
Series: Health at a Glance
ISSN 1995-3992 (print)
ISSN 1999-1312 (online)
Revised version, February 2018
Details of revisions available at: http://www.oecd.org/about/publishing/Corrigendum_Health_at_a_Glance_2017.pdf
The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities.
The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and
Israeli settlements in the West Bank under the terms of international law.
Photo credits: Cover © chombosan/Shutterstock.com.
Corrigenda to OECD publications may be found on line at: www.oecd.org/publishing/corrigenda.htm.
© OECD 2017
You can copy, download or print OECD content for your own use, and you can include excerpts from OECD publications, databases and
multimedia products in your own documents, presentations, blogs, websites and teaching materials, provided that suitable
acknowledgement of OECD as source and copyright owner is given. All requests for public or commercial use and translation rights
should be submitted to rights@oecd.org. Requests for permission to photocopy portions of this material for public or commercial use shall
be addressed directly to the Copyright Clearance Center (CCC) at info@copyright.com or the Centre français d’exploitation du droit de
copie (CFC) at contact@cfcopies.com.
Foreword
Foreword
H ealth at a Glance 2017 presents the latest comparable data and trends on key indicators of
health outcomes and health systems across the 35 OECD member countries. These indicators shed
light on the performance of health systems, with indicators reflecting health outcomes, non-medical
determinants of health, the degree of access to care, the quality of care provided, and the financial
and material resources devoted to health. For a subset of indicators, data are reported for partner
countries, including Brazil, China, Colombia, Cost Rica, India, Indonesia, Lithuania, the Russian
Federation and South Africa.
The production of Health at a Glance would not have been possible without the contribution
of OECD Health Data National Correspondents, Health Accounts Experts, and Health Care Quality
Indicators Experts from the 35 OECD countries. The OECD gratefully acknowledges their effort in
supplying most of the data contained in this publication. The OECD also acknowledges the contribution
of other international organisations, especially the World Health Organization and Eurostat, for sharing
some of the data presented here, and the European Commission for supporting data development work.
This publication was prepared by a team from the OECD Health Division under the coordination
of Chris James. Chapter 1 was prepared by Chris James and Alberto Marino; Chapter 2 by Chris
James and Marion Devaux; Chapter 3 by Eileen Rocard, Chris James, Marie-Clémence Canaud and
Emily Hewlett; Chapter 4 by Sahara Graf, Marion Devaux and Michele Cecchini; Chapter 5 by Alberto
Marino, Chris James, Rie Fujisawa, Akiko Maeda, David Morgan and Eileen Rocard; Chapter 6 by Ian
Brownwood, Frédéric Daniel, Rie Fujisawa, Rabia Khan, Michael Padget and Niek Klazinga; Chapter 7
by David Morgan, Michael Mueller and Michael Gmeinder; Chapter 8 by Akiko Maeda, Gaëlle Balestat
and Michael Gmeinder; Chapter 9 by Chris James, Gaëlle Balestat and Alberto Marino; Chapter 10 by
Rabia Khan, Gaëlle Balestat, Marie-Clémence Canaud, Michael Mueller, Martin Wenzl, Chris James
and Valérie Paris; Chapter 11 by Tim Muir, Eileen Rocard, Michael Mueller and Elina Suzuki. The OECD
databases used in this publication are managed by Gaëlle Balestat, Ian Brownwood, Marie-Clémence
Canaud, Frédéric Daniel, Michael Gmeinder, Gaétan Lafortune and David Morgan.
Detailed country comments improved the quality of this publication, as did comments from
Francesca Colombo, Gaétan Lafortune, Mark Pearson and Stefano Scarpetta. Format and editing
support from Marlène Mohier, Kate Lancaster and Andrew Esson are also gratefully acknowledged.
Health at a Glance 2017 © OECD 2017 3
Table of contents
Table of contents
Executive summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Reader’s guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Chapter 1. Indicator overview: OECD snapshots and country dashboards . . . . . . . . . 17
OECD snapshots and country dashboards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Health status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Risk factors for health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Access to care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Quality and outcomes of care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Health care resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Chapter 2. What has driven life expectancy gains in recent decades?
A cross-country analysis of OECD member states . . . . . . . . . . . . . . . . . . . . 31
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Understanding the determinants of health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Gains in life expectancy over time reflect increased health spending,
healthier lifestyles and improving socio-economic conditions. . . . . . . . . . . . . . . . . 34
Unpacking the mechanisms by which socio-economic factors
and a person’s living environment affect health is essential for policy . . . . . . . . . . 39
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Chapter 3. Health status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Life expectancy at birth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Life expectancy by sex and education level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Main causes of mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Mortality from circulatory diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Mortality from cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Infant health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Mental health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Perceived health status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Cancer incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Diabetes prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Chapter 4. Risk factors for health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Smoking among adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Alcohol consumption among adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Smoking and alcohol consumption among children . . . . . . . . . . . . . . . . . . . . . . . . . 74
Healthy lifestyles among adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Health at a Glance 2017 © OECD 2017 5
Table of contents
Healthy lifestyles among children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Overweight and obesity among adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Overweight and obesity among children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Air pollution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Chapter 5. Access to care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Population coverage for health care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Unmet needs for health care due to cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Out-of-pocket medical expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Geographic distribution of doctors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Waiting times for elective surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Chapter 6. Quality and outcomes of care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Patient experience with ambulatory care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Prescribing in primary care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Avoidable hospital admissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Diabetes care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Mortality following ischaemic stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Mortality following acute myocardial infarction (AMI) . . . . . . . . . . . . . . . . . . . . . . . 110
Hospital mortality rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Waiting times for hip fracture surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Surgical complications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Obstetric trauma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Care for people with mental health disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Screening, survival and mortality for breast cancer . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Survival and mortality for colorectal cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Survival and mortality for leukaemia in children . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Vaccinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Chapter 7. Health expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Health expenditure per capita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Health expenditure in relation to GDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Financing of health care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Sources of health care financing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Health expenditure by type of service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Health expenditure by provider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Capital expenditure in the health sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Chapter 8. Health workforce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Health and social care workforce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Doctors (overall number) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Doctors by age, sex and category . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Medical graduates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Remuneration of doctors (general practitioners and specialists) . . . . . . . . . . . . . . . 156
Nurses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Nursing graduates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Remuneration of nurses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Foreign-trained doctors and nurses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
6 Health at a Glance 2017 © OECD 2017
Table of contents
Chapter 9. Health care activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Consultations with doctors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Medical technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Hospital beds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Hospital discharges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Average length of stay in hospitals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Hip and knee replacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Caesarean sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Ambulatory surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Chapter 10. Pharmaceutical sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Pharmaceutical expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Pharmacists and pharmacies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Pharmaceutical consumption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Generics and biosimilars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Research and development in the pharmaceutical sector . . . . . . . . . . . . . . . . . . . . . 194
Chapter 11. Ageing and long-term care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Demographic trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Life expectancy and healthy life expectancy at age 65 . . . . . . . . . . . . . . . . . . . . . . . 200
Self-reported health and disability at age 65 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Dementia prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
Recipients of long-term care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Informal carers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Long-term care workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Long-term care beds in institutions and hospitals . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Long-term care expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Follow OECD Publications on:
http://twitter.com/OECD_Pubs
http://www.facebook.com/OECDPublications
http://www.linkedin.com/groups/OECD-Publications-4645871
http://www.youtube.com/oecdilibrary
OECD
Alerts http://www.oecd.org/oecddirect/
This book has... StatLinks2
A service that delivers Excel® files from the printed page!
Look for the StatLinks2at the bottom of the tables or graphs in this book.
To download the matching Excel® spreadsheet, just type the link into your
Internet browser, starting with the http://dx.doi.org prefix, or click on the link from
the e-book edition.
Health at a Glance 2017 © OECD 2017 7
Health at a Glance 2017
© OECD 2017
Executive summary
H ealth at a Glance 2017 presents up-to-date cross-country comparisons of the health status
of populations and health system performance in OECD and partner countries. Alongside
indicator-by-indicator analysis, this edition offers snapshots and dashboard indicators that
summarise the comparative performance of countries, and a special chapter on the main
factors driving life expectancy gains.
Most OECD countries have universal health coverage systems which promote equitable
access for needed health services. Quality of care has also generally improved, but this has
come at a cost: health spending now accounts for about 9% of GDP on average. Investing in
cost-effective health promotion interventions is one important way to improve value for
money and reduce health inequities.
People in OECD countries are living longer, but the burden of mental illness
and chronic disease is rising
●● Life expectancy at birth is 80.6 years, on average, across OECD countries. Japan and Spain
lead a group of 25 OECD countries with life expectancies over 80 years.
●● Turkey, Korea and Chile have experienced the largest gains in life expectancy since 1970.
●● Health spending contributes to longevity, but only explains part of the cross-country
differences and gains in life expectancy over time. New regression estimates suggest
healthier habits and wider social determinants of health are also key.
●● Women can expect to live just over five years longer than men, while people with tertiary
level education live around six years longer than those with the lowest level of education.
●● Across the OECD, more than one in three deaths are caused by ischaemic heart disease,
stroke or other circulatory diseases; one in four deaths are due to cancer.
●● Mortalityrates for circulatory diseases have fallen rapidly, with 50% fewer deaths due to
ischaemic heart disease, on average, since 1990. Cancer mortality rates have also fallen,
though less markedly, by 18% since 1990.
While smoking rates continue to decline, there has been little success in
tackling obesity and harmful alcohol use, and air pollution is often neglected
●● Smoking rates have decreased in most OECD countries, but 18% of adults still smoke daily.
Rates are highest in Greece, Hungary and Turkey, and lowest in Mexico.
●● Alcohol consumption in the OECD averaged 9 litres of pure alcohol per person per year,
equivalent to almost 100 bottles of wine. This figure is driven by the sizeable share of heavy
drinkers: 30% of men and 12% of women binge-drink at least once per month.
●● In 13 OECD countries alcohol consumption has increased since 2000, most notably in
Belgium, Iceland, Latvia and Poland.
Health at a Glance 2017 © OECD 2017 9
Executive summary
●● Since
the late 1990s, obesity has risen quickly in many OECD countries, and more than
doubled in Korea and Norway, albeit from low levels.
●● 54%of adults in OECD countries today are overweight, including 19% who are obese.
Obesity rates are higher than 30% in Hungary, Mexico, New Zealand and the United States.
●● Among 15 year olds, 25% are overweight and only 15% do enough physical activity. Further,
12% smoke weekly and 22% have been drunk at least twice in their lives.
●● In 21 countries, over 90% of people are exposed to unsafe levels of air pollution.
Most OECD countries have achieved universal or near-universal health
coverage, but access to care needs to be improved
●● Populationcoverage for a core set of services is 95% or higher in all but seven OECD
countries and lowest in Greece, the United States and Poland.
●● Out-of-pocketpayments by households make up 20% of all health spending on average
in the OECD, and over 40% in Latvia and Mexico.
●● Costconcerns lead about 10% of people to skip consultations, while 7% do not purchase
prescribed medicines. Poorer households are most affected.
●● Thenumber of physicians per 1000 people is much higher in capitals and other cities,
with variation between areas most marked in the United States and the Netherlands.
●● Waiting times for elective surgery are long in a number of countries, particularly Estonia,
Poland and Chile.
Patient experiences and outcomes of care are improving, with lower
mortality rates after a heart attack or stroke and higher survival rates
for people with cancer
●● Over 80% of patients report positive experiences in terms of their time spent with a doctor,
easy-to-understand explanations and involvement in treatment decisions.
●● Avoidable hospital admissions for chronic conditions have fallen in most OECD countries,
indicating an improving quality of primary care.
●● In terms of acute care, fewer people are dying following heart attack or stroke. Improvements
are particularly striking among heart attack patients in Finland, and stroke patients in
Australia.
●● Timeliness of hip fracture surgery (a measure of patient safety) has improved in most
countries, with over 80% occurring within two days of admission.
●● Ratesof obstetric trauma have remained relatively unchanged, with tearing of the
perineum in 5.7% of instrument-assisted vaginal deliveries.
●● Across the OECD, five-year survival rates for breast cancer were 85% and just over 60%
for colon and rectal cancers, with survival rates improving in most countries over time.
●● Childhood vaccinations are near universal in most OECD countries, though measles
coverage has fallen slightly in Australia and Italy in recent years.
Having sufficient financial and material resources is critical to the functioning
of a health system. These resources need to be used wisely to avoid
ineffective spending
●● Spending on health in the OECD was about USD 4 000 per person on average (adjusted for
purchasing powers). The United States spends almost USD 10 000 per person.
10 Health at a Glance 2017 © OECD 2017
Executive summary
●● Health spending was 9% of GDP on average in the OECD, ranging from 4.3% in Turkey to
17.2% in the United States.
●● In all countries except the United States, government schemes and compulsory health
insurance are the main health care financing arrangements.
●● Hospitals account for nearly 40% of health spending.
●● Since 2000, the number of doctors and nurses has grown in nearly all OECD countries.
There are about three nurses per doctor, with the nurse-to-doctor ratio highest in Japan,
Finland and Denmark.
●● Hospital beds per capita have fallen in all OECD countries except Korea and Turkey, linked
to lower hospitalisation rates and increased day surgery.
●● Increased use of generics in most OECD countries has generated cost-savings, though
generics still represent less than 25% of the volume of pharmaceuticals sold in Luxembourg,
Italy, Switzerland and Greece.
●● Population ageing has increased the demand for long-term care, with spending increasing
more than for any other type of health care.
●● On average, 13% of people aged 50 and older provide weekly care for a dependent relative
or friend; 60% of informal carers are women.
Health at a Glance 2017 © OECD 2017 11
Reader’s guide
Reader’s guide
H ealth at a Glance 2017 presents comparisons of key indicators for health and health system
performance across the 35 OECD countries. Candidate and key partner countries are also
included where possible (Brazil, China, Colombia, Costa Rica, India, Indonesia, Lithuania,
the Russian Federation and South Africa). The data presented in this publication come from
official national statistics, unless otherwise stated.
Structure of the publication
The general framework underlying this publication assesses the performance of health
systems within the context of a broader view of public health (Figure 0.1). It is based on a
framework that has been endorsed and updated for the OECD Health Care Quality Indicators
project (see source to Figure 0.1). This framework recognises that the ultimate goal of
health systems is to improve the health status of the population. Many factors influence
health status, including those that fall outside health care systems, such as the physical
environment in which people live, and individual lifestyles and behaviours. The demographic,
economic and social context also affects the demand for and supply of health services, and
ultimately health status.
Conceptual framework for health system performance assessment
Health status
(dashboard 1, chapter 3)
Risk factors for health
(dashboard 2, chapter 4)
Health care system performance
How does the health system perform? What is the level of quality of care and access to services?
What does the performance cost?
Access Quality Health expenditure and financing
(dashboard 3, chapter 5) (dashboard 4, chapter 6) (dashboard 5, chapter 7)
Health care resources and activities (dashboard 5) Sub-sector analysis (dashboards 1 & 5)
Health workforce (chapter 8) Pharmaceutical sector (chapter 10)
Health care activities (chapter 9) Ageing and long-term care (chapter 11)
Demographic, economic & social context
Source: Adapted from Carinci, F. et al. (2015), “Towards Actionable International Comparisons of Health System
Performance: Expert Revision of the OECD Framework and Quality Indicators”, International Journal for Quality in Health
Care, Vol. 27, No. 2, pp. 137-146.
Health at a Glance 2017 © OECD 2017 13
Reader’s guide
At the same time, the performance of health care systems is clearly crucial. Core
dimensions of performance include the degree of access to care and the quality of care
provided. Performance measurement needs to take into account the financial resources
required to achieve these access and quality goals. Health system performance also depends
critically on the health workers providing services, and the goods and services at their
disposal.
Health at a Glance 2017 compares OECD countries on each component of this general
framework. It is structured around eleven chapters. The first two chapters offer an overview
of health and health system performance. The next nine chapters then provide detailed
country comparisons across a range of health indicators, including where possible time
trend analysis.
In Chapter 1, a series of dashboards present the relative strengths and weaknesses of
OECD countries’ health systems, alongside OECD-wide summary data. These dashboards use
a subset of the indicators that are presented in more detail in later chapters of the publication.
Chapter 2 provides a complementary thematic analysis on the determinants of health
across OECD countries. It assesses the relative contributions of health systems vis-à-vis
wider social factors to life expectancy.
Following these overview chapters, Chapter 3 on health status highlights variations
across countries in life expectancy, the main causes of mortality and other measures of
population health status. This chapter also includes measures of inequality in health status
by education and income level for key indicators such as life expectancy and perceived
health status.
Chapter 4 examines major risk factors for health. The focus is on health-related lifestyles
and behaviours, most of which can be modified by public health and prevention policies.
These include the major risk factors for non-communicable diseases of smoking, alcohol
and obesity, for children and adults. At the same time, healthy lifestyles are assessed in
terms of nutrition and physical activity. Population exposure to air pollution is also analysed.
Chapter 5 on access to care presents a set of indicators related to financial access,
geographic access and timely access (waiting times). This includes analysis of self-reported
unmet needs for medical care. Overall measures of population coverage are also presented.
Chapter 6 assesses quality and outcomes of care in terms of clinical effectiveness, patient
safety and the person responsiveness of care. The chapter seeks to reflect the lifecycle of
care by presenting indicators related to preventive, primary, chronic and acute care. This
includes analysis of patient experiences, prescribing practices, management of chronic
conditions, acute care for heart attack and stroke, patient safety, mental health, cancer care
and prevention of communicable diseases.
Chapter 7 on health expenditure and financing compares how much countries spend on
health, both on a per capita basis and in relation to GDP. The chapter analyses how health
care is paid for, through a mix of government funding, compulsory and voluntary health
insurance and direct out-of-pocket payments by households. The breakdown of spending
by health provider and by the type of health care provided is also examined.
Chapter 8 looks at the health workforce, particularly the supply and remuneration of
doctors and nurses. The chapter also presents data on the number of new graduates from
medical and nursing education programmes. It features indicators on the international
14 Health at a Glance 2017 © OECD 2017
Reader’s guide
migration of doctors and nurses, comparing countries in terms of their reliance on foreign-
trained workers as well as trends over time.
Chapter 9 on health care activities describes some of the main characteristics of health
service delivery. It starts with the number of consultations with doctors, often the “entry
point” of patients to health care systems. Country comparisons on hospital discharges
and lengths of stay, the utilisation rates of surgical procedures, and the increased use of
ambulatory surgery for minor surgeries are also included.
Chapter 10 takes a closer look at the pharmaceutical sector. Analysis of pharmaceutical
spending gives a sense of the varying scale of the market in different countries. The number
of pharmacists and pharmacies; consumption on certain high-volume drugs; and the use of
generics and bio-similars are also compared. Finally, spending on research and development
in the pharmaceutical sector is assessed.
Chapter 11 focuses on ageing and long-term care. It assesses key factors affecting the
current and future demand for long-term care. This includes demographic trends, and
health status indicators for elderly populations, such as life expectancy and self-reported
measures of health and disability at age 65. Dementia is compared across countries in terms
of prevalence today and in the future, and in terms of indicators for quality of care. The
recipients of long-term care and the formal and informal workers providing care for these
people are also assessed, as are trends in long-term care expenditure in different countries.
Presentation of indicators
With the exception of the first two chapters, indicators covered in the rest of the
publication are presented over two pages. The first page defines the indicator, provides
a brief commentary highlighting key findings conveyed by the data, and signals any
significant national variation from the definition which might affect data comparability.
On the facing page is a set of figures. These typically show current levels of the indicator
and, where possible, trends over time. Where an OECD average is included in a figure, it is
the unweighted average of the OECD countries presented, unless otherwise specified. The
number of countries included in this OECD average is indicated in the figure, and for charts
showing more than one year this number refers to the latest year.
Data limitations
Limitations in data comparability are indicated both in the text (in the box related to
“Definition and comparability”) as well as in footnotes to figures.
Data sources
Readers interested in using the data presented in this publication for further analysis
and research are encouraged to consult the full documentation of definitions, sources
and methods presented in OECD Health Statistics on OECD.Stat (http://stats.oecd.org/index.
aspx, then choose “Health”). More information on OECD Health Statistics is available at
http://www.oecd.org/health/health-data.htm.
Population figures
The population figures used to calculate rates per capita throughout this publication
come from Eurostat for European countries and from OECD data based on UN Demographic
Yearbook and UN World Population Prospects (various editions) or national estimates for
non-European OECD countries (data extracted as of early June 2017), and refer to mid-year
Health at a Glance 2017 © OECD 2017 15
Reader’s guide
estimates. Population estimates are subject to revision, so they may differ from the latest
population figures released by the national statistical offices of OECD member countries.
Note that some countries such as France, the United Kingdom and the United States
have overseas colonies, protectorates or territories. These populations are generally excluded.
The calculation of GDP per capita and other economic measures may, however, be based on
a different population in these countries, depending on the data coverage.
OECD country ISO codes
Australia AUS Korea KOR
Austria AUT Latvia LVA
Belgium BEL Luxembourg LUX
Canada CAN Mexico MEX
Chile CHL Netherlands NLD
Czech Republic CZE New Zealand NZL
Denmark DNK Norway NOR
Estonia EST Poland POL
Finland FIN Portugal PRT
France FRA Slovak Republic SVK
Germany DEU Slovenia SVN
Greece GRC Spain ESP
Hungary HUN Sweden SWE
Iceland ISL Switzerland CHE
Ireland IRL Turkey TUR
Israel ISR United Kingdom GBR
Italy ITA United States USA
Japan JPN
Partner country ISO codes
Brazil BRA Indonesia IDN
China CHN Lithuania LTU
Colombia COL Russian Federation RUS
Costa Rica CRI South Africa ZAF
India IND
16 Health at a Glance 2017 © OECD 2017
Health at a Glance 2017
© OECD 2017
Chapter 1
Indicator overview: OECD snapshots
and country dashboards
This chapter presents a set of selected indicators on health and health system
performance, designed to shed light on how well OECD countries perform along
five dimensions: health status, risk factors for health, access to care, quality and
outcomes of care, and health care resources. These indicators, taken from the main
chapters of the publication, are presented in the form of OECD snapshots and
country dashboards. The former illustrates time trends for the OECD as a whole,
together with a snapshot of the latest available data (OECD average, top and
bottom performers). The dashboards summarise how each country performs on all
indicators compared to the OECD average.
The selection of the indicators presented in this chapter was based on policy
relevance, data availability and ease of interpretation. The selection and comparison
of indicators is meant to capture relative strengths and weaknesses of countries
to help identify possible areas for priority action, though not to identify which
countries have the best health system overall.
The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli
authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights,
East Jerusalem and Israeli settlements in the West Bank under the terms of international law.
17
1. Indicator overview: OECD snapshots and country dashboards
OECD snapshots and country dashboards
Policy makers in OECD countries have a keen interest to understand how well their
health systems perform. A look at indicators contained in this publication shows that
significant progress has already been achieved. People in OECD countries are living longer
than ever before, with life expectancy at birth now exceeding 80 years on average, thanks
to improvements in living conditions and educational attainments, but also to healthier
lifestyles and progress in health care.
In most countries, universal health coverage provides financial protection against the
cost of illness and promotes access to care for the whole population. Quality of care has also
generally improved, as shown by the reduction in deaths after heart attacks and strokes,
and the earlier detection and improved treatments for diseases such as diabetes and cancer.
But these improvements have come at a cost: health spending now accounts for about
9% of GDP on average in OECD countries, and exceeds 10% in many countries. Higher health
spending is not necessarily a problem when the benefits exceed the costs, but there is ample
evidence of inequities and inefficiencies in health. There is also a need to achieve a better
balance between spending on curative care and disease prevention.
Despite these improvements, important questions remain about how successful
countries are in achieving good results on different dimensions of health system
performance. For example, what are the main factors explaining differences in health
status and life expectancy across OECD countries? Is the increase in the prevalence of
certain risk factors, such as obesity, offsetting some of the gains from the reduction in
other risk factors like smoking? To what extent can citizens benefit from adequate and
timely access to care, and good financial protection against the costs of health care? What
do we know about the quality and safety of care provided to people for a range of common
health conditions? What are the financial, human and technical resources allocated to
health systems in different countries?
Answering these questions is by no means an easy task, but the snapshots and
dashboards presented in this chapter can help shed light on how well countries do in
promoting the health of their population and on several dimensions of health system
performance. They do not have the ambition of identifying which countries have the overall
best health system; rather, they summarise some of the relative strengths and weaknesses
of OECD countries on a selected set of indicators of health and health system performance.
They can be useful to identify areas for priority action, but should be complemented by a
more in-depth review of the data and factors influencing cross-country variations, presented
in the main chapters of this publication.
18 Health at a Glance 2017 © OECD 2017
1. Indicator overview: OECD snapshots and country dashboards
This chapter presents five sets of indicators, which are discussed in full in the chapters
in parentheses, highlighting how well countries fare in each of the following dimensions:
●● Health status (Chapters 3 and 11)
●● Risk factors for health (Chapter 4)
●● Access to care (Chapter 5)
●● Quality of care (Chapter 6)
●● Health care resources (Chapters 7, 8 and 9)
For each of these dimensions, a set of 4-5 relevant indicators is presented in the form
of OECD snapshots and country dashboards. These indicators are selected from the publication
based on their policy relevance and importance as key factors to monitor in a health system,
but also on data availability and interpretability. Therefore, indicators for which country
coverage is highest are prioritised to improve comparability.
OECD snapshots, newly introduced, provide summary statistics for key indicators in
the five dimensions listed above. They complement the country dashboards by visualising:
●● the latest OECD average (for quick comparison with country figures in the dashboards)
●● the distribution of top and bottom values (for a general sense of the dispersion surrounding
each indicator)
●● the overall OECD trend since 2005 (to highlight changes over time)
The snapshots complement the country dashboards, helping the reader make a first
assessment of a country’s performance vis-à-vis the OECD average and value range before
delving into the more detailed indicator chapters of the publication.
Country dashboards, in the form of summary tables, compare a country’s performance
to one another and the OECD average. Countries are classified for each indicator into three
colour-coded groups:
●● Blue, when the country’s performance is within close distance of the OECD average
●● Green, when the country’s performance is considerably better than the OECD average
●● Red, when the country’s performance is considerably worse than the OECD average
The only exception to this grouping is for the dashboard on health care resources
(Table 1.5), where the indicators presented cannot be strictly classified as better or worse
performance. For this reason, the colour coding in this dashboard uses a lighter and darker
shade of blue to signal that a country is considerably below or above the OECD average.
Values for each indicator are shown for all countries and for the OECD as a whole.
Health at a Glance 2017 © OECD 2017 19
1. Indicator overview: OECD snapshots and country dashboards
Methodology, interpretation and use
OECD snapshots
For each indicator, the OECD average, highest and lowest values for the latest available year are shown,
corresponding to the data presented in the main chapters of the publication. The spark lines on the right
show OECD-wide trends in recent years. These are calculated based on an unweighted mean of the data
available for each year (data linearly interpolated when unavailable, for consistency regarding the number
of countries used for the means). These spark lines are intended to give a broad overview of trends, given
potential differences in methodology or country composition over time.
Country dashboards
The classification of countries as better, worse or within close distance of the OECD average is based on
each indicator’s standard deviation. This method is preferred to using a fixed percentage or fixed number of
countries per category, since it reflects variation (how far a country is from the OECD average) in a dynamic way.
The standard deviation is a common statistical indicator of variation in a distribution, measuring how close
values are to the central tendency. Countries are classified as “close to the OECD average” (blue) whenever
the value for any indicator is within one standard deviation from the OECD mean for the latest year. In rare
cases, particularly large outliers are excluded from the calculation of the standard deviation. These exceptions
are noted under the relevant dashboards.
For a typical indicator, about 65% of the countries (23 countries) will be close to the OECD average, with
the remaining 35% performing significantly better (green) or worse (red). When the number of countries that
are close to the OECD average is higher (lower) than 23, it means that cross-country variation is relatively
low (high) for that indicator. For example, indicators such as male life expectancy and alcohol consumption
show that 28 countries are close to the OECD average, meaning that countries show relatively less variation
compared to other indicators.
Health status
Life expectancy is a key indicator to understanding the overall health of a population. It
therefore is the focus of this section, with three indicators reflecting gender and age-specific
life expectancies. More specific indicators on ischaemic heart mortality and dementia
prevalence are also shown, two major causes of mortality and morbidity today and in the
future. Figure 1.1 provides a snapshot on health status across the OECD and Table 1.1 provides
more detailed country comparisons.
Figure 1.1. Snapshot on health status across the OECD
LOW OECD HIGH 2005 2015
Latvia Iceland 77.9
Life expectancy (male)
Years of life at birth 68 69.7 77.9 81.2 88 75.2
Mexico Japan 83.1
Life expectancy (female)
Years of life at birth 73 77.7 83.1 87.1 93
81.2
Hungary Japan 19.5
Life expectancy at age 65
Years of life 15 25
16.4 19.5 21.9 18.1
Ischaemic mortality Japan Latvia 160
Age-standardised rate per
100 000 population 0 34 112 328 400
112
Mexico Japan
Dementia prevalence
Cases per 1 000 population 5
7.2 14.8 23.3 25
Note: the Y-axis for OECD trends is standardised to have fixed height, based on the minimum and maximum values
of the indicator. The high-low X-axis is standardised with constant distance from the OECD average whenever the
indicator is not truncated at 0.
20 Health at a Glance 2017 © OECD 2017
1. Indicator overview: OECD snapshots and country dashboards
In general, Japan, France and Spain have the best overall health outcomes in terms
of life expectancy and ischaemic heart mortality. Hungary, Latvia, Mexico and the Slovak
Republic are consistently below the OECD average for these indicators. Across the OECD, life
expectancy has increased steadily over time, though about half of the countries reported
slight falls in life expectancy between 2014 and 2015. At the same time, some of the countries
with the highest rates of dementia prevalence are the countries with longer life expectancies,
such as Japan and Italy. Dementia prevalence also shows the greatest amount of variation
across countries, amongst these indicators.
Important variations in life expectancy by gender and age exist. Women in Japan,
Spain and France live much longer than the OECD average; while male life expectancy
is particularly high in Iceland. Life expectancy at 65 is noticeably lower than the OECD
average in 12 countries, and noticeably higher than the average in Japan, France and Spain.
Life expectancy is affected by a range of factors within and beyond the health system.
Higher health spending per capita is positively associated with life expectancy, though
this relationship is less pronounced in countries with the highest health spending, such as
Luxembourg, Norway, Switzerland and the United States. Differences in risky behaviours
such as smoking and obesity, which have a major impact on health, can also partly explain
cross-country variation and differences in life expectancies. Factors beyond the health
system are also important determinants of health, including income, education and other
socio-economic factors.
Ischaemic heart disease remains the highest cause of mortality in most OECD countries,
though there has been an average decline of more than 50% since 1990. Mortality rates
are considerably above the OECD average in five countries, and are highest in Latvia, the
Slovak Republic and Hungary; whereas they are relatively low in Japan, Korea, France and
the Netherlands.
The prevalence of dementia, a variety of brain disorders of which Alzheimer’s disease
is the most common form, is a core indicator to monitor the health of ageing populations.
Dementia prevalence is noticeably higher than the OECD average in seven countries and
highest in Japan, Italy and Germany. Countries with younger populations typically have
lower dementia prevalence, with Mexico, Turkey and the Slovak Republic having the lowest
rates of dementia.
Health at a Glance 2017 © OECD 2017 21
1. Indicator overview: OECD snapshots and country dashboards
Table 1.1. Dashboard on health status
✓ Better than OECD average
Close to OECD average Ð Worse than OECD average
– Missing data
LIFE LIFE LIFE ISCHAEMIC DEMENTIA
EXPECTANCY (M) EXPECTANCY (F) EXPECTANCY (65) MORTALITY PREVALENCE
Years of life at birth, Years of life at birth, Years of life at age 65, Age-standardised rate Cases per
males females total per 100 000 population 1 000 population
OECD 77.9 83.1 19.5 112 14.8
Australia 80.4 84.5 20.9 85 14.2
Austria 78.8 83.7 19.7 131 18.0
Belgium 78.7 83.4 19.9 54 18.0
Canada 79.6 83.8 20.2 93 13.0
Chile 76.5 81.7 18.5 64 11.7
Czech Republic 75.7 81.6 17.7
Ð 237
Ð 10.4 ✓
Denmark 78.8 82.7 19.4 60 16.4
Estonia 73.2
Ð 82.2 18.1
Ð 211
Ð 14.7
Finland 78.7 84.4 20.1 147 18.5
France 79.2 85.5 ✓
21.5 ✓
39 ✓
19.7
Ð
Germany 78.3 83.1 19.5 106 20.2
Ð
Greece 78.5 83.7 19.9 82 19.6
Ð
Hungary 72.3
Ð 79.0
Ð 16.4
Ð 288
Ð 10.6
Iceland 81.2 ✓
83.8 20.4 100 13.0
Ireland 79.6 83.4 19.7 127 11.5
Israel 80.1 84.1 20.2 64 10.5
Italy 80.3 84.9 20.6 84 22.5
Ð
Japan 80.8 87.1 ✓
21.9 ✓
34 ✓
23.3
Ð
Korea 79.0 85.2 20.3 38 ✓
9.6 ✓
Latvia 69.7
Ð 79.5
Ð 16.6
Ð 328
Ð 14.6
Luxembourg 80.0 84.7 20.4 59 13.3
Mexico 72.3
Ð 77.7
Ð 17.7
Ð 144 7.2 ✓
Netherlands 79.9 83.2 19.8 46 ✓
16.1
New Zealand 79.9 83.4 20.4 129 13.5
Norway 80.5 84.2 20.3 72 15.1
Poland 73.5
Ð 81.6 17.9
Ð 98 9.9 ✓
Portugal 78.1 84.3 19.9 55 19.9
Ð
Slovak Republic 73.1
Ð 80.2
Ð 16.9
Ð 291
Ð 8.3 ✓
Slovenia 77.8 83.9 19.5 82 11.8
Spain 80.1 85.8 ✓
21.0 ✓
53 19.4
Ð
Sweden 80.4 84.1 20.2 95 18.1
Switzerland 80.8 85.1 20.9 78 17.2
Turkey 75.3 80.7
Ð 17.8
Ð 146 8.0 ✓
United Kingdom 79.2 82.8 19.7 98 17.1
United States 76.3 81.2 19.3 113 11.6
Note: All data refer to 2015 or nearest year, except for dementia prevalence, which refers to 2017. Indicators are taken from Chapter 3 (life
expectancy, male and female; ischaemic mortality) and Chapter 11 (life expectancy at 65; dementia prevalence).
Source: OECD Health Statistics 2017; OECD analysis of data from the World Alzheimer Report 2015 and the United Nations (for prevalence
of dementia).
22 Health at a Glance 2017 © OECD 2017
1. Indicator overview: OECD snapshots and country dashboards
Risk factors for health
Smoking, alcohol consumption and obesity are three major risk factors for non-
communicable diseases. Population exposure to air pollution is also a critical non-medical
determinant of health. Figure 1.2 provides a snapshot on risk factors for health across the
OECD and Table 1.2 provides more detailed country comparisons.
Figure 1.2. Snapshot on risk factors for health across the OECD
LOW OECD HIGH 2005 2015
23.7
Daily smoking Mexico Greece
% of population (15 years+) who
smokes daily 0 7.6 18.4 27.3 37
18.4
9.7
Alcohol consumption Turkey Belgium
Liters per capita (15 years+) 9.0
consumed in a year 0 1.4 9.0 12.6 18
Obesity Japan United States 19.4
% of population (15 years+) with 16.5
BMI>30 0 3.7 19.4 38.2 40
Air pollution New Zealand Turkey 14.2 15.1
Mean annual population exposure
to PM 2.5, µg/m 3 0 5.5 15.1 36.4 40
Note: The Y-axis for OECD trends is standardised to have fixed height, based on the minimum and maximum values
of the indicator. The high-low X-axis is standardised with constant distance from the OECD average whenever the
indicator is not truncated at 0. Air pollution shows data for 2005 and 2010 to 2015.
In general, Iceland, Norway and Sweden perform well across these indicators. Smoking
rates are considerably below the OECD average in Mexico, Iceland, Sweden, the United States,
Norway and Australia; whereas they are much higher in Greece, Turkey, Hungary, Austria and
Latvia. Although most countries have managed to reduce smoking rates in recent decades,
there is still significant progress to be made. Tobacco smoking has been estimated by the
World Health Organization to kill 7 million people each year, yet, on average 18.5% of adults
still report daily smoking across the OECD.
Excessive alcohol consumption is also a considerable health burden, associated to
increased risk for a range of illnesses, including cancer, stroke, liver disease, as well as social
problems, with an estimated 2.3 million deaths per year. Populations in Belgium, Austria and
France consume considerably more litres per capita than the OECD average, while it is relatively
low in Turkey, Israel, Mexico and Norway. Alcohol consumption has been fluctuating over the
past 15 years, with a slight reduction across the OECD as a whole in average litres consumed
(based on sales figures). Binge drinking is of particular concern in certain countries, notably
Germany, Finland, Luxembourg and Denmark, and is more predominant among men.
Obesity is a major risk factor for many chronic diseases, including diabetes, cardiovascular
diseases and cancer. Obesity rates have been increasing in recent decades in almost all OECD
countries, with an average of 54% people overweight, of which 19% are obese. Obesity rates
are considerably higher than the OECD average in five countries, with rates highest in the
United States and Mexico. Obesity is lowest in Japan, Korea, Italy and Switzerland. The measure
reported here is for obese adults based on both measured and self-reported data. Caution
should be taken in comparing countries with reporting differences, since measured data is
generally much higher (and more accurate).
Health at a Glance 2017 © OECD 2017 23
1. Indicator overview: OECD snapshots and country dashboards
Air pollution is a major environmental threat, with health links to lung cancer, respiratory
and cardiovascular disease, low birth weight, dementia and other health problems.
Population exposure to fine particulates (PM2.5) is particularly high in Turkey, Korea, Poland
and Hungary. It is considerably below the OECD average in Australia, New Zealand, Sweden,
Canada, Finland and Iceland. While the overall trend since 1990 has been downward, there
have been some increases in population exposure to PM2.5 in more recent years.
Table 1.2. Dashboard on risk factors for health
✓ Better than OECD average
Close to OECD average Ð Worse than OECD average
– Missing data
SMOKING ALCOHOL OBESITY AIR POLLUTION
% of population who Liters per capita consumed % of population Mean annual exposure
smokes daily in a year with BMI > 30 to PM2.5, mg/m3
OECD 18.4 9.0 19.4 15.1
Australia 12.4 ✓
9.7 27.9
Ð 5.9 ✓
Austria 24.3
Ð 12.3
Ð 14.7 17.0
Belgium 18.9 12.6
Ð 18.6 15.6
Canada 14.0 8.1 25.8 7.2 ✓
Chile – 7.2 – 20.9
Czech Republic 18.2 11.5 18.7 21.4
Denmark 17.0 9.4 14.9 11.0
Estonia 21.3 10.3 18.0 9.4
Finland 17.4 8.5 24.8 7.4 ✓
France 22.4 11.9
Ð 17.0 12.4
Germany 20.9 11.0 23.6 14.0
Greece 27.3
Ð 7.5 17.0 13.5
Hungary 25.8
Ð 10.9 30.0
Ð 23.1
Ð
Iceland 10.2 ✓
7.5 19.0 7.8 ✓
Ireland 19.0 10.9 23.0 9.8
Israel 19.6 2.6 ✓
16.6 21.1
Italy 20.0 7.6 9.8 ✓
19.9
Japan 18.2 7.2 3.7 ✓
13.3
Korea 17.3 9.1 5.3 ✓
28.7
Ð
Latvia 24.1
Ð 10.8 23.2 20.4
Luxembourg 14.9 11.1 22.6 16.6
Mexico 7.6 ✓
5.2 ✓
33.3
Ð 20.2
Netherlands 19.0 8.0 12.8 14.6
New Zealand 14.2 8.9 31.6
Ð 5.5 ✓
Norway 12.0 ✓
6.0 ✓
12.0 ✓
9.1
Poland 22.7 10.5 16.7 24.3
Ð
Portugal 16.8 9.9 16.6 9.8
Slovak Republic 22.9 10.2 16.3 20.5
Slovenia 18.9 11.5 19.2 20.3
Spain 23.0 9.3 16.7 9.7
Sweden 11.2 ✓
7.2 12.3 ✓
6.2 ✓
Switzerland 20.4 9.5 10.3 ✓
12.9
Turkey 27.3
Ð 1.4 ✓
22.3 36.4
Ð
United Kingdom 16.1 9.5 26.9
Ð 12.4
United States 11.4 ✓
8.8 38.2
Ð 8.4
Note: All data refer to 2015 or nearest year. Indicators are taken from Chapter 4.
Obesity data reports a mix of measured and self-reported weights, with measured data often being higher and more accurate compared
to self-reported weight. Chapter 4 details the country coverage for each measure.
Source: OECD Health Statistics 2017; World Development Indicators (for air pollution).
24 Health at a Glance 2017 © OECD 2017
1. Indicator overview: OECD snapshots and country dashboards
Access to care
Access to care is a critical measure of health system performance. Indicators presented
here include population coverage, an overall measure of health care coverage, alongside
indicators reflecting financial and timely access. The access to care chapter also includes
geographic accessibility measures, not included here because of the complexity of cross-
country comparisons. Figure 1.3 provides a snapshot on access to care across the OECD and
Table 1.3 provides more detailed country comparisons.
Figure 1.3. Snapshot on access to care across the OECD
LOW OECD HIGH 2005 2015
Greece OECD 96.8 97.9
Population coverage
% of population covered by insurance for a
core set of services 80 86.0 97.9 100
3.0
Share of OOP expenditure France Switzerland
% of medical expenses on final household
consumption 0 1.4 3.0 5.3 7 2.8
Waiting times for cataract Netherlands Poland 133
Number of days from referral to procedure
0 37 121 464 500 121
Germany Poland
Doctor consultations skipped
Age-sex standardised rate per 100 0 40
2.6 10.5 33.0
(skipped due to cost)
Note: the Y-axis for OECD trends is standardised to have fixed height, based on the minimum and maximum values
of the indicator. The high-low X-axis is standardised with constant distance from the OECD average whenever the
indicator is not truncated at 0.
In terms of population coverage, most OECD countries have achieved universal (or near-
universal) coverage of health care costs for a core set of services, except for six countries
which remain considerably below the OECD average – Chile, Greece, Mexico, Poland, the
Slovak Republic and the United States.
Population coverage, though, is not sufficient by itself. The degree of cost-sharing
applied to those services also affects access to care. Out-of-pocket (OOP) expenditures and
consultations skipped due to cost are two indicators measuring financial access, which is
of particular concern for low-income population groups. OOP expenditures can create
financial barriers to health care. Across the OECD, they have made up a slightly increasing
share of household consumption over time, and are relatively high in Korea, Switzerland,
Greece, Hungary, Mexico, Portugal and Chile. The rate of consultations skipped due to cost is
particularly high in Poland, the United States and Switzerland (for the subset of 17 countries
with comparable data).
Long waiting times are also an important barrier to access in many OECD countries. They
are the result of a complex interaction between supply and demand of health services, with
doctors playing a crucial role on both sides. Long waiting times for elective (non-emergency)
surgery lead to patients suffering unnecessary pain and disability. Waiting times for cataract
surgery, one of the most commonly reported indicators, are particularly high in Poland and
Estonia (for the subset of 16 countries with comparable data), while numbers are very low
for Canada, Italy and the Netherlands.
Health at a Glance 2017 © OECD 2017 25
1. Indicator overview: OECD snapshots and country dashboards
Table 1.3. Dashboard on access to care
✓ Better than OECD average
Close to OECD average Ð Worse than OECD average
– Missing data
POPULATION SHARE OF OUT OF POCKET WAITING TIMES FOR CONSULTATIONS SKIPPED
COVERAGE EXPENDITURE CATARACT SURGERY ** DUE TO COST *
% of population covered % of final household Number of days from Age-sex standardised
by insurance consumption referral to procedure rate per 100 population
OECD 97.9 3.0 121 10.5
Australia 100 3.1 93 ✓
16.2
Austria 99.9 3.0 – –
Belgium 99.0 3.2 – –
Canada 100 2.2 58 ✓
6.6
Chile 92.1
Ð 4.1
Ð 103 –
Czech Republic 100 2.3 – –
Denmark 100 2.6 87 ✓
–
Estonia 94.0 2.7 253
Ð 9.7
Finland 100 3.0 103 –
France 99.9 1.4 ✓
– 8.5
Germany 100 1.8 ✓
– 2.6 ✓
Greece 86.0
Ð 4.4
Ð – –
Hungary 95.0 4.4
Ð 88 ✓
–
Iceland 100 3.1 – –
Ireland 100 2.5 – –
Israel 100 3.4 132 4.7
Italy 100 3.1 50 ✓
4.8
Japan 100 2.6 – –
Korea 100 5.1
Ð – –
Latvia – 3.9
Ð – –
Luxembourg 95.2 1.4 ✓
– –
Mexico 92.3
Ð 3.6 – –
Netherlands 99.9 2.4 37 ✓
10.3
New Zealand 100 2.1 73 ✓
14.5
Norway 100 2.9 117 5.9
Poland 91.0
Ð 2.5 464
Ð 33.0
Ð
Portugal 100 3.8
Ð 104 8.3
Slovak Republic 93.8
Ð 2.4 – –
Slovenia 100 2.0 – –
Spain 99.8 3.7 105 2.8 ✓
Sweden 100 3.3 – 3.9 ✓
Switzerland 100 5.3
Ð – 20.9
Ð
Turkey 98.4 – – –
United Kingdom 100 1.5 ✓
72 ✓
4.2 ✓
United States 90.9
Ð 2.5 – 22.3
Ð
* Poland is excluded from the standard deviation calculation. ** Estonia and Poland are excluded from the standard deviation calculation.
The values for Australia and Canada are reported in median number of days, rather than mean.
Note: Data on population coverage, share of OOP and waiting times refers to 2015, consultations skipped due to cost refer to 2016.
Indicators are taken from Chapter 5.
Source: OECD Health Statistics 2017; Commonwealth Fund International Health Policy Survey 2016 and other national sources.
26 Health at a Glance 2017 © OECD 2017
1. Indicator overview: OECD snapshots and country dashboards
Quality and outcomes of care
Measures of the quality and outcomes of care should reflect appropriateness of
care, clinical effectiveness, patient safety and the person responsiveness of care. The
appropriateness of care is measured by antibiotics prescribed and asthma/COPD admissions
as an indicator of avoidable admissions. 30-day mortality following acute myocardial
infarction (AMI) and colon cancer survival are indicators of clinical effectiveness; obstetric
trauma is a measure of patient safety. Figure 1.4 provides a snapshot on quality and outcome
of care across the OECD and Table 1.4 provides more detailed country comparisons.
Figure 1.4. Snapshot on quality and outcomes of care across the OECD
LOW OECD HIGH 2010 2015
241.7
Asthma/COPD admissions Japan Hungary
236.5
Age-sex standardised rate
per 100 000 population 0 58.1 236.5 427.5 470
20.6
Antibiotics prescribed Netherlands Greece
Defined daily dose per 1 000 20.0
population 0 10.7 20.6 36.1
40
8.5
AMI mortality Norway Mexico
Age-sex standardised rate 7.5
per 100 000 population 0 3.7 7.5 28.1 30
62.8
Colon cancer survival Chile Israel
Age-standardised survival in % 57.0
40 80
51.5 62.8 71.7
Obstetric trauma Poland Canada 5.9 5.7
Crude rates per 100 instrument-
assisted vaginal deliveries 0 0.7 5.7 16.9
20
Note: the Y-axis for OECD trends is standardised to have fixed height, based on the minimum and maximum values
of the indicator. The high-low X-axis is standardised with constant distance from the OECD average whenever the
indicator is not truncated at 0. Asthma/COPD admissions and antibiotics prescribed report 2011 as the baseline year.
Obstetric trauma reports 2010.
Asthma and COPD admissions are conditions for which effective treatment at the
primary care level is well established, but they vary significantly across countries. They
are considerably higher than the OECD average in Hungary, Turkey, Ireland, Australia,
New Zealand and Latvia; but much lower than the OECD average in Japan, Italy, Portugal,
Mexico and Chile. The number of antibiotics prescribed is higher than the OECD average in
Greece, France, Belgium and Italy. Antibiotic prescriptions are considerably below the OECD
average in the Netherlands, Estonia, Sweden, Latvia and Austria. The number of antibiotics
prescribed has increased slightly over time, with overuse of antibiotics not only a wasteful
use of resources, but also responsible for increased antimicrobial resistance.
Mortality following acute myocardial infarction (admission-based) is a long-established
indicator of the quality of acute care. It has been steadily declining since the 1970s in most
countries, yet important cross-country differences still exist. Mexico shows very high
mortality following AMI; rates are also relatively high in Latvia, Japan, Chile and Estonia.
Eight countries have mortality rates considerably below the OECD average, with Norway,
Australia and Denmark having the lowest rates.
Health at a Glance 2017 © OECD 2017 27
1. Indicator overview: OECD snapshots and country dashboards
Colon cancer survival rates vary relatively less than AMI, with only Israel and Korea
performing better than the average, and five countries performing considerably worse, with
Chile and Slovenia having the lowest rates.
Obstetric trauma (with instrument) is the most robust measure available for the dimension
of patient safety. For the subset of 21 countries with comparable data, obstetric trauma is
highest in Canada, followed by Sweden, Denmark and the United States. In contrast, rates
of obstetric trauma are considerably lower than the OECD average in Poland, Israel, Italy,
Slovenia and Portugal.
Table 1.4. Dashboard on quality of care
✓ Better than OECD average
Close to OECD average Ð Worse than OECD average
– Missing data
ASTHMA AND COPD ANTIBIOTICS ACUTE MYOCARDIAL COLON OBSTETRIC TRAUMA
HOSPITAL ADMISSIONS PRESCRIBED INFARCTION MORTALITY* CANCER SURVIVAL (INSTRUMENT) **
Age-sex standardised Age-sex standardised
Defined daily dose Age-standardised Crude rates per 100
rate per 100 000 rate per 100 000
per 1 000 population survival rate in % vaginal deliveries
population population
OECD 236 20.6 7.5 62.8 5.7
Australia 371
Ð 23.4 4.0 ✓
70.6 ✓
7.2
Austria 330 14.0 ✓
7.4 63.7 –
Belgium 286 29.2
Ð 7.0 67.8 3.4
Canada 247 20.8 5.1 ✓
67.2 16.9
Ð
Chile 99 ✓
– 11.3
Ð 51.5
Ð –
Czech Republic 193 19.6 6.9 56.1
Ð –
Denmark 333 16.1 4.0 ✓
61.6 10.9
Ð
Estonia 137 12.1 ✓
10.6
Ð 58.4 3.9
Finland 184 17.2 5.6 64.8 3.7
France 150 29.9
Ð 5.6 63.7 –
Germany 284 14.4 ✓
7.7 64.8 6.4
Greece – 36.1
Ð – – –
Hungary 428
Ð 17.0 – – –
Iceland 223 19.9 5.9 68.2 –
Ireland 411
Ð 25.6 6.4 60.5 4.2
Israel 259 21.4 6.7 71.7 ✓
1.9 ✓
Italy 64 ✓
27.5
Ð 5.4 64.1 1.9 ✓
Japan 58 ✓
– 11.7
Ð 67.8 –
Korea 309 24.3 8.1 71.6 ✓
–
Latvia 341
Ð 13.3 ✓
13.4
Ð 56.4
Ð –
Luxembourg 186 26.3 7.3 – –
Mexico 96 ✓
– 28.1
Ð – –
Netherlands 202 10.7 ✓
5.4 63.0 3.2
New Zealand 363
Ð 25.8 4.7 ✓
64.0 8.5
Norway 261 15.8 3.7 ✓
66.6 2.5
Poland 234 26.2 4.4 ✓
52.8
Ð 0.7 ✓
Portugal 74 ✓
21.3 7.9 60.9 2.5 ✓
Slovak Republic 238 24.5 6.4 51.7
Ð –
Slovenia 146 14.5 6.1 61.9 2.1 ✓
Spain 234 21.6 7.9 63.3 4.8
Sweden 184 12.3 ✓
4.2 ✓
64.9 11.3
Ð
Switzerland 138 – 5.1 ✓
67.2 7.4
Turkey 414
Ð 17.3 8.6 54.6
Ð –
United Kingdom 303 20.1 7.1 60.0 6.8
United States 262 – 6.5 64.9 9.6
Ð
Note: All data refer to 2015 or nearest year. Indicators are taken from Chapter 6.
* Mexico is excluded from the calculation of the standard deviation. ** Canada is excluded from the calculation of the standard deviation.
Source: OECD Health Statistics 2017.
28 Health at a Glance 2017 © OECD 2017
1. Indicator overview: OECD snapshots and country dashboards
Health care resources
Having sufficient health care resources is critical to the functioning of health systems.
But higher resources do not automatically translate into better health outcomes – the
effectiveness of spending is also important. Health care expenditure per capita is the most
immediate summary measure of health care resources. The supply of health workers (doctors
and nurses) and hospital beds are also reported, since higher health spending is not always
closely related to these indicators. Figure 1.5 provides a snapshot on health care resources
across the OECD and Table 1.5 provides more detailed country comparisons.
Figure 1.5. Snapshot on health care resources across the OECD
LOW OECD HIGH 2005 2015
4.0K
Health care expenditure Mexico United States
Total current expenditure per
capita, USD PPP 1.0K 4.0K 9.8K
2.4K
3.4
Doctors per capita Turkey Greece
Number of practising physicians
per 1 000 population 0 1.8 3.4 6.3 8
3.0
9.0
Nurses per capita Turkey Switzerland
Number of practising nurses per
1 000 population 0 2.0 9.0 18.0 20
8.0
5.3
Beds per capita Mexico Japan
Number of beds per 1 000
population 0 1.5 4.7 13.2 15
4.7
Note: the Y-axis for OECD trends is standardised to have fixed height, based on the minimum and maximum values
of the indicator. The high-low X-axis is standardised with constant distance from the OECD average whenever the
indicator is not truncated at 0.
In general, countries with higher health spending and higher numbers of health workers
and other resources have better health outcomes, quality and access to care. However, the
absolute number of resources invested is not a perfect predictor of better outcomes – efficient
use of health resources is also critical.
In terms of overall health care expenditure, the United States spends considerably
more per person than any other country. Health care spending is also high in Switzerland,
Luxembourg and Norway. Nine countries spend less than the OECD average, with health
spending per capita lowest in Mexico, Turkey and Latvia. Health spending has been
consistently growing in all countries over the past decades, other than a slowdown following
the financial crisis. Looking at growth rates of spending as a share of GDP, in addition to
absolute levels of spending, can give a better perspective on how much countries spend
relative to the general economy.
A large part of health spending is translated into wages for the workforce. The number
of doctors and nurses in a health system is therefore an important way of monitoring how
resources are being used. The number of doctors per capita is relatively high in Greece, Austria,
Portugal and Norway. Among these countries, Greece has one of the lowest numbers of
nurses per capita, suggesting the potential to decrease the doctors to nurses ratio. This could
generate significant cost savings in the long run. In contrast, Norway has one of the highest
numbers of nurses (Austria and Portugal nurses per capita are close to the OECD average).
Health at a Glance 2017 © OECD 2017 29
1. Indicator overview: OECD snapshots and country dashboards
Nurses per capita are particularly high in Switzerland, Germany and Nordic countries. While
the total number of nurses has grown more than doctors in absolute terms, both have grown
at similar rates in recent years, at around 13%.
Hospitals also take an important share of health care resources, with hospital beds
per capita a marker of the physical and technical resources available in a health system.
Reductions in the number of beds in many OECD countries over the past years have been
a voluntary effort to encourage a shift to day surgery and primary care. Nevertheless, the
number of beds per capita remains particularly high in Japan and Korea.
Table 1.5. Dashboard on health care resources
Above OECD average
Close to OECD average Below OECD average
– Missing data
HEALTH CARE DOCTORS PER NURSES PER BEDS PER
EXPENDITURE * CAPITA CAPITA CAPITA **
Total spending per capita, Number of practising pysicians Number of practising nurses Number of beds
USD PPP per 1 000 population per 1 000 population per 1 000 population
OECD 4 003 3.4 9.0 4.7
Australia 4 708 3.5 11.5 3.8
Austria 5 227 5.1
8.1 7.6
Belgium 4 840 3.0 10.8 6.2
Canada 4 753 2.7 9.9 2.6
Chile 1 977
2.1
2.1
2.1
Czech Republic 2 544 3.7 8.0 6.5
Denmark 5 199 3.7 16.7
2.5
Estonia 1 989
3.4 6.0 5.0
Finland 4 062 3.2 14.7
4.4
France 4 600 3.3 9.9 6.1
Germany 5 551 4.1 13.3
8.1
Greece 2 223
6.3
3.2
4.3
Hungary 2 101
3.1 6.5 7.0
Iceland 4 376 3.8 15.5
3.1
Ireland 5 528 2.9 11.9 3.0
Israel 2 822 3.4 4.9 3.0
Italy 3 391 3.8 5.4 3.2
Japan 4 519 2.4
11.0 13.2
Korea 2 729 2.2
5.9 11.5
Latvia 1 466
3.2 4.7 5.7
Luxembourg 7 463
2.9 11.9 4.8
Mexico 1 080
2.4
2.8
1.5
Netherlands 5 385 3.5 10.5 4.2
New Zealand 3 590 3.0 10.3 2.7
Norway 6 647
4.4
17.3
3.8
Poland 1 798
2.3
5.2 6.6
Portugal 2 734 4.6
6.3 3.4
Slovak Republic 2 150
3.5 5.7 5.8
Slovenia 2 835 2.8 8.8 4.5
Spain 3 248 3.9 5.3 3.0
Sweden 5 488 4.2 11.1 2.4
Switzerland 7 919
4.2 18.0
4.6
Turkey 1 088
1.8
2.0
2.7
United Kingdom 4 192 2.8 7.9 2.6
United States 9 892
2.6 11.3 2.8
Note: All data refer to 2015 or nearest year, except for health care expenditure, which refers to 2016. Indicators are taken from Chapter 7
(health expenditure), Chapter 8 (doctors and nurses per capita) and Chapter 9 (beds per capita).
* United States is excluded from the standard deviation calculation. ** Japan and Korea are excluded from the standard deviation
calculation. For Ireland, private hospitals beds are excluded.
Source: OECD Health Statistics 2017.
30 Health at a Glance 2017 © OECD 2017
Health at a Glance 2017
© OECD 2017
Chapter 2
What has driven life expectancy gains
in recent decades? A cross-country
analysis of OECD member states
Countries with higher national income and health spending tend to have longer
life expectancies. But these factors can only account for a part of life expectancy
differences across countries. This chapter analyses the factors contributing to health
status, including a closer assessment of the determinants of health that go beyond
the health system. It shows that on average, a 10% increase in health spending per
capita is associated with a gain of 3.5 months of life expectancy. The same rate of
improvement in healthier lifestyles (10%) is associated with a gain of 2.6 months
of life expectancy. Wider social determinants are also important: a 10% increase in
income per capita is associated with a gain of 2.2 months of life expectancy, and a
10% increase in primary education coverage with 3.2 months. For income, minimum
absolute levels are particularly critical to protecting people’s health.
The main policy implication emerging from this analysis is the significant opportunities
for health improvement from coordinated action across ministries responsible for
education, the environment, income and social protection, alongside health ministries.
This includes inter-sectoral action to address health-related behaviours. Collaboration
with the private sector will also be important, especially with employers in relation to
working conditions.
31
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
Introduction
Life expectancy has risen steadily in most OECD countries, increasing over ten years on
average since 1970. Mortality rates from the main causes of death, cardiovascular diseases
and cancer, have generally fallen. Today, countries with higher national income and health
spending tend to have longer life expectancies. But these factors can only account for a part
of life expectancy differences across countries. Furthermore, life expectancy varies across
population groups. For example, life expectancy is lower amongst individuals with lower
levels of education across all OECD countries (Murtin et al., 2017).
This chapter explores the determinants of life expectancy gains in OECD countries. These
include drivers beyond the health system – the demographic, economic and social context
– alongside health system factors. Such analysis complements subsequent chapters in this
Health at a Glance edition, which focus predominantly on cross-country comparisons of
health care system performance. Referring back to the conceptual framework underpinning
Health at a Glance, this chapter analyses the factors contributing to health status, including
a closer assessment of the determinants of health that go beyond the health system
(Figure 2.1).
Figure 2.1. Determinants of health and the Health at a Glance
conceptual framework
Health status
(dashboard 1, chapter 3)
Risk factors for health
(dashboard 2, chapter 4)
Health care system performance
How does the health system perform? What is the level of quality of care and access to services?
What does the performance cost?
Access Quality Health expenditure and financing
(dashboard 3, chapter 5) (dashboard 4, chapter 6) (dashboard 5, chapter 7)
Health care resources and activities (dashboard 5) Sub-sector analysis (dashboards 1 & 5)
Health workforce (chapter 8) Pharmaceutical sector (chapter 10)
Health care activities (chapter 9) Ageing and long-term care (chapter 11)
Demographic, economic & social context
Drivers beyond the health system, including income,
education, working and living conditions
32 Health at a Glance 2017 © OECD 2017
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
Analysis is based on country-level data for the time period 1995-2015, and covers all
35 OECD member states. Empirical findings are complemented by an assessment of the
mechanisms by which drivers within and beyond the health system affect health.
Understanding the determinants of health
Health outcomes depend on investments both within and beyond the health system
Biological endowment and health service availability are not sufficient to explain
differences in individuals’ health. But a growing body of evidence has demonstrated that
an individual’s health also depends on factors that go beyond the medical care received
(Marmot and Wilkinson, 2006; WHO, 2008). Some of these factors can still be influenced
by health systems directly, through public health and prevention measures. In particular,
non-medical determinants related to lifestyle choices are important. These include major
risk factors such as smoking, alcohol and unhealthy diet, and conversely health-seeking
activities such as physical activity.
But broader social determinants of health also matter. Income, education, working
and living conditions are all also important factors. Having a sufficient income allows
people to purchase essential goods and services that sustain or improve health, such as
nutritious food and shelter; though higher income can also involve longer work hours
and greater stress (Fuchs, 2004). The more educated, as well as often being richer, may be
better informed about health-seeking activities (Mackenbach et al., 2008). Unemployment
and poor working conditions adversely affect mental health, and certain occupations
carry a greater risk of injury (Bassanini and Caroli, 2014). Living in an unsanitary, unsafe
or polluted environment also increases the risk of illness or death (Gibson et al., 2011;
Deguen and Zmirou-Navier, 2010).
The social determinants of health are closely inter-linked. Indeed, this makes it hard
to empirically disentangle the individual effects of different factors on health (Fuchs, 2004).
But what is evident is that these factors will, in general, reinforce each other. For example,
the better educated are also likely to be richer, live in healthier environments, and be less
likely to smoke. Further, some researchers argue that large income differences not only
cause health inequalities, but may also be detrimental to population health (Pickett and
Wilkinson, 2015). Finally, health inequalities are likely to persist over the life cycle and
across generations, with early life circumstances influencing future health and economic
prospects.
Further, despite the fact that most OECD countries have achieved universal health
coverage, individuals from the most disadvantaged groups tend to have worse access to
health services. For example, some individuals may be unaware or unwilling to use the full
range of health services available to them. Quality of care may be worse in more socially
deprived areas; co-payments and other direct payments by users without effective exemption
mechanisms will disproportionately affect the poor (OECD, 2014, 2015a).
Studies using aggregated data highlight the contribution of socio-economic factors
to health
A range of studies have estimated an empirical ‘health production function’ using
aggregated data. Such analyses have been used to assess the contribution of health spending,
socio-economic and other factors on population health.1
Health at a Glance 2017 © OECD 2017 33
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
In general, health spending, income and education have significant beneficial impacts
on population health (Berger and Messer, 2002; OECD, 2010; Heijink et al., 2013; Moreno-
Serra and Smith, 2015); with pollution and lifestyle factors (particularly smoking and alcohol
consumption) typically having significant adverse effects (Shaw, 2005; Blázquez-Fernádez
et al., 2013). Far fewer studies have incorporated variables reflecting unemployment,
occupational category or income inequality, and when included they have had more mixed
results (Or, 2000; Lin, 2009).
Note that health spending and income have typically had a stronger impact on reducing
avoidable mortality or infant mortality than on increasing life expectancy (Heijink et al., 2013;
Nixon and Ulmann, 2006). Dynamic factors may also be important. For example, temporary
economic downturns have shown more mixed effects on health outcomes, worsening
mental health but also potentially reducing mortality through reduced traffic fatalities and
possibly lower pollution (Ruhm, 2012; van Gool and Pearson, 2014; Laliotis et al., 2016). More
generally, differences in the countries analysed explains variability in the impact of different
factors on health outcomes.
Gains in life expectancy over time reflect increased health spending,
healthier lifestyles and improving socio-economic conditions
All OECD and partner countries have experienced gains in life expectancy over time,
but the rate of increase varies markedly across countries
Life expectancy at birth increased in all the countries analysed. Gains have been
particularly rapid in Turkey, India, Korea and China, countries which have had sustained
periods of economic growth alongside improved health care coverage (Figure 2.2). In the
United States and Mexico, gains have been more modest. There has also been slower
progress in South Africa (due mainly to the epidemic of HIV/AIDS), Lithuania and the Russian
Federation (due mainly to the impact of the economic transition in the 1990s and a rise in
risk increasing behaviors among men). Life expectancy at birth is currently the highest in
Japan, at 83.9 years.
Figure 2.2. Trends in life expectancy at birth, selected countries, 1970-2015
Japan Korea United States
China Turkey India
Years
90
85
80
75
70
65
60
55
50
45
40
0
74
00
02
04
06
08
10
12
14
7
7
7
8
8
9
9
9
8
9
19
20
20
20
19
19
19
19
19
19
19
19
19
19
20
20
19
19
19
20
20
19
20
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602139
34 Health at a Glance 2017 © OECD 2017
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
Increased health care spending had a strong positive impact on life expectancy,
but wider social determinants are also important
New analysis provides estimates of the relative contribution of health systems and
healthy lifestyles vis-à-vis socio-economic, and environmental factors across OECD countries.
This analysis uses the latest cross-country data and follows best methodological practices
(Box 2.1). Life expectancy gains from 1995 to 2015 are assessed. Data on explanatory factors
were lagged by five years (i.e. using data from 1990 to 2010) to account for the delayed effects
on health.
Box 2.1. Data and methods
The analysis assessed the relative contribution of factors within and beyond the health
system to life expectancy gains between 1995 and 2015 in all 35 OECD countries. Macro-level
panel data from OECD Health Statistics and the World Bank Databank was used.
An empirical health production function was developed, taking the following general form:
LEi,t = α i + β1Wi,t −5 + β 2 Xi,t −5 + β3 Yi,t −5 + β 4 Zi,t −5 + ei,t
where LEi,t is the life expectancy at birth for country i in year t; α the country effect; and
e is the error term. Explanatory variables are 5-year lagged in order to capture the delayed
effects of key determinants on life expectancy, with variable selection based on key
determinants identified in the literature. Lags of 5 years were chosen to strike a balance
between accounting for delayed effects on health and maintaining a sufficient number of
observations for the time-series analysis.
W is a vector of health system variables in year t-5 (health care spending, including both
curative and preventive care, measured by total health expenditure expressed in per capita
constant USD PPP; financial protection using the share of out-of-pocket spending in total
health expenditure as a proxy). X is a vector of lifestyle factors in year t-5 (prevalence of daily
smokers; alcohol consumption in litres per capita; healthy diet, measured by the share of the
population consuming vegetables daily). Y is a vector of income and other socio-economic
variables in year t-5 (income measured by GDP per capita at constant USD PPP, net of total
health expenditure; education measured as the share of the population attaining above
primary school education; and the long-term unemployment rate). Z is an environmental
variable in year t-5 (air pollution measured by the share of the population exposed to fine
particulates PM2.5).
A Cobb-Douglas production function is used, where all variables are expressed in
logarithmic form. The general econometric specification is a GLS model with country
fixed effects, country-specific autocorrelation structures for errors, a correction for
heteroscedasticity, and lagged explanatory variables. Data gaps in specific years were
addressed using linear interpolation. Further empirical models are examined in a related
working paper (James et al., forthcoming). Although the analysis follows best methodological
practice, associations between life expectancy and explanatory variables do not guarantee
causality.
Results from this analysis show that increased health spending, healthier lifestyles,
higher incomes and better education coverage over time have positive and statistically
significant associations with life expectancy gains (Figure 2.3). In particular, a 10% increase
in health spending per capita (in real terms) is associated with a gain of 3.5 months of life
expectancy. The same rate of improvement in healthier lifestyles (10%) is associated with
Health at a Glance 2017 © OECD 2017 35
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
a gain of 2.6 months of life expectancy (fewer smokers with 1.6 months, decreased alcohol
use with 1.0 month). Wider social determinants also matter. A 10% increase in income
per capita (in real terms) is associated with a gain of 2.2 months of life expectancy, and a
10% increase in primary education coverage with 3.2 months.
The share of out-of-pocket spending in total health spending did not have a significant
association with life expectancy gains, mainly because of its very small reduction over the
time period studied. Healthy diet had a positive but not significant association with life
expectancy. This may be explained by the very limited improvements to people’s diet over
time, and the difficulty to capture nutritional effects at the macro level. The association
between long-term unemployment rates and life expectancy was also not significant.2
More surprisingly, air pollution was also not significantly associated with life expectancy
gains, despite there being clear evidence elsewhere of the adverse effects of air pollution
on health (OECD 2016). This result reflects the long lag in time before air pollution affects a
person’s health, and also the relatively small decreases in air pollution over time in many
OECD countries. These results are explored further in a related working paper (James et al.,
forthcoming).
Figure 2.3. Life expectancy gains associated with a 10% change in the main
determinants of health
Analysis based on 35 OECD countries for the time period 1995-2015
Health spending 3.5
Out-of-pocket spending °
Smoking 1.6
Alcohol 1.0
Healthy diet °
Income 2.2
Education 3.2
Unemployment °
Air pollution °
0 1 2 3 4
Months
Note: ◦ stands for a contribution near zero.
12 http://dx.doi.org/10.1787/888933602158
While the effect on life expectancy of a 10% change in the main determinants of
health is useful for comparative purposes, in practice larger changes may be feasible,
leading to larger life expectancy gains. For example, if smoking rates and alcohol
consumption could be halved, together these could lead to a gain of 13 months of life
expectancy. Figure 2.4 illustrates the impact of more ambitious changes for selected
factors, notably a doubling of health spending and income, primary education coverage
reaching 100%, and more marked improvements in healthy lifestyles (a halving of
smoking rates and alcohol consumption).
The actual evolution in the main determinants of health over the past 20 years has
often been much more substantial than the 10% change used in Figure 2.3. From a policy
perspective, this is relevant because it means the positive impacts on life expectancy can
be substantial – given the right investments within and beyond the health system.
36 Health at a Glance 2017 © OECD 2017
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
Figure 2.4. Life expectancy gains from more substantial changes in the main
determinants of health
Analysis based on 35 OECD countries for the time period 1995-2015
Health spending 35.2
Smoking 8.1
Alcohol 4.9
Income 22.4
Education 23.8
0 10 20 30 40
Months
Note: Figures represent the gains in life expectancy that could be expected with doubling health spending, doubling
income, reaching 100% of tertiary education, and halving smoking and alcohol use. Unemployment, healthy diet, out-
of-pocket spending and air pollution are excluded because they were not statistically significant.
12 http://dx.doi.org/10.1787/888933602177
Figure 2.5 shows the percentage change of these determinants of health between
1990 and 2010. For example, while a 10% increase health spending is associated with a
gain of 3.5 months of life expectancy, health spending actually grew by 98% from 1990
to 2010 (from USD PPP 1 624 in 1990 to USD PPP 3 212 in 2010 in constant terms). Income
increased by 42% over the same time period, and education coverage by 44%. Improvements
in healthy lifestyles have been less marked: smoking rates were reduced by 31%, but
alcohol use only fell by 8% and the rate of daily vegetable consumption only increased by
2% from 1990 to 2010.
Figure 2.5. Evolution of the main determinants of life expectancy:
OECD 1990 to 2010
Health spending 98
Out-of-pocket spending -9
Smoking -31
Alcohol -8
Healthy diet 2
Income 42
Education 44
Unemployment 14
Air pollution -14
-40 -20 0 20 40 60 80 100 120
Growth %
12 http://dx.doi.org/10.1787/888933602196
As a result of the evolution of these determinants over time, health spending has been
the major contributing factor to gains in life expectancy over the last two decades, followed
by education then income (Table 2.1). The contributions of lifestyle factors (smoking, alcohol,
Health at a Glance 2017 © OECD 2017 37
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
healthy diet) have been smaller, largely because there have been smaller improvements in
these factors over the time period studied. Table 2.1 also shows regression coefficients and
values for 1990 and 2010, alongside the relative contributions of each of these determinants
of life expectancy.
Table 2.1. Determinants of life expectancy gains over time: regression
coefficients, relative contributions, 1990 and 2010 values
Regression Contribution to life
Explanatory variables 1990 value 2010 value
coefficient expectancy (months)
Health system factors
Health expenditure (per capita in constant USD PPP) + 0.039* 42.4 1 624 3 212
Out-of-pocket spending (as % of health expenditure) ns ns 22 20
Lifestyle factors
Smoking (% daily smokers) - 0.018* 5 30.3 21
Alcohol (litres of pure alcohol per capita) - 0.011* 0.4 10.1 9.2
Healthy diet (% daily consumers of vegetables) ns ns 64.2 65.3
Income and other socio-economic factors
Income (GDP per capita in constant USD PPP) + 0.025* 13.4 22 479 31 900
Education (% with above primary education) + 0.035* 15.1 57 82
Unemployment (% long-term unemployed) ns ns 3.2 3.6
Environmental factors
Air pollution (% of population exposed to PM2.5) ns ns 75.7 65.2
Note: * statistically significant at the 5% level, ‘ns’ means not significant. Regression based on 718 observations across
35 countries. The sum of the contributions and the residual (not shown here) is equal to the total gain of life years
over the studied period.
Supplementary analyses were carried out to test a range of common econometric
specification issues, as well as alternative explanatory variables. These analyses showed
consistent results (see James et al., forthcoming). Additional analysis adding OECD partner
countries to the sample shows some differences in the determinants of health by a country’s
level of economic development. For high-income countries, health care spending has been
the main driver of life expectancy gains, whereas income was the main driver in emerging
economies. This analysis, though, was limited by data only being available for a shorter
time period.
Most OECD countries have steadily increased health care spending in recent
decades, but accompanying gains in life expectancy vary markedly across countries
While empirical analysis showed that health care spending has made a marked
contribution to life expectancy gains across OECD countries as a whole, there are important
cross-country differences. These are illustrated in Figure 2.6, which shows the trajectories
of life expectancy gains alongside increase in health expenditure since 1995 for selected
high-income countries.
In all OECD countries, both life expectancy and health spending have been increasing
over time. But these rates of increase vary significantly across countries. The notable outlier
is the United States, where health spending has increased far more rapidly over time than
in other OECD countries, yet life expectancy gains have been smaller. On the other hand,
life expectancy at birth in Japan has reached almost 84 years, but health expenditure per
capita is less than half of the United States.
38 Health at a Glance 2017 © OECD 2017
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
Figure 2.6. Life expectancy gains and increased health spending, selected
high-income countries, 1995-2015
Japan Italy France Australia
United Kingdom Canada Netherlands Germany
Norway United States
Life expectancy at birth
84
83
82
81
80
79
78
77
76
75
0 1 000 2 000 3 000 4 000 5 000 6 000 7 000 8 000 9 000 10 000
Health expenditure per capita (2010 USD, adjusted using 2010 PPPs)
12 http://dx.doi.org/10.1787/888933602215
These varying trajectories for health expenditure and life expectancy across countries
over time suggest the critical role healthy lifestyles and the wider social determinants of
health have in increasing life expectancy. But these trajectories also point to the importance
of improving value for money in health systems. This includes placing greater emphasis
on health promotion and other highly cost-effective interventions, but also eliminating
ineffective spending and waste (see OECD, 2017 for an in-depth discussion).
Unpacking the mechanisms by which socio-economic factors and a person’s
living environment affect health is essential for policy
The empirical results presented offer insights on the strength and relative contribution
of different determinants of health. This section complements the macro-level analysis
by assessing exactly how socio-economic factors and a person’s living environment affect
health and health-seeking behaviours, drawing on insights from more micro-level evidence.
The nature of income trajectories matter
The positive association between income and health is an important general finding. But
examining how different income trajectories influence health status offers further guidance
for policymakers. A first observation is the importance of minimum absolute levels of income.
Whereas low income and poverty has a clear detrimental effect on health, health differences
between individuals with average or high income are far less pronounced (Deaton, 2003). In
other words, there is a non-linear relationship between income and health.
Second, whilst current income matters, long-term income has a much greater impact on health.
That is, it takes time for higher (lower) incomes to have a beneficial (adverse) effect on health.
For example, studies in the United Kingdom concluded that persistent poverty carries a much
greater health risk than occasional episodes, and income level appears more important than
income change (Benzeval and Judge, 2001; Contoyannis et al., 2004).
Third, income reductions generally seem to have a larger impact on health than income gains,
irrespective of whether they are temporary or more permanent (O’Donnell et al., 2013).
Health at a Glance 2017 © OECD 2017 39
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
For example, McInerney et al. (2013) found that wealth losses following the 2008 global
financial crisis led to increased depression and use of antidepressants in the United
States. In contrast, they observed no health improvements from wealth gains in the
same study sample. In Sweden, self-assessed health responded to decreases in income
to a greater extent than to income gains over time (Miething and Aberg-Yngwe, 2014).
Similarly, most (but not all) studies of sudden wealth gains from inheritance, the stock
market and lotteries find limited or no evidence of associated improvements in health
status (O’Donnell et al., 2013).
Indeed, income payments can trigger adverse health events in some circumstances,
probably reflecting an increase in more risky behaviours. For example, Dobkin and Puller
(2007) found elevated drug-related admissions and within-hospital mortality in California for
recipients of federal disability payments around the time of payment. Evans and Moore (2011)
found increased risks of traffic accidents and heart attacks immediately after social security
payments, wage payments for military personnel, tax rebates and dividend payments.
Unemployment worsens mental and physical health; employment conditions
are also important
As discussed earlier, macro-level studies of unemployment on health find mixed effects.
In contrast, micro-level studies more consistently find that being unemployed adversely affects
both mental and physical health. For example, a meta-analysis of studies using individual data
found that unemployment is associated with a 63% higher risk of mortality after controlling
for age and other control factors (Roelfs et al., 2011), although this may partly reflect pre-
existing health conditions. Unemployment also affects mental health. In Australia, Canada
and the United Kingdom, evidence from panel data shows that changing from employment
to unemployment significantly increased mental distress (Llena-Nozal, 2009).
Employment conditions also matter. Working longer hours are harmful to health, raising
general stress levels but also increasing the risk of stroke and coronary heart disease
(Kivimaki et al., 2015). In extreme cases, it may raise the risk of major accidents (Harrington,
2001). Choice over working hours has also been shown to be crucial, irrespective of the
number of hours worked (Bassanini and Caroli, 2014). Other aspects of job quality are also
important. Exposure to hazardous substances and risk of injury is typically concentrated
amongst low-skilled menial labour (Clougherty et al., 2013). Job insecurity and job
dissatisfaction has also been shown to adversely affect health (Caroli and Godard, 2014;
Datta Gupta and Kristensen, 2008).
Education encourages healthier lifestyles
Better educated individuals and their offspring are healthier, independent of income
and employment-related effects. A large part of this difference has been attributed to
healthier lifestyles. In particular, the more educated are typically better informed about the risks
and benefits of different behaviours, but also more likely to process and act upon this information.
For example, people with lower education levels are more likely to smoke, be obese, have
less well-balanced diets and be less physically active (Mackenbach et al., 2008; Cutler and
Lleras-Muney, 2010). The evidence on alcohol, however, is more mixed. A recent OECD report
found that in general better educated women were more likely to drink excessively, though
the opposite held true for men (OECD, 2015b). At the same time, alcohol-related harm is
more prevalent among less educated and low-income groups, partly because of multiple
comorbidities (coexisting risk factors) and lower access to health care.
40 Health at a Glance 2017 © OECD 2017
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
The better educated are also more knowledgeable about exactly which health services are
available to them, with consequently greater use of certain services. This is particularly
noticeable in terms of use of preventive health services and specialist consultations (OECD,
2006). Further, education may improve self-management (and therefore the efficacy) of medical
treatment, particularly for chronic diseases (Goldman and Smith, 2002).
Disadvantaged population groups are more likely to experience inadequate living
conditions, and adverse health effects from pollution
Air pollution was not significantly associated with life expectancy changes in the
empirical analysis presented earlier, principally due to there being rather small decreases
in air pollution over time in many OECD countries and because of the lagged effects of
air pollution on health. Nevertheless, air pollution is a major health concern, linked to
respiratory diseases, lung cancer and cardiovascular diseases.
The level of pollution varies greatly across different neighbourhoods, with consequent
effects on health. A review found that poorer and less educated populations often (but not
always) lived in areas with worse air pollution, but also were far more likely to experience
negative health effects from air pollutants (Deguen and Zmirou-Navier, 2010). The authors
posit this reflects a greater susceptibility because of factors such as higher prevalence of
chronic conditions and greater long-term exposure to pollutants. More generally, children
and the elderly are particularly vulnerable to air pollution.
Alongside pollution, other aspects of a person’s living environment also impact upon
their health. Poor housing conditions and certain neighbourhood characteristics such as the
risk of crime have frequently been shown to adversely affect health (Gibson et al., 2011).
Households with low-incomes and many ethnic minorities are more likely to experience
these inadequate living conditions. Policies targeting better housing infrastructure (home
visits, removal of hazards) and rental assistance policies, have had positive health effects
(Bambra et al., 2010).
Conclusion
Empirical results demonstrate that while life expectancy depends on factors both within
and beyond the health system, health spending has been a major driver of life expectancy
gains in recent decades. In particular, a 10% increase in health spending per capita (in real
terms) is associated with a gain of 3.5 months of life expectancy. Given the notable evolution
in health spending in the last 20 years, higher health spending is associated with 42.4 months
of life expectancy gains in this time period.
Education and income have also made significant contributions to life expectancy
gains. A 10% increase in education coverage is associated with a gain of 3.2 months of life
expectancy, and a 10% increase in income per capita with 2.2 months. The same rate of
improvement in healthier lifestyles (10%) is associated with a gain of 2.6 months of life
expectancy (fewer smokers with 1.6 months, decreased alcohol use with 1 month). Other
factors – out-of-pocket spending, healthy diet, unemployment, air pollution – had smaller
effects at the aggregate level. For some of these factors, notably air pollution and healthy
diet, this may reflect long time lags before they affect an individual’s health.
These empirical results provide a useful aggregate picture of the relative importance of
investments within and beyond the health system. Looking forward, future analysis using
such macro-level data could include variables that proxy health policies and institutional
characteristics, and sub-national analysis.
Health at a Glance 2017 © OECD 2017 41
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
It is important, though, to reiterate that observed associations between life expectancy
and explanatory factors at this macro-level does not guarantee causality. Indeed, it is
important to recognise two-way causality, as ill-health worsens productivity, hinders job
prospects, and adversely affects human capital development. For this reason, a review of
more micro-level evidence was also undertaken. Such evidence was generally consistent
with the macro-level analysis, while also providing further precision on the mechanisms
by which different socio-economic factors and a person’s living environment affect health.
For example, the empirical results showed that income has a strong positive association
with life expectancy. Micro-level evidence adds to this by demonstrating that the nature of
income trajectories matter: persistent poverty has particularly adverse health effects, and
falls in income have a larger impact on health than income gains.
Taken together, the main policy implication emerging from this analysis is the significant
opportunities for health improvement from coordinated action across ministries responsible
for education, the environment, income and social protection, alongside health ministries.
This includes inter-sectoral action to address health-related behaviours. In this regard, the
WHO Health in All Policies (HiAP) framework provides countries with an approach that
systematically accounts for the health implications of public policies across sectors (WHO,
2013). Collaboration with the private sector will also be important, especially with employers
in relation to working conditions. Particular attention should be paid to early childhood,
since early life circumstances are crucial to future health and economic prospects, as well
as to shaping health-related behaviours later in life. Such policies can help reduce health
inequalities and achieve better health outcomes for all.
Notes
1. The studies referenced in the text are based on a systematic review of the literature, based on
studies from 1995 or later that included OECD and/or BRIICS countries. Note that such econometric
analyses face some common methodological issues, including two-way causality and delayed effects
of certain factors on health outcomes. James et al. (forthcoming) explores these methodological
issues in more detail.
2. A positive association with life expectancy is consistent with other country-level studies that have
typically shown decreases in mortality (as well as morbidity) during economic downturns, when
unemployment levels are higher (Ruhm, 2012). However, much of the observed correlation between
unemployment and life expectancy in these studies has been explained by fewer traffic accidents
and lower pollution (particularly as decreases in deaths have been concentrated among the elderly),
rather than unemployment per se (Miller et al., 2009; van Gool and Pearson, 2014). Moreover, auxiliary
regressions with interaction terms between unemployment and country dummies showed large
variability in the sign and strength of this coefficient across countries.
References
Bambra, C. et al. (2010), “Tackling the Wider Social Determinants of Health and Health Inequalities:
Evidence from Systematic Reviews”, Journal of Epidemiology and Community Health, Vol. 64, pp. 284-291.
Bassanini, A. and E. Caroli (2014), “Is Work Bad for Health? The Role of Constraint Versus Choice”, IZA
Discussion Paper No. 7891.
Benzeval, M. and K. Judge (2001), “Income and Health: The Time Dimension”, Social Science and Medicine,
Vol. 52, pp. 1371-1390.
Berger, M. and J. Messer (2002), “Public Financing of Health Expenditures, Insurance, and Health
Outcomes”, Applied Economics, Vol. 34, pp. 2105-2113.
Blázquez-Fernández, C., N. González-Prieto and P. Moreno-Mencía (2013), “Pharmaceutical Expenditure
as a Determinant of Health Outcomes in EU Countries”, Estudios de Economía Aplicada, Vol. 31,
pp. 379-396.
42 Health at a Glance 2017 © OECD 2017
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
Caroli, E. and M. Godard (2014), “Does Job Insecurity Deteriorate Health?”, Health Economics, Vol. 27.
Clougherty, J., K. Souza and M. Cullen (2013), “Work and Its Role in Shaping the Social Gradient in Health”,
Annals of New York Academy of Sciences, Vol. 1186, pp. 102-124.
Contoyannis, P., A.M. Jones and N. Rice (2004), “The Dynamics of Health in the British Household Panel
Survey”, Journal of Applied Econometrics, Vol. 19, No. 4, pp. 473-503.
Cutler, D. and A. Lleras-Muney (2010), “Understanding Differences in Health Behaviours by Education”,
Journal of Health Economics, Vol. 29, No. 1, pp. 1-28.
Datta Gupta, N. and N. Kristensen (2008), “Work Environment Satisfaction and Employee Health: Panel
Evidence from Denmark, France and Spain, 1994–2001”, European Journal of Health Economics, Vol. 9,
No. 1, pp. 51-61.
Deaton, A. (2003), “Health, Inequality, and Economic Development”, Journal of Economic Literature, Vol. 41,
No. 1, pp. 113-158.
Deguen, S. and D. Zmirou-Navier (2010). Social inequalities resulting from health risks related to ambient
air quality – a European review. European Journal of Public Health, 20(1): 27–35.
Dobkin, C. and S. Puller (2007), “The Effects of Government Transfers on Monthly Cycles in Drug Abuse,
Hospitalization and Mortality”, Journal of Public Economics, Vol. 91, pp. 2137–2157.
Evans, N. and T. Moore (2011), “The Short-term Mortality Consequences of Income Receipt”, Journal of
Public Economics, Vol. 95, pp. 1410-1424.
Fuchs, V. (2004), “Reflections on the Socio-economic Correlates of Health”, Journal of Health Economics,
Vol. 23, pp. 653-661.
Gibson, M. et al. (2011), “Housing and Health Inequalities: A Synthesis of Systematic Reviews of
Interventions Aimed at Different Pathways Linking Housing and Health”, Health and Place, Vol. 17,
pp. 175-184.
Goldman, D.P. and J.P. Smith (2002), “Can Patient Self-management Help Explain the SES Health
Gradient?”, Proceedings of the National Academy of Science, Vol. 99, No. 16.
Harrington, J.M. (2001), “Health Effects of Shift Work and Extended Hours of Work”, Occupational and
Environmental Medicine, Vol. 58, No. 1, pp. 68-72.
Heijink, R., X. Koolman and G.P. Westert (2013), “Spending More Money, Saving More Lives? The
Relationship Between Avoidable Mortality and Healthcare Spending in 14 Countries”, European
Journal of Health Economics, Vol. 14, pp. 527-538.
James, C., M. Devaux and F. Sassi (forthcoming), “Inclusive growth and health”, OECD Health Division
Working Papers, OECD Publishing, Paris
Kivimaki, M. et al. (2015), “Long Working Hours and Risk of Coronary Heart Disease and Stroke:
A Systematic Review and Meta-analysis of Published and Unpublished Data for 603 838 Individuals”,
The Lancet, Vol. 386, pp. 1739-1746.
Laliotis, I., J.P.A. Ioannidis and C. Stavropoulou (2016), “Total and Cause-specific Mortality Before and
After the Onset of the Greek Economic Crisis: An Interrupted Time-series Analysis”, The Lancet,
Vol. 12, pp. 56-65.
Lin, S.-J. (2009), “Economic Fluctuations and Health Outcome: A Panel Analysis of Asia-Pacific Countries”,
Applied Economics, Vol. 41, pp. 519-530.
Llena-Nozal, A. (2009), “The Effect of Work Status and Working Conditions on Mental Health in Four
OECD Countries”, National Institute Economic Review, Vol. 209, No. 1, pp. 72-87.
Mackenbach, J.P. et al. (2008), “Socioeconomic Inequalities in Health in 22 European Countries”, New
England Journal of Medicine, Vol. 358, pp. 2468–2483.
Marmot, M. and R. Wilkinson (2006), Social Determinants of Health, 2nd edition, Oxford University Press.
McInerney, M., J.M. Mellor and L.H. Nicholas (2013), “Recession Depression: Mental Health Effects of the
2008 Stock Market Crash”, Journal of Health Economics, Vol. 32, No. 6, pp. 1090-1104.
Miething, A. and M. Aberg-Yngwe (2014), “Stability and Variability in Income Position Over Time: Exploring
their Role in Self-rated Health in Swedish Survey Data”, BMC Public Health, Vol. 14:1300.
Miller, D. et al. (2009), “Why Are Recessions Good for Your Health?”, AER Papers and Proceedings, Vol. 99,
No. 2, pp. 122-127.
Health at a Glance 2017 © OECD 2017 43
2. What has driven life expectancy gains in recent decades? A cross-country analysis of OECD member states
Moreno-Serra, R. and P. Smith (2015), “Broader Health Coverage Is Good for the Nation’s Health: Evidence
from Country Level Panel Data”, Journal of the Royal Statistical Society, Vol. 178, pp. 101-124.
Murtin, F. et al. (2017), “Inequalities in longevity by education in OECD countries: Insights from new
OECD estimates”, OECD Statistics Working Papers, No. 2017/02, OECD Publishing, Paris, http://dx.doi.
org/10.1787/6b64d9cf-en.
Nixon, J. and P. Ulmann (2006), “The Relationship Between Health Care Expenditure and Health Outcomes:
Evidence and Caveats for a Causal Link”, European Journal of Health Economics, Vol. 7, pp. 7-18.
O’Donnell, O., E. van Doorslaer and T. van Ourti (2013), “Health and Inequality”, Netspar Discussion Papers
No. 10/2013-060.
OECD (2017), Tackling Wasteful Spending on Health, OECD Publishing, Paris, http://dx.doi.org/10.1787/
9789264266414-en.
OECD (2016), The Economic Consequences of Outdoor Air Pollution, OECD Publishing, Paris, http://dx.doi.
org/10.1787/9789264257474-en.
OECD (2015a), Fiscal Sustainability of Health Systems: Bridging Health and Finance Perspectives, OECD
Publishing, Paris, http://dx.doi.org/10.1787/9789264233386-en.
OECD (2015b), Tackling Harmful Alcohol Use: Economics and Public Health Policy, OECD Publishing, Paris,
http://dx.doi.org/10.1787/9789264181069-en.
OECD (2014), Geographic Variations in Health Care: What Do We Know and What Can Be Done to Improve Health
System Performance?, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264216594-en.
OECD (2010), Health Care Systems: Efficiency and Policy Settings, OECD Publishing, Paris, http://dx.doi.
org/10.1787/9789264094901-en.
OECD (2006), “Measuring the Effects of Education on Health and Civic Engagement: Proceedings
of the Copenhagen Symposium”, OECD, Paris, http://www.oecd.org/edu/innovation-education/
measuringtheeffectsofeducationonhealthandcivicengagement.htm.
Or, Z. (2000), “Determinants of Health Outcomes in Industrialised Countries: A Pooled, Cross-Country,
Time-Series Analysis”, OECD Economic Studies, Vol. 30, pp. 53-77, http://dx.doi.org/10.1787/eco_
studies-v2000-1-en.
Pickett, K.E. and R.G. Wilkinson (2015), “Income Inequality and Health: A Causal Review”, Social Science
and Medicine, Vol. 128, pp. 316-326.
Roelfs, D.J. et al. (2011), “Losing Life and Livelihood: A Systematic Review and Meta-analysis of
Unemployment and All-cause Mortality”, Social Science and Medicine, Vol. 72, No. 6, pp. 840-854.
Ruhm, C. (2012), “Understanding the Relationship Between Macroeconomic Conditions and Health”, in
A. Jones (ed.), The Elgar Companion to Health Economics, pp. 5-14.
Shaw, J. (2005), “The Determinants of Life Expectancy: An Analysis of the OECD Health Data”, Southern
Economic Journal, Vol. 71, pp. 768-783.
van Gool, K. and M. Pearson (2014), “Health, Austerity and Economic Crisis: Assessing the Short-term
Impact in OECD countries”, OECD Health Working Papers, No. 76, OECD Publishing, Paris, http://dx.doi.
org/10.1787/5jxx71lt1zg6-en.
WHO – World Health Organization (2013), Health in All Policies – Seizing Opportunities, Implementing Policies,
edited by K. Leppo, E. Ollila, S. Peña, M. Wismar and S. Cook, WHO, Geneva.
WHO (2008), “Closing the Gap in a Generation”, Commission on social determinants of health.
44 Health at a Glance 2017 © OECD 2017
3. HEALTH STATUS
Life expectancy at birth
Life expectancy by sex and education level
Main causes of mortality
Mortality from circulatory diseases
Mortality from cancer
Infant health
Mental health
Perceived health status
Cancer incidence
Diabetes prevalence
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.
HEALTH AT A GLANCE 2017 © OECD 2017 47
3. HEALTH STATUS
Life expectancy at birth
Life expectancy at birth was on average 80.6 years across primary care, and a large share of the population uninsured;
OECD countries in 2015 (Figure 3.1). There have been 2) health-related behaviours, including greater obesity
substantial gains in life expectancy over time, with life rates, higher consumption of prescription and illegal
expectancy at birth on average ten years higher today than drugs, more deaths from road traffic accidents and higher
it was in 1970. A number of countries reported slight falls in homicide rates; and 3) higher rates of poverty and income
life expectancy between 2014 and 2015, though preliminary inequality than in most other OECD countries (National
data for 2016 suggest these reductions were temporary. Research Council and Institute of Medicine, 2013).
Among OECD countries, Turkey, Korea and Chile have Higher national income (as measured by GDP per capita) is
experienced the largest gains since 1970, with increases generally associated with higher life expectancy at birth,
of 24, 20 and 17 years respectively. Gains in longevity over although the relationship is less pronounced at the highest
time can be attributed to a number of factors within and levels of national income (Figure 3.2). There are also notable
beyond the health system. These include rising incomes, differences in life expectancy between countries with
better education, healthier lifestyles and progress in similar income per capita. For example, Japan and Spain
health care (see Chapter 2 for further analysis). Indeed, have higher, and Luxembourg, the United States and the
each of these countries has experienced rapid economic Russian Federation lower, life expectancies than would be
growth alongside expanded health care coverage in recent predicted by their GDP per capita alone.
decades. Figure 3.3 shows the relationship between life expectancy at
Although the life expectancy in partner countries such birth and health spending per capita across OECD, candidate
as India, Indonesia, Brazil and China remains well below and partner countries. Higher health spending per capita is
the OECD average, these countries have also achieved generally associated with higher life expectancy at birth,
considerable gains in longevity over the past decades, with although this relationship tends to be less pronounced
the level converging rapidly towards the OECD average. in countries with the highest health spending per capita.
There has been less progress in South Africa (due mainly Japan, Spain and Korea stand out as having relatively high
to the epidemic of HIV/AIDS), Lithuania and the Russian life expectancies, and the United States and the Russian
Federation (due mainly to the impact of the economic Federation relatively low life expectancies, given their
transition in the 1990s and a rise in risky health behaviours levels of health spending.
among men).
Japan, Spain and Switzerland lead a large group of 25 OECD
countries in which life expectancy at birth now exceeds Definition and comparability
80 years. A second group, including the United States, Chile
and a number of central and eastern European countries, Life expectancy at birth measures how long, on
has a life expectancy between 75 and 80 years. average, people would live based on a given set of age-
specific death rates. However, the actual age-specific
Among OECD countries, Latvia and Mexico had the lowest
death rates of any particular birth cohort cannot be
life expectancy in 2015, at around 75 years. Since 2000,
known in advance. If age-specific death rates are
life expectancy in Mexico has increased more slowly than
falling (as has been the case over the past decades),
in other OECD countries, with a gain of just over a year
actual life spans will be higher than life expectancy
compared with an average gain of more than three years
calculated with current death rates.
across OECD countries. Slow progress in life expectancy in
Mexico is due to a number of factors, including harmful The methodology used to calculate life expectancy
health-related behaviours such as poor nutrition and high can vary slightly between countries. This can change
obesity rates, increasing mortality rates from diabetes and a country’s estimates by a fraction of a year.
a lack of progress in reducing mortality from circulatory Life expectancy at birth for the total population
diseases, high death rates from road traffic accidents and is calculated by the OECD Secretariat for all OECD
homicides, as well as persistent barriers of access to quality countries, using the unweighted average of life
care. expectancy of men and women.
In the United States, gains in life expectancy over the past
few decades have also been more modest than in most
other OECD countries. While life expectancy in the United
References
States used to be one year above the OECD average in 1970,
it is now almost two years below the average. Many factors National Research Council and Institute of Medicine,
can explain these lower gains in life expectancy, including: S. Woolf and L. Aron (eds) (2013), U.S. Health in International
1) the highly fragmented nature of the US health system, Perspective: Shorter Lives, Poorer Health, National Academies
with relatively few resources devoted to public health and Press, Washington, DC.
48 HEALTH AT A GLANCE 2017 © OECD 2017
3. HEALTH STATUS
Life expectancy at birth
3.1. Life expectancy at birth, 1970 and 2015 (or nearest year)
1970 2015
Years
90
83.9
83.0
83.0
82.6
82.5
82.5
82.4
82.4
82.4
82.3
82.1
82.1
81.7
81.7
81.6
81.6
81.5
81.3
81.2
80.9
81.0
81.1
81.1
80.8
80.7
80.6
79.6
79.1
78.8
78.7
78.0
77.7
77.6
80
76.7
76.0
75.7
75.0
74.7
74.6
74.5
74.2
71.3
69.1
68.3
70
57.4
60
50
40
a
n ia
Br o
m ce
S w S an
do on
er n
nd
st ly
a
h ia
Ic a li a
Lu F an d
No ur g
S w way
Is n
Ne K r ael
Ze ea
Ne F in n d
er d
Ca nd s
Ir e d a
Au nd
rt a
Tu lic
lg l
i t e G ium
ng e
C ic
rm r k
Co ECD y
st 3 5
C z i t e Ch a
e c d S il e
pu s
E s ke y
ak Po ia
pu d
Sl dom
ng a
M ar y
L il
s s C thu ia
n lo ia
a
Be uga
Po s t r i
ric
O an
de bi
si
Re t e
Hu hin
K i ec
ic
az
i t z pai
ic
th lan
Re lan
De ven
bl
u t In d
n
Au It a
ia o an
Li at v
Ge ma
xe r a n
w or
b
ed
na
p
la
In r a t i
la
a
ne
aR
Fe m
to
ex
h ta
bo
r
d re
r
Ja
el
al
la
Af
r
So
Un
ov
Un
Sl
Ru
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602234
3.2. Life expectancy at birth and GDP per capita, 2015 3.3. Life expectancy at birth and health spending
(or nearest year) per capita, 2015 (or nearest year)
Life expectancy in years Life expectancy in years
85 85
ITA FRA JPN FINCAN ITA JPN
FRA SWE
ISL AUS ISR ESP ISL
ISR ESP CHE LUX AUS CHE
KOR NOR KOR
SWE NLD LUX
IRL NZL
GRC NZL GRC IRL
AUT NLD PRT FIN
GBR AUT
NOR
80 PRT SVN 80
CRI DNK CRI CHL SVN DEU
GBR
CHL USA
CZE USA
EST CZE DNK
TUR POL BEL TUR CAN BEL
EST DEU
POL
CHN SVK CHN SVK
MEX HUN MEX HUN
75 BRA 75 BRA
LTU
COL LVA LTU LVA
COL
RUS RUS
70 70
IDN IDN
IND IND
R 2 = 0.57 R 2 = 0.54
65 65
0 20 000 40 000 60 000 80 000 100 000 0 2 000 4 000 6 000 8 000 10 000
GDP per capita (USD PPP) Health spending per capita (USD PPP)
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602253 12 http://dx.doi.org/10.1787/888933602272
HEALTH AT A GLANCE 2017 © OECD 2017 49
3. HEALTH STATUS
Life expectancy by sex and education level
There remain large gaps in life expectancy between inequalities are less pronounced. Differences in lifespan
women and men in all OECD countries. On average across between people with low and high education have been
OECD countries, life expectancy at birth for women was estimated to account for about 10% of overall inequalities
83.1 years in 2015, compared with 77.9 years for men, a gap in ages at death (Murtin et al., 2017).
of 5.2 years (Figure 3.4). The gender gap in life expectancy
increased substantially in many OECD countries during
the 1970s and early 1980s to reach a peak of almost seven Definition and comparability
years in the mid-1980s, but it has narrowed since, reflecting
higher gains in life expectancy among men than women. Life expectancy at birth measures how long, on average,
This can be attributed at least partly to narrowing of people would live based on a given set of age-specific
differences in risk-increasing behaviours such as smoking, death rates. However, the actual age-specific death
accompanied by sharp reductions in mortality rates from rates of any particular birth cohort cannot be known
circulatory diseases among men. in advance. If age-specific death rates are falling (as
In 2015, life expectancy for women in OECD countries has been the case over the past decades), actual life
ranged from less than 80 years in Hungary, Latvia and spans will be higher than life expectancy calculated
Mexico to more than 85 years in Japan, Spain, France, Korea with current death rates. Data for life expectancy at
and Switzerland. Life expectancy for men ranged from birth comes from Eurostat for EU countries, and from
less than 75 years in Latvia, Mexico, Hungary, the Slovak national sources elsewhere. The methodology used to
Republic, Estonia and Poland to over 80 years in Iceland, calculate life expectancy can vary slightly between
Japan, Switzerland, Norway, Australia, Sweden, Italy, Israel countries. This can change a country’s estimates by
and Spain. a fraction of a year.
Among OECD countries, the gender gap in life expectancy Data for life expectancy by education level come from
is relatively narrow in Iceland, the Netherlands, New national suveys provided for the OECD Health Data
Zealand, the United Kingdom, Norway, Sweden, Ireland questionnaire for Israel, Mexico and the Netherlands;
and Denmark (a gap of less than four years), but much from the OECD Statistics Directorate project (see
larger in Latvia (around ten years) Estonia (around Murtin et al. below) for Australia, Austria, Belgium,
nine years), Poland (around eight years), the Slovak Canada, France, Latvia, the United Kingdom and the
Republic and Hungary (around seven years). In this latter United States; and from Eurostat for the remaining
group of countries, gains in life expectancy of men over 14 European countries shown in Chart 3.5.
the past few decades have been much more modest than To calculate life expectancies by education level,
in other countries. For partner countries, the gender detailed data on deaths by sex, age and education
gap is also large in the Russian Federation, Lithuania level are needed. However, not all countries have
and Colombia (seven years or more), and small in China information on education as part of their deaths data.
(around three years). In such cases, data linkage to another source (e.g. a
Life expectancy in OECD countries varies by socio-economic census) which does have information on education
status as measured, for instance, by education level may be required (Corsini, 2010). Note further that
(Figure 3.5). A higher education level not only provides the data disaggregated by education are only available
means to improve the socio-economic conditions in which for a subset of the population for Belgium, the
people live and work, but may also promote the adoption Czech Republic and Norway, and that there are more
of healthier lifestyles and facilitate access to appropriate missing data on education among the deceased than
health care. the population at large. In these three countries, the
large share of the deceased population with missing
On average among 25 OECD countries for which recent data
education (above 40%) could affect the accuracy of
are available, people with the highest level of education can
results.
expect to live around six years longer than people with the
lowest level of education at age 30 (53.4 versus 47.8 years).
These differences in life expectancy by education level
are particularly pronounced for men, with an average gap References
of seven years. Differences are especially pronounced in
central and eastern European countries (Slovak Republic, Corsini, V. (2010), “Highly Educated Men and Women
Estonia, Poland, Hungary, Latvia and the Czech Republic), Likely to Live Longer: Life Expectancy by Educational
where the life expectancy gap between higher and lower Attainment”, Eurostat Statistics in Focus 24/2010, European
educated men is more than ten years. This is largely Commission, Luxembourg.
explained by older people in these countries having Murtin, F. et al. (2017), “Inequalities in Longevity by Education
lower levels of education, and the greater prevalence of in OECD Countries: Insights from New OECD Estimates”,
risk factors among men, such as tobacco and alcohol use. OECD Statistics Working Papers, No. 2017/02, OECD
In other countries such as Turkey, Sweden and Canada, Publishing, Paris, http://dx.doi.org/10.1787/6b64d9cf-en.
50 HEALTH AT A GLANCE 2017 © OECD 2017
3. HEALTH STATUS
Life expectancy by sex and education level
3.4. Life expectancy at birth by sex, 2015 (or nearest year)
Total Men Women
Years
90
85
80
75
70
83.9
83.0
83.0
82.6
82.5
82.5
82.4
82.4
82.4
82.3
82.1
82.1
81.7
81.7
81.6
81.6
81.5
81.3
81.2
80.9
81.0
80.8
81.1
81.1
80.6
80.7
79.6
79.1
65
78.8
78.7
78.0
77.7
77.6
76.7
76.0
75.7
75.0
74.7
74.6
74.5
74.2
71.3
69.1
60
68.3
55
57.4
50
Br c o
a
b e
Ge ma a
Tu blic
ic
ne n
S w S pan
er in
Au It a d
st ly
Ic a li a
xe r a d
g
ed y
Is e n
w Ko e l
a a
Ca land
th nla a
er nd
Ir e n d s
r ia
lg al
d Gr e um
ng c e
o m
r rk
C o E C ny
aR 5
Cz ite C ic a
h St le
pu es
E s r ke y
ak Po ni a
pu d
ng a
M ar y
A u land
L i a t il
ia Co ua ia
Fe m a
In r a t i a
u t In i a
Af a
n i
ric
m nc
Swr wa
N e F i ad
Hu Chin
h di
n lo ni
Ze re
az
n
do io
Re lan
Lu F lan
No ou r
st D3
Po s t r
De ven
bl
e c d hi
de b
ra
Be tug
s
th v
i t z pa
Sl do
Ki e
Re a t
O ma
i
la
to
ex
i
r
n
Ja
la
e
So
Ne
Un
ite
ov
Un
Sl
ss
Ru
Note: Countries are ranked in descending order of life expectancy for the whole population.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602291
3.5. Gap in life expectancy at age 30 between highest and lowest education level, by sex, 2015
(or nearest year)
Male Female
Gap in years
16
14
12
10
0
o
ic
ic
ia
a
nd
ia
el
es
ce
li a
r ia
ce
l
ay
ly
da
en
ey
25
ga
ni
ar
nd
ar
ic
an
bl
bl
en
ra
It a
tv
iu
do
rk
rw
an
at
ee
ra
ed
na
la
st
CD
to
ex
r tu
nm
ng
pu
pu
nl
La
la
lg
Is
ov
St
Tu
Po
st
ng
Au
Gr
Fr
Sw
Ca
No
Es
er
Fi
OE
Be
Hu
Po
Re
Re
Au
De
Sl
d
Ki
th
ite
h
ak
Ne
d
ec
Un
ite
ov
Cz
Un
Sl
Note: The figures show the gap in the expected years of life remaining at age 30 between adults with the highest level (“tertiary education”) and the
lowest level (“below upper secondary education”) of education.
Source: Eurostat database complemented with OECD Statistics Directorate data and national data for Israel, Mexico and the Netherlands.
12 http://dx.doi.org/10.1787/888933602310
HEALTH AT A GLANCE 2017 © OECD 2017 51
3. HEALTH STATUS
Main causes of mortality
Over 10 million people died in 2015 across OECD lung cancer and accident-related deaths were higher for
countries, which equates to an average of 793 deaths per men than for women. A body of evidence suggests that
100 000 population. Diseases of the circulatory system alongside intrinsic gender differences, women are more
and cancer are the two leading causes of death in most likely to choose healthy behaviours (Gore et al., 2011).
countries. Across the OECD, more than one in three deaths It is also worth noting that the main causes of death
were caused by ischaemic heart diseases, stroke or other diverge between socio-economic groups. Social disparities
circulatory diseases; and one in four deaths were related are generally larger for the most preventable diseases, as
to cancer. deaths are amenable to medical intervention, behaviour
Two factors can explain certain commonalities in causes of change and injury prevention (Mackenbach et al., 2015).
death across OECD and partner countries. First, population
ageing is important since the main causes of death change
with age. Among younger adults, cancer-related deaths
occur more frequently than many other causes. After Definition and comparability
age 50, deaths due to diseases of the circulatory system rise
Mortality rates are based on numbers of deaths
steadily, and become one of the major causes of death after
registered in a country in a year divided by the size
age 80, along with dementia. Second is the epidemiological
of the corresponding population. The rates have been
transition from communicable to non-communicable
directly age-standardised to the 2010 OECD population
diseases, which has already taken place in high-income
(available at http://oe.cd/mortality) to remove variations
countries and is rapidly occurring in many middle-income
arising from differences in age structures across
countries (GBD, 2013).
countries and over time. The source is the WHO
Variation across OECD and partner countries is substantial. Mortality Database.
All-cause mortality rates (age-standardised) ranged
Deaths from all causes are classified to ICD-10, Codes
from 583 deaths per 100 000 population in Japan to over
A00-Y89, excluding S00-T98. The classification of
1 000 deaths per 100 000 in Hungary, Latvia, Lithuania,
causes of death defines groups and subgroups. Groups
the Russian Federation and the Slovak Republic in 2015
are umbrella terms covering diseases that are related
(Figure 3.6). Looking at specific causes, diseases of the
to each other; subgroups refer to specific diseases.
circulatory system were the main cause of mortality in
For example, the group diseases of the respiratory
most OECD countries. They caused over 600 deaths per
system comprises 4 subgroups: influenza, pneumonia,
100 000 population in Latvia and Lithuania, and 869 deaths
chronic obstructive pulmonary diseases and asthma.
per 100 000 in the Russian Federation. Japan and France had
the lowest rates, at 152 and 164 deaths per 100 000 population
respectively. Diet, smoking and alcohol consumption
play important roles in these diseases, as does access to
References
treatment.
Variations in cancer-related deaths was less substantial but GBD 2013 Mortality and Causes of Death Collaborators
still significant, ranging from 123 to 286 deaths per 100 000 (2015), “Global, Regional, and National Age-sex Specific
in 2015. Other causes of death were particularly important in All-cause and Cause-specific Mortality for 240 Causes of
specific countries. For example, respiratory system diseases Death, 1990–2013: A Systematic Analysis for the Global
(predominantly chronic obstructive pulmonary diseases) Burden of Disease Study 2013”, The Lancet, Vol. 385,
caused over 100 deaths per 100 000 in Ireland, the United pp. 117–171.
Kingdom, Brazil and Colombia. External causes (predominantly Gore, F. et al. (2011), “Global Burden of Disease in Young
assault, accidents and intentional self-harm) accounted for People Aged 10–24 Years: A Systematic Analysis”, The
over 80 deaths per 100 000 in Brazil, Latvia, Lithuania, South Lancet, Vol. 377, pp. 2093–2102.
Africa and the Russian Federation. HIV-AIDS caused more Mackenbach, J. et al. (2015), “Variations in the Relation
than 50 deaths per 100 000 population in South Africa. Between Education and Cause-specific Mortality in
The main causes of death also differ by gender (Figure 3.7). 19 European Populations: A Test of the ‘Fundamental
For example, dementia is a more important cause of Causes’ Theory of Social Inequalities in Health”, Social
death for women than for men. In contrast, the rates of Science & Medicine, Vol. 127, pp. 51-62.
52 HEALTH AT A GLANCE 2017 © OECD 2017
3. HEALTH STATUS
Main causes of mortality
3.6. Main causes of mortality per country, 2015 (or nearest year)
Circulatory system Cancer Respiratory system
Dementia External causes Other
Age-standardised rates per 100 000 population
1 600
1 400
1 200
1 000
211
800
237
245
286
242
260
600
225
246
245
869
159
169
203
400
207
163
202
209
235
177
215
215
193
193
123
197
201
231
202
654
643
177
232
207
184
210
238
203
212
192
566
192
187
209
183
511
494
182
453
438
200
349
336
317
314
302
294
289
282
280
277
268
255
251
251
249
239
235
234
230
222
218
212
206
204
203
191
189
185
184
179
166
164
152
0
Is e
Fi ico
Fr an
n
Ko e l
Ca rea
S a
nm n
Ne st k
er ia
No d s
L e ay
i t e mb m
lo ia
S w ingd g
pu c
nd
s t il e
Po Ric a
Ne Sw nd
Un Z e en
i t z om
Ic g a l
d nd
es
Ir e l y
M nd
O E lan d
Gr 3 5
rm e
A u ny
B a
o l
Tu ia
Es lic
e P ey
ov e nd
Hu ni a
th y
n L ia
de via
Sl r a zi
c
ri
d
Re bli
A u ar
Ge eec
tio
De pai
L i gar
r
ra
th r al
Co ven
b
n
It a
U n u x e l gi u
d ou
an
rw
rk
C o Ch
at
b
na
p
w ed
a
n
la
ite ala
la
S l ch R o l a
st
CD
Fe at
m
r tu
ex
to
ua
ak pu
Ja
el
la
ra
St
er
n
a
B
ia
Cz
ss
Ru
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602329
3.7. Main causes of mortality by gender, 2015 (or nearest year)
Women Men
0.7% Parkinson’s disease 0.9% Parkinson’s disease
0.8% Intential self-harm 1.3% Alzheimer’s disease
Ischaemic
heart 10.6% Ischaemic
2.5% Colorectal cancer heart 12.8% 2.1% Dementia
diseases
diseases
3.0% Alzheimer’s disease 2.2% Intentional self-harm
2.5% Prostate cancer
3.1% Accidents
Stroke 8.2% 2.9% Colorectal cancer
3.2% Breast cancer
Lung cancer 7.0%
3.1% Diabetes
3.4% Diabetes
Dementia 4.5% 4.1% COPD
3.4% COPD Stroke 6.0%
3.9% 4.8%
Lung cancer Accidents
Note: Shares of the sum of all deaths across OECD countries, by gender.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602348
HEALTH AT A GLANCE 2017 © OECD 2017 53
3. HEALTH STATUS
Mortality from circulatory diseases
Despite substantial declines in recent decades, circulatory burden from stroke and other cerebrovascular diseases is
diseases remain the main cause of mortality in most OECD also substantial (Feigi et al., 2016).
countries, accounting for more than one-third (36%) of all There are large variations in cerebrovascular disease
deaths in 2015. Prospects for further reductions may be mortality rates across countries (Figure 3.9). Among OECD
hampered by a rise in certain risk factors such as obesity countries, Latvia, Hungary and the Slovak Republic report
and diabetes (OECD, 2015). Circulatory diseases cover a cerebrovascular mortality that is more than three times
a range of illnesses related to the circulatory system, higher than that of Switzerland, Canada and France, and have
particularly ischaemic heart disease (including heart the highest mortality rates for both IHD and cerebrovascular
attack) and cerebrovascular diseases such as stroke. disease. Rates are also high in the partner countries of the
Ischaemic heart disease (IHD) is caused by the accumulation Russian Federation and South Africa. The high prevalence
of fatty deposits lining the inner wall of a coronary artery, of risk factors common to both diseases (e.g. smoking and
restricting blood flow to the heart. IHD alone was responsible high blood pressure) may explain this link.
for nearly 12% of all deaths in OECD countries in 2015. Since 1990, cerebrovascular disease mortality has
However, mortality from IHD varies considerably across decreased in all OECD countries, although to a lesser
countries (Figure 3.8). Among OECD countries, Central and extent in Poland and the Slovak Republic. On average, the
Eastern European countries report the highest IHD mortality mortality burden from cerebrovascular disease has halved
rates. Rates are also high in the Russian Federation. across OECD countries. In Estonia, Luxembourg, Portugal,
Japan, Korea and France report the lowest rates. Across the Czech Republic and Austria, the rates have been cut
OECD countries, IHD mortality rates in 2015 were around by over 70%. As with IHD, the reduction in mortality from
82% higher for men than women. cerebrovascular disease can be attributed at least partly
IHD mortality rates have declined in nearly all OECD to a reduction in risk factors as well as improvements in
countries, with an average reduction of more than 50% medical treatments (OECD, 2015; see indicator “Mortality
since 1990, contributing greatly to gains in life expectancy, following ischaemic stroke” in Chapter 6) but rising obesity
particularly among men. The decline has been most and diabetes threatens progress in tackling cerebrovascular
remarkable in Denmark, the Netherlands, Norway and disease (OECD, 2015).
Israel, where rates fell by over 70%. Declining tobacco
consumption contributed significantly to reducing the
incidence of IHD (see indicator on “Smoking among adults”
in Chapter 4), and consequently to reducing mortality Definition and comparability
rates. Improvements in medical care have also contributed
Mortality rates are based on numbers of deaths
to reduced mortality rates (see indicators on “Mortality
registered in a country in a year divided by the size
following acute myocardial infarction” in Chapter 6 and
of the corresponding population. The rates have been
“Hospital discharges” in Chapter 9).
directly age-standardised to the 2010 OECD population
In Korea, IHD mortality rates have increased substantially (available at http://oe.cd/mortality) to remove variations
since 1990, although they remain low compared with arising from differences in age structures across
nearly all other OECD countries and have started to fall countries and over time. The source is the WHO
after peaking in 2006. The initial rise in IHD mortality rates Mortality Database.
in Korea has been attributed to changes in lifestyle and
Deaths from ischaemic heart disease are classified to
dietary patterns as well as environmental factors at the
ICD-10 codes I20-I25, and cerebrovascular disease to
time of birth, with people born between 1940 and 1950
I60-I69.
facing higher relative risks. In 2006, Korea introduced a
Comprehensive Plan to tackle circulatory diseases that
encompassed prevention and primary care as well as better
acute care, contributing to the reduction in mortality in References
recent years (OECD, 2012).
Feigi, V. et al. (2016), “Global Burden of Stroke and Risk
Cerebrovascular disease was the underlying cause
Factors in 188 Countries, During 1990–2013: A Systematic
for about 7% of all deaths in OECD countries in 2015.
Analysis for the Global Burden of Disease Study 2013”,
Cerebrovascular disease refers to a group of diseases that
The Lancet Neurology, Vol. 15, pp. 913–924.
relate to problems with the blood vessels that supply
the brain. Common manifestations of cerebrovascular OECD (2015), Cardiovascular Disease and Diabetes: Policies for
disease include ischaemic stroke, which develops when Better Health and Quality of Care, OECD Publishing, Paris,
the brain’s blood supply is blocked or interrupted, and http://dx.doi.org/10.1787/9789264233010-en.
haemorrhagic stroke which occurs when blood leaks from OECD (2012), OECD Reviews of Health Care Quality: Korea –
blood vessels into the surface of the brain. In addition Raising Standards, OECD Publishing, Paris, http://dx.doi.
to being an important cause of mortality, the disability org/10.1787/9789264173446-en.
54 HEALTH AT A GLANCE 2017 © OECD 2017
3. HEALTH STATUS
Mortality from circulatory diseases
3.8. Ischaemic heart disease mortality, 2015 and change 1990-2015 (or nearest year)
2015 Change 1990-2015
34 Japan -40
38 Korea 43
39 France -56
46 Netherlands -75
53 Spain -49
54 Belgium -55
55 Portugal -53
59 Luxembourg -61
60 Denmark -81
64 Israel -71
64 Chile -61
66 South Africa -40
72 Norway -73
78 Switzerland -52
82 Greece -37
82 Slovenia -54
84 Italy -38
85 Australia -68
93 Brazil -37
93 Canada -60
95 Sweden -65
98 Poland -33
98 United Kingdom -67
100 Iceland -60
106 Germany -52
112 OECD35 -52
113 United States -55
120 Costa Rica -47
127 Ireland -62
129 New Zealand -56
131 Austria -42
144 Mexico 1
146 Turkey na
147 Finland -57
175 Colombia -4
211 Estonia -67
237 Czech Republic -46
288 Hungary -13
291 Slovak Republic -21
328 Latvia -42
424 Lithuania -34
458 Russian Federation -2
500 400 300 200 100 0 -100 -50 0 50
Age-standardised rates per 100 000 population Change in %
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602367
3.9. Cerebrovascular disease mortality, 2015 and change 1990-2015 (or nearest year)
2015 Change 1990-2015
36 France -59
36 Canada -54
37 Switzerland -60
39 Israel -62
41 Luxembourg -80
42 United States -45
43 Spain -70
45 Austria -71
46 Australia -60
47 Belgium -58
47 Norway -65
48 Germany -66
48 Sweden -56
50 Iceland -52
50 Netherlands -53
51 Japan -63
51 Denmark -54
53 United Kingdom -61
55 Ireland -59
57 Costa Rica -44
58 Mexico -40
58 Estonia -82
63 Finland -58
64 New Zealand -47
65 OECD35 -59
67 Italy -54
71 Chile -48
72 Korea -59
74 Colombia -32
78 Slovenia -59
80 Poland -21
85 Portugal -74
86 Czech Republic -72
89 Greece -59
91 Brazil -48
101 Turkey na
111 Slovak Republic -21
115 Hungary -55
142 South Africa -4
154 Lithuania -14
197 Latvia -44
275 Russian Federation -25
300 250 200 150 100 50 0 -100 -80 -60 -40 -20 0
Age-standardised rates per 100 000 population Change in %
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602386
HEALTH AT A GLANCE 2017 © OECD 2017 55
3. HEALTH STATUS
Mortality from cancer
Cancer is the second leading cause of mortality in OECD in several countries such as the Netherlands, France and
countries after circulatory diseases, accounting for 25% of Spain where it has more than doubled since 1990. These
all deaths in 2015, up from 15% in 1960. In a number of conflicting trends are, to a large degree, explained by the
countries such as Denmark, France, Japan, the Netherlands, high number of females who started smoking several
Canada, the United Kingdom, Spain, Belgium and Australia, decades later than males.
the mortality rate for cancer is higher than for circulatory Breast cancer is the second most common cause of cancer
diseases. The rising share of deaths due to cancer reflects mortality in women in many OECD countries. While there
the fact that mortality rates from other causes, particularly has been an increase in the incidence of breast cancer over
circulatory diseases, has been declining more rapidly than the past decade, mortality has declined in most countries
for cancer. due to earlier diagnosis and better treatment. Mortality
There are more than 100 different types of cancers. For a from breast cancer increased in Korea and Japan, though
large number of cancer types, the risk of developing the the rates there remained the lowest in 2015. Mortality rates
disease rises with age. While genetics is a risk factor, only from breast cancer in 2015 were highest in Ireland, Iceland,
about 5% to 10% of all cancers are inherited. Modifiable Hungary, Denmark and the Netherlands.
risk factors such as smoking, obesity, lack of exercise and Colorectal cancer is a major cause of cancer mortality
excess sun exposure, as well as environmental exposures, among both men and women (second-highest cause of
explain up to 90-95% of all cancer cases (Anand et al., cancer mortality in men and third in women). In Japan,
2008). Prevention, early detection and treatment remain it is the leading cause of cancer mortality in women. In
at the forefront in the battle to reduce the burden of cancer 2015, colorectal cancer mortality was lowest in Mexico and
(OECD, 2013). Turkey, and highest in Hungary and the Slovak Republic.
In 2015, the average rate of mortality attributable to Prostate cancer has become the most common cancer
cancer across OECD countries was just over 200 per among men in many OECD countries, particularly among
100 000 population (Figure 3.10). Mortality due to cancer was men aged 65 years and over.
lowest in Mexico, Turkey, Finland, Switzerland, Japan, Israel
and Korea, with rates less than 180 per 100 000 population.
Among partner countries, rates were also less than 180 per Definition and comparability
100 000 in Colombia, Brazil, Costa Rica and South Africa.
Hungary, the Slovak Republic, Slovenia and Latvia bear the Mortality rates are based on numbers of deaths
highest cancer mortality burden, with rates in excess of 240 registered in a country in a year divided by the size
per 100 000 population. of the corresponding population. The rates have been
In most OECD countries, cancer-related mortality rates have directly age-standardised to the 2010 OECD population
fallen since 1990, with the largest reductions in the Czech (available at http://oe.cd/mortality) to remove variations
Republic and Switzerland. On average, rates fell by 18% arising from differences in age structures across
between 1990 and 2015. Substantial declines in mortality countries and over time. The source is the WHO
from stomach cancer, colorectal cancer, lung cancer for Mortality Database.
men, breast, cervical and ovarian cancer for women, as well Deaths from all cancers are classified to ICD-10
as prostate cancer for men contributed to this reduction. codes C00-C97. The international comparability of
However, these gains were partially offset by increases in cancer mortality data can be affected by differences
the number of deaths due to cancer of the liver, skin and in medical training and practices as well as in death
pancreas for both sexes, as well as lung cancer for women. certification across countries.
Mortality due to cancer is consistently higher for men than
for women in all countries (Figure 3.11). The gender gap was
particularly wide in Korea, Turkey, Latvia, Estonia, Spain References
and Portugal, with rates among men more than twice those
for women. This gender gap can be explained partly by Anand, P. et al. (2008), “Cancer Is a Preventable Disease
the greater prevalence of risk factors among men, notably that Requires Major Lifestyle Changes”, Pharmaceutical
smoking. Research, Vol. 25, No. 9, pp. 2097-2116.
Among men, lung cancer imposes the highest mortality OECD (2013), Cancer Care: Assuring Quality to Improve
burden, accounting for 22% of all cancer-related deaths Survival, OECD Publishing, Paris, http://dx.doi.
(Figure 3.12). For women, lung cancer accounted for 16% org/10.1787/9789264181052-en.
of all cancer-related deaths. In many countries, lung cancer Slawomirski, L., A. Auraaen and N. Klazinga (2017), “The
mortality rates for men have decreased over the last Economics of Patient Safety: Strengthening a Value-based
25 years, in particular in Mexico, the Netherlands, Czech Approach to Reducing Patient Harm at National Level”,
Republic, Finland and the United Kingdom where they fell OECD Health Working Papers, No. 96, OECD Publishing,
by about 50%. But lung cancer mortality has risen for women Paris, http://dx.doi.org/10.1787/5a9858cd-en.
56 HEALTH AT A GLANCE 2017 © OECD 2017
100
150
250
300
50
100
150
200
250
300
350
400
450
200
350
M M
Co e x i c Co e x i c
lo o lo o 114.6
m m
bi b
C o Br a Br i a 158.3
st a zi az
aR l Co Turk il 160.3
Lung
ic st ey
Breast
Is a
r
aR 166.1
ic
Colorectal
Ic a el
el
a
S w F in a 167.2
Fi nd i la
nl So t zer nd
ut lan 172.3
17.5%
h
14.4%
Un S w an d
11.3%
ite ed Af d 175.9
d e ric
Sw St n Ja a 176.2
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
pa
7.5%
it z ate
HEALTH AT A GLANCE 2017 © OECD 2017
er s Is n 176.6
Pancreas
Au lan ra
3.9%
3.3%
178.8
Women
st d Ko e l
1.9%
1.7%
1.1%
ra
Age-standardised rates per 100 000 population
Age-standardised rates per 100 000 population
li a Sw rea
Ch ed 178.9
No il e e
rw Sp n 184.7
A
Liver
So Tu ay Un u s ain
185.8
Bladder
ut rk i t e tr a
4.8% Ovary
Leukemia
h ey d li a
187.3
Cervix uteri
Af St
r at
es
Ca ica
4.6% Stomach
187.8
Men
na Ch
1990
Melanoma of skin
Ne J da Ic il e
w ap el
a 193.9
Ze an Fr n d
al
a 194.4
Au nd Po a n c
r tu e 196.3
Ge stri
rm a Au gal
s 196.9
B e a ny N o t r ia
lg
iu rw 197.2
m Gr a y
Un Sp ee 198.2
O
i t e E ain Lu B e l c e
d CD x e gi u 199.0
Ki 3 m m
ng 5 b
Note: Shares of the sum of cancer-related deaths across OECD countries, by gender.
do 199.4
G e ou r
Fr m rm g 199.9
an
L u Ir e c e O E a ny
CD 200.9
Lung
xe la
m nd Ru 35
bo ss 203.7
u ia C It al
Gr r g n y 205.4
Colorectal
ee Fe ana
ce N e d er d a
Ne
th It a C z w tio a 207.3
Prostate
er l y e Z n 208.9
Women
2015
la
25.5%
Un ch e a l a
10.7%
Po nd i t e Rep nd
r tu s d ub 209.9
K
9.2%
De ga
nm l N e ing lic 220.5
3.10. Cancer mortality, 1990 and 2015 (or nearest year)
th do
3.11. Cancer mortality by gender, 2015 (or nearest year)
Ru C z e ar er m
c la 221.9
6.1%
s s h Ko k n
Stomach
i a Re r e
n p a Ir e d s 223.6
5.9%
Men
F e ub
3.6%
3.5%
Li lan
1.3%
de lic th d
ra
ti u 227.0
3.12. Main causes of cancer mortality by gender, 2015 (or nearest year)
De ani
Po on nm a 232.6
Sl land a
o Po r k 233.4
Liver
Sl L ve la
Bladder
6.1% Pancreas
ov i t h nia Es nd
ak u 234.0
Leukemia
to
Re ani n
pu a L a ia 237.0
b Sl
ov Sl t v i a
Es lic
Melanoma of skin
to ak ov 240.7
n Re eni
L a ia pu a 243.3
Hu t v i Hu bli
ng a ng c 258.6
ar ar
y y 281.6
57
12 http://dx.doi.org/10.1787/888933602443
12 http://dx.doi.org/10.1787/888933602424
12 http://dx.doi.org/10.1787/888933602405
Mortality from cancer
3. HEALTH STATUS
3. HEALTH STATUS
Infant health
Infant mortality, the rate at which babies and children of to be taken into account when interpreting differences
less than one year of age die, is the most fundamental (Euro-Peristat, 2013).
measure of infant health. In OECD countries, around two- Comparisons of different population groups within
thirds of the deaths that occur during the first year of life countries indicate that both infant mortality and the
are neonatal deaths (i.e. during the first four weeks). Birth proportion of low birth weight infants may be influenced
defects, prematurity and other conditions arising during by differences in education level, income and associated
pregnancy are the main factors contributing to neonatal living conditions. For example, in the United States, black
mortality in developed countries. For deaths beyond a women are more likely to give birth to low birth weight
month (post-neonatal mortality), there tends to be a infants, with an infant mortality more than double that for
greater range of causes – the most common being SIDS white women (NCHS, 2015). Similar differences have also
(sudden infant death syndrome), birth defects, infections been observed among the indigenous and non-indigenous
and accidents. populations in Australia, Mexico and New Zealand,
In most OECD countries infant mortality is low and there reflecting the disadvantaged living conditions of many of
is little difference in rates (Figure 3.13). In 2015, the average these mothers.
in OECD countries was less than four deaths per 1 000 live
births. Turkey and Mexico still have comparatively high
infant mortality at above ten deaths per 1 000 live births.
Definition and comparability
In some large partner countries (India, South Africa and
Indonesia), infant mortality remains above 20 deaths per The infant mortality rate is the number of deaths of
1 000 live births, although in these three countries infant children under one year of age, expressed per 1 000 live
mortality has reduced considerably in recent decades. births. Some of the international variation in infant
Indeed, infant mortality has fallen significantly in all mortality rates is related to variations in registering
OECD and partner countries, with reductions since 1990 practices for very premature infants. While some
particularly large in Slovenia, Estonia, Poland, Korea and countries register all live births including very small
China. babies with low odds of survival, several countries
Despite this progress in reduced infant mortality, apply a minimum threshold of a gestation period of
increasing numbers of low birth weight infants is a concern 22 weeks (or a birth weight threshold of 500 g) for
in some OECD countries. In a number of countries, this babies to be registered as live births (Euro-Peristat,
has contributed to a levelling-off of the downward trend 2013). To remove this data comparability limitation,
in infant mortality over the past few years. On average, the data presented in this section are based on a
one in 15 babies born in the OECD (or 6.5% of all births) minimum threshold of 22 weeks of gestation period
weighed less than 2 500 grams at birth in 2015 (Figure 3.14). (or 500 grams birth weight) for a majority of OECD
In almost all OECD countries, the proportion of low birth countries that have provided these data. However, the
weight infants has increased over the past two decades, data for some countries (e.g., Canada and Australia)
mainly due to increases in pre-term births (Euro-Peristat, continue to be based on all registered live births,
2013). Korea, Spain, Portugal, Greece and Japan have seen resulting in some over-estimation.
large increases (50% or more) of low birth weight babies Low birth weight is defined by the World Health
since 1990, although the proportions remain below the Organization as the weight of an infant at birth of
OECD average in Korea. less than 2 500 grams (5.5 pounds) irrespective of the
Low birth weight can occur as a result of restricted foetal gestational age of the infant. This threshold is based on
growth or from pre-term birth. Low birth weight infants epidemiological observations regarding the increased
have a greater risk of poor health or death, require a longer risk of death to the infant and serves for international
period of hospitalisation after birth, and are more likely comparative health statistics. The number of low
to develop significant disabilities. Risk factors for low weight births is expressed as a percentage of total
birth weight include maternal smoking, excessive alcohol live births.
consumption, poor nutrition, low body mass index, lower
socio-economic status, having had in-vitro fertilisation
treatment and multiple births, and a higher maternal age. References
The increased use of delivery management techniques
such as induction of labour and caesarean delivery, which Euro-Peristat (2013), European Perinatal Health Report: The
have increased the survival rates of low birth weight babies, Health and Care of Pregnant Women and their Babies in
may also explain the rise in low birth weight infants. 2010, Luxembourg.
Despite the widespread use of a 2 500 grams limit for low NCHS – National Centre for Health Statistics (2016), Health,
birth weight, physiological variations in size occur across United States, 2015, with Special Feature on Racial and Ethnic
different countries and population groups, and these need Health Disparities, NCHS, Hyattsville, United States.
58 HEALTH AT A GLANCE 2017 © OECD 2017
3. HEALTH STATUS
Infant health
A corrigendum has been issued for this page. See: http://www.oecd.org/about/publishing/Corrigendum_Health_at_a_Glance_2017.pdf
3.13. Infant mortality, 2015 and change 1990-2015 (or nearest year)
1.6 Slovenia -81
1.7 Finland -70
2.0 Iceland¹ -65
2.1 Japan -54
2.3 Norway -67
2.5 Czech Republic -77
2.5 Estonia -80
2.5 Sweden -58
2.7 Korea -78
2.7 Spain -64
2.9 Italy -64
2.9 Portugal -73
3.1 Austria -60
3.1 Israel -69
3.2 Luxembourg¹ -63
3.2 Australia -61
3.3 Belgium -59
3.3 Germany -53
3.3 Netherlands -54
3.4 Ireland -59
3.7 Denmark -51
3.7 France -49
3.9 Switzerland -43
3.9 United Kingdom -51
3.9 OECD35 -64
4.0 Greece -59
4.0 Poland -79
4.1 Latvia -70
4.2 Hungary -72
4.2 Lithuania -59
4.8 Canada -29
5.0 New Zealand -40
5.1 Slovak Republic -58
5.8 United States -37
6.5 Russian Federation -63
7.2 Chile -55
8.5 Costa Rica -41
9.2 China -78
10.7 Turkey -74
12.5 Mexico -62
13.6 Colombia -53
14.6 Brazil -71
22.8 Indonesia -63
33.6 South Africa -29
37.9 India -57
40 30 20 10 0 -100 -80 -60 -40 -20 0
Deaths per 1 000 live births in 2015 Change 1990-2015 in %
1. Three-year average (1988-90 and 2013-15).
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602462
3.14. Low birth weight infants, 2015 and change 1990-2015 (or nearest year)
2.4 China -59
4.0 Iceland¹ 38
4.2 Estonia -5
4.2 Finland 17
4.4 Sweden 2
4.5 Latvia 0
4.6 Lithuania 24
4.7 Norway 2
5.0 Denmark -4
5.6 Ireland 33 119
5.7 Korea
5.8 Mexico -31
5.8 Netherlands 4
5.8 Poland -31
5.9 New Zealand -5
6.2 Chile 15
6.2 France 5
6.2 Russian Federation 9
6.3 Canada 15
6.4 Australia 5
6.4 Slovenia 23
6.5 Austria 16
6.5 Switzerland 27
6.5 OECD35 15
6.6 Luxembourg¹ 18
6.6 Germany 16
6.9 Belgium 13
6.9 United Kingdom 3
7.3 Costa Rica 12
7.4 Italy 32
7.7 Slovak Republic 33
7.8 Czech Republic 42
7.9 Spain 76
8.0 Israel 11
8.1 United States 13
8.5 Brazil -18
8.6 Hungary -8
8.6 Turkey 9
8.9 Portugal 59
9.2 Greece 53
9.5 Japan 51
9.5 Colombia 32
11.1 Indonesia 22
12 10 8 6 4 2 0 -100 -50 0 50 100
% of total live births in 2015 Change 1990-2015 in %
1. Three-year average (1988-90 and 2013-15).
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602481
HEALTH AT A GLANCE 2017 © OECD 2017 59
3. HEALTH STATUS
Mental health
Mental illness represents a considerable – and growing – declines in suicide can be attributed at least in part to
proportion of the global burden of disease. An estimated targeted mental health promotion and suicide prevention
one in two people will experience a mental illness in programmes, as well as to improved mental health care.
their lifetime, and around one in five working-age adults In some other countries suicides have increased in recent
suffer from mental ill-health at any given time (OECD, years. In Mexico the suicide rate increased from 4.8 per
2012; OECD, 2015). Depression alone affects millions of 100 000 population in 2010 to 5.5 in 2015, while in the
individuals each year. Figure 3.17 shows self-reported United States the rate rose from 12.5 to 13.5. A range of
prevalence of depression in Europe. On average, 12-month interventions can both prevent and treat depression, and
prevalence of depression was 7.9% of the population. prevent suicide, but in many countries people with mental
Women reported higher rates of depression in all countries; ill-health have difficulties accessing appropriate mental
in Spain, Lithuania, Hungary, Poland women were more health care in a timely way.
than 50% more likely to report experiencing depression
in the previous year than men, rising to 66% in Portugal.
People in Iceland or Ireland were close to three times Definition and comparability
more likely to report depression than people in the Czech
Republic (Figure 3.17). These differences are in part driven The registration of a suicide is a complex procedure,
by different attitudes and understandings around mental which is affected by factors including how intent is
ill-health and depression. Lower stigma around depression ascertained, who is responsible for completing the
may contribute to higher rates of self-reported illness, and death certificate, and cultural dimensions including
higher rates of diagnosis. stigma around suicide. Caution is therefore needed
When people are suffering from a mental disorder, it has when comparing suicide rates between countries.
significant consequences across their lives, contributing Mortality rates are based on numbers of deaths divided
to poorer educational outcomes, higher rates of by the size of the corresponding population. The rates
unemployment, and poorer physical health. In serious have been age-standardised to the OECD population.
cases depression and other mental illnesses, such as The source is the WHO Mortality Database; suicides are
bipolar disorder and schizophrenia, can lead to people classified under ICD-10 codes X60-X84, Y870.
harming themselves, or even dying from suicide (McDaid et Estimates of the prevalence of depression are derived
al., 2017). There are other complex reasons that contribute from the second wave of the European Health
to the rate of death by suicide. The social context, poverty, Interview Survey. Respondents were asked: “During
substance abuse, and unemployment are all associated the past 12 months, have you had any of the following
with higher rates of suicide. diseases or conditions?” with the list including
Suicide remains a significant cause of death in many depression. Self-reported data on depression may
OECD countries. Figure 3.15 shows that in 2015 suicide be subject to under-diagnosis and reporting errors.
rates were lowest in South Africa, Turkey, Greece and Studies from several European countries show more
Colombia with fewer than five deaths by suicide per variation between countries in self-reported data on
100 000 population. Lithuania had the highest suicide mental illness than on other survey methods.
rate, with 29 deaths per 100 000, followed by Korea and
the Russian Federation. Some caution is needed when
comparing suicide rates. Stigma associated with suicide, References
or problems with recording suicides mean that in some
countries deaths by suicide may be under-reported. McDaid, D., Hewlett, E. and A. Park (2017), “Understanding
Unlike depression prevalence, mortality rates for suicide Effective Approaches to Promoting Mental Health
are three-to-four times higher for men than for women. and Preventing Mental Illness”, OECD Health Working
Studies suggest that the gender gap for attempted suicide Papers, OECD Publishing, Paris, http://dx.doi.org/10.1787/
is smaller, but men tend to use more lethal means when bc364fb2-en.
attempting suicide. OECD (2015), Fit Mind, Fit Job: From Evidence to Practice in
Suicide rates have decreased steadily across the OECD, Mental Health and Work, OECD Publishing, Paris, http://
falling by close to 30% between 1990 and 2015. In some dx.doi.org/10.1787/9789264228283-en.
countries the declines have been significant, including in OECD (2012), Sick on the Job? Myths and Realities about Mental
Estonia, Finland and Hungary where suicide rates have Health and Work, Mental Health and Work, OECD Publishing,
fallen by 40% or more (Figure 3.16). In Finland significant Paris, http://dx.doi.org/10.1787/9789264124523-en.
60 HEALTH AT A GLANCE 2017 © OECD 2017
3. HEALTH STATUS
Mental health
3.15. Suicide, 2015 (or nearest year)
Total Men Women
Age-standardised rates per 100 000 population
50
40
30
20
10
0
Tu ca
C o Br o
a
Un Z e blic
Po ce
r tu c
E s um
Ja ia
La n
G ey
Is a
lo e
M ael
aR l
a
ng y
ov m ain
pu ¹
Ne C al
er le
No nd s
n y
rm k
C a ny
Ir e d a
Sw and
OE en
Sw st 5
e c z e li a
N e R e p nd
d nd
Au tes
I c r ia
F i n d¹
Fr n d
Be land
ov a
de ar y
K n
th a
a
m
st a zi
Re u r g
i a Hu eni
Po bli
bi
De r wa
Sl t v i
ni
Co r eec
L i or e
K i It al
ic
Ge mar
ic
pa
tio
Au D3
t h hi
n
g
do
rk
ri
an
Cz it ra
ed
a
na
h rla
a
ite ala
st
r
m
Sl uxe Sp
to
ex
ua
a
F e ng
a
w u
nl
l
la
lg
Af
ra
ak bo
St
el
h
ut
d
So
ite
n
L
Un
ss
Ru
1. Three-year average.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602500
3.16. Trends in suicide, selected OECD countries, 1990-2015
Estonia Finland Israel Japan Korea Mexico
Hungary OECD United States OECD
Age-standardised rates per 100 000 population Age-standardised rates per 100 000 population
45 45
40 40
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
0
02
05
08
11
14
02
05
08
11
14
9
9
20
20
20
20
20
20
20
20
20
20
19
19
19
19
19
19
19
19
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933602519
3.17. Prevalence of chronic depression, 2014
% Men Women
20
18
16
14
12
10
8
6
4
2
0
en
ia
26
ce
ic
ic
l
nd
ia
r ia
ay
ly
ce
nd
a
ga
an
ke
ni
ni
ar
nd
ar
ai
an
an
ur
en
bl
bl
tv
It a
iu
do
rw
an
ee
ed
la
la
EU
st
Sp
r tu
ua
to
nm
ng
r
bo
rm
pu
pu
el
nl
La
la
lg
ov
Tu
Ir e
Po
ng
Au
Gr
Fr
Sw
No
Es
th
Ic
er
Fi
Be
Hu
m
Po
Re
Re
Ge
De
Sl
Ki
Li
th
xe
h
ak
Ne
d
Lu
ec
ite
ov
Cz
Un
Sl
Note: Self-reported prevalence of depression in the past 12 months.
Source: Eurostat Database, 2017.
12 http://dx.doi.org/10.1787/888933602538
HEALTH AT A GLANCE 2017 © OECD 2017 61
3. HEALTH STATUS
Perceived health status
Most OECD countries conduct regular health surveys which causal link is also possible, with poor health status leading
allow respondents to report on different aspects of their to lower employment and lower income.
health. A commonly asked question is of the type: “How Greater emphasis on public health and disease prevention
is your health in general?”. Despite the subjective nature among disadvantaged groups, and improving access to
of this question, indicators of perceived general health are health services may contribute to further improvements
a good predictor of people’s future health care use and in population health status in general and reducing health
mortality (Palladino et al., 2016). inequalities.
For the purpose of international comparisons, cross-
country variations in perceived health status are difficult
to interpret because responses may be affected by the
Definition and comparability
formulation of survey questions and responses, and by
social and cultural factors. For example, a central tendency Perceived health status reflects people’s overall
bias in self-reporting health has been noted in Japan and perception of their health. Survey respondents are
Korea (Lee et al., 2003). In addition, since older people report typically asked a question such as: “How is your
poor health more often than younger people, countries health in general?”. Caution is required in making
with a larger proportion of aged persons will also have a cross-country comparisons of perceived health status
lower proportion of people reporting to be in good health. for at least two reasons. First, people’s assessment
With these limitations in mind, in almost all OECD of their health is subjective and can be affected by
countries a majority of adults report being in good health cultural factors. Second, there are variations in the
(Figure 3.18). New Zealand, Canada, the United States and question and answer categories used to measure
Australia are the four leading countries, with more than perceived health across surveys and countries. The
85% of people reporting to be in good health. However, the response scale used in the United States, Canada, New
response categories offered to survey respondents in these Zealand, Australia and Chile is asymmetric (skewed on
four countries are different from those used in European the positive side), including the following response
countries and Asian OECD countries, which introduce an categories: “excellent, very good, good, fair, poor”.
upward bias (see box on “Definition and comparability”). In most other OECD countries the response scale
On the other hand, less than half of adults in Japan, is symmetric, with response categories being: “very
Korea, Latvia and Portugal rate their health as being good. good, good, fair, poor, very poor”. In Israel, the scale
The proportion is also relatively low in Estonia, Hungary, is symmetric but there is no middle category related to
Poland and Chile, where less than 60% of adults consider “fair” health. Such differences in response categories
themselves to be in good health. In many of these cases, bias upward the results from those countries that are
though, adults consider themselves to be in fair health. A using an asymmetric scale or a symmetric scale but
potentially clearer distinction is on adults who consider without any middle category.
themselves to be in bad health. Across the OECD, on average Self-reported health by income level is reported for
9% of adults consider themselves to be in bad health. The the first quintile (lowest 20% of income group) and the
share is over 15% in Portugal, Korea, Latvia, Israel, Hungary fifth quintile (highest 20%). Depending on the surveys,
and Estonia. the income may relate either to the individual or the
In all OECD countries, men are more likely than women household (in which case the income is equivalised
to report being in good health, except in New Zealand, to take into account the number of persons in the
Canada and Australia where the proportion is almost equal. household).
As expected, people’s rating of their own health tends to
decline with age. In many countries, there is a particularly
marked decline in how people rate their health after age References
45 and a further decline after age 65.
Lee,Y. et al (2003), “A Comparison of Correlates of Self-rated
There are large disparities in self-reported health across
Health and Functional Disability of Older Persons in the
different socio-economic groups. Figure 3.19 shows that,
Far East: Japan and Korea”, Archives of Gerontology and
in all countries, people with a lower level of income tend
Geriatrics, Vol. 37, pp. 63-76.
to report poorer health than people with higher income,
although the gap varies. On average across OECD countries, Lumsdaine, R. and A. Exterkate (2013), “How Survey Design
nearly 80% of people in the highest income quintile report Affects Self-assessed Health Responses in the Survey
being in good health, compared with just under 60% for of Health, Ageing, and Retirement in Europe (SHARE)”,
people in the lowest income group. These disparities European Economic Review, Vol. 63, pp. 299-307.
may be explained by differences in living and working Palladino, R. et al. (2016), “Associations Between
conditions, as well as differences in smoking and other risk Multimorbidity, Healthcare Utilisation and Health
factors. People in low-income households may also have Status: Evidence from 16 European Countries”, Age
limited access to certain health services for financial or and Ageing, Vol. 45, pp. 431-435, http://dx.doi.org/10.1093/
other reasons (see Chapter 5 on “Access to care”). A reverse ageing/afw044.
62 HEALTH AT A GLANCE 2017 © OECD 2017
Ne Ne
w w
60
20
30
40
50
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
Ze Ze
al al
a a
U n C a n d¹ U n C a n d¹
ite na 88.9 ite na 88.9 8.7 2.4
d da d da
St ¹ 88.1 8.8 3.1
St ¹ 88.1 Au ate
Au a tes
st ¹ st s¹ 88.1 9.2 2.7
ra 88.1 ra
l li
resulting in an upward bias.
Ire ia¹ 85.2 Is a¹ 85.2 10.4 4.4
la ra
S nd Ir e e l ¹ 83.9 16.0
Sw we 82.3 la
HEALTH AT A GLANCE 2017 © OECD 2017
i t z de n S n
er 79.7 Sw we d 82.3 13.8 3.9
la i t z de n
N o nd
79.3
er
la 79.7 16.5 3.8
rw
a N o nd
79.3 16.9 3.8
% of population aged 15 years and over
N e Ic e l y 78.3 rw
th an a
er d
76.3
N e Ic e y 78.3 15.3 6.4
la th lan
er d
Be nd s
76.2 la 76.3 18.0 5.7
lg
iu Be nd s
lg 76.2 18.5 5.2
Gr m
ee 74.6 iu
Good or very good
Good or very good
c Gr m 74.6 16.0 9.4
S e 74.4 ee
c
De pain S e 74.4 15.6 10.0
Lu nm 72.4 De pai
xe a
m rk Lu nm n 72.4 20.2 7.5
bo 71.6 xe a
Source: OECD Health Statistics 2017 (EU-SILC for European countries).
Source: OECD Health Statistics 2017 (EU-SILC for European countries).
u m rk
bo 71.6 20.8 7.6
% of population aged 15 years and over reporting to be in good health
Au rg
st 70.4 u
Un Au rg 70.4 20.9 8.7
i t e F i n r ia 69.8 Un st
d i t e F i r ia
K i lan d
ng d nla 69.8 21.4 8.8
do 69.8 Ki n
ng d
resulting in an upward bias. In Israel, there is no category related to fair health.
d 69.8 24.4 5.8
Fr m
an 69.8 OE om
69.8 21.6 8.7
Fair
OE ce CD
CD 67.8 3
Sl
ov T 3 3 Sl
Fr 4
a 68.2 23.1 9.4
67.7
Highest income
a k ur ov T n c e
Re ke y a k ur 67.8 24.4 7.8
pu 66.4 Re ke y
bl
ic
pu 66.4 21.6 12.0
bl
I 65.9 ic
Sl t al y I 65.9 21.8 12.3
65.6
Note: Countries are ranked in descending order of perceived health status for the whole population.
ov Sl t al y
C z Ge eni ov 65.6 22.0 12.4
C z G eni
ec rm a
h 64.8
R e a ny
ec erm a
h 64.8 23.6 11.6
pu 64.5 R e a ny
27.1 8.3
b pu 64.5
b
Po lic
la 61.2 Po lic 61.2 27.7 11.1
nd la
57.8 n
Ch Ch d 57.8 27.9 14.3
Hu ile¹ Hu le i
57.4
3.18. Perceived health status among adults, 2015 (or nearest year)
ng ng ¹ 57.4 35.4 7.3
Lowest income
Bad or very bad
a
3.19. Perceived health status by income level, 2015 (or nearest year)
Es r y a
to 56.3 Es r y
to 56.3 27.9 15.8
Po ni a Po nia
r tu 51.4 r tu 51.4 32.8 15.7
ga g
La l 46.4 L a al 46.4 35.6 18.0
tv tv
i i
Ja a 46.2 Ja a 46.2 37.5 16.3
pa pa
Ko n 35.4 Ko n 35.4 49.1 14.5
re re
a a 32.5 49.9 17.6
32.5
1. Results for these countries are not directly comparable with those for other countries, due to methodological differences in the survey questionnaire
63
1. Results for these countries are not directly comparable with those for other countries, due to methodological differences in the survey questionnaire
12 http://dx.doi.org/10.1787/888933602557
Perceived health status
3. HEALTH STATUS
12 http://dx.doi.org/10.1787/888933602576
3. HEALTH STATUS
Cancer incidence
In 2012, an estimated 5.8 million new cases of cancer were OECD countries, prostate cancer accounted for 24% of all
diagnosed in OECD countries, 54% (around 3.1 million) new cancer diagnoses in men in 2012, followed by lung (14%)
occurring in men and 46% (around 2.7 million) in women. and colorectal (12%). Similar to breast cancer, the causes
The most common were breast cancer (12.9% of all new of prostate cancer are not well-understood but age, ethnic
cancer cases) and prostate cancer (12.8%), followed by lung origin, family history, obesity, lack of exercise and poor
cancer (12.3%) and colorectal cancer (11.9%). These four nutrition are the main risk factors. Incidence in 2012 was
cancers represented half of the estimated overall burden highest in Norway, Sweden, Australia and Ireland, with rates
of cancer in OECD countries (Ferlay et al., 2014). more than 50% higher than the OECD average (Figure 3.22).
Large variations exist in cancer incidence across OECD Greece had the lowest rates, followed by Mexico, Korea and
countries. Cancer incidence rates are highest in Denmark, Japan. Prostate cancer incidence rates have increased in
Australia, Belgium, Norway, United States, Ireland, Korea, most OECD countries since the late 1990s with increased
Netherlands and France registering more than 300 new use of prostate specific antigen (PSA) tests having led to
cancer cases per 100 000 population in 2012 (Figure 3.20). greater detection (Ferlay et al., 2014). Differences between
The lowest rates were reported in some Latin American countries’ rates can be partly attributed to differences in
and Mediterranean countries such as Mexico, Greece, the use of PSA testing. Mortality rates from prostate cancer
Chile and Turkey, with around 200 new cases or less per have decreased in some OECD countries as a consequence
100 000 population. These variations reflect not only of early detection and improvements in treatments (see
variations in the prevalence of risk factors for cancer, indicator “Mortality from cancer” in Chapter 3).
but also national policies regarding cancer screening and
differences in quality of reporting.
Cancer incidence was higher for men in all OECD countries Definition and comparability
in 2012 except in Mexico. However, the gender gap varies
widely across countries. In Turkey, Estonia and Spain, Cancer incidence rates are based on numbers of
incidence among men were around 60% higher than among new cases of cancer registered in a country in a year
women, whereas in the United Kingdom, Denmark and per 100 000 population. The rates have been directly
Iceland, the gap was less than 10%. age-standardised based on Segi’s world population
to remove variations arising from differences in age
Breast was by far the most common primary sites in
structures across countries and over time. The data
women (28% on average), followed by colorectal (12%),
come from the International Agency for Research on
lung (10%), and cervical (3%). The causes of breast cancer
Cancer (IARC), GLOBOCAN 2012, available at globocan.
are not fully understood, but the risk factors include age,
iarc.fr. GLOBOCAN estimates for 2012 may differ from
family history, breast density, exposure to oestrogen, being
national estimates due to differences in methods.
overweight or obese, alcohol intake, radiation and hormone
replacement therapy. Incidence rates in 2012 were highest Cancer registration is well established in most OECD
in Belgium, Denmark and Netherlands, with rates 25% or countries, although the quality and completeness
more than the OECD average (Figure 3.21). Chile and Mexico of cancer registry data may vary. In some countries,
had the lowest rate, followed by Turkey and Greece. The cancer registries only cover subnational areas. The
variation in breast cancer incidence across OECD countries international comparability of cancer incidence data
may be at least partly attributed to variation in the extent can also be affected by differences in medical training
and type of screening activities. Although mortality rates and practice.
for breast cancer have declined in most OECD countries The incidence of all cancers is classified to ICD-
since the 1990s due to earlier detection and improvements 10 codes C00-C97 (excluding non-melanoma skin
in treatments, breast cancer continues to be one of the cancer C44). Breast cancer corresponds to C50, and
leading causes of death from cancer among women prostate cancer to C61.
(see indicator “Mortality from cancer” in Chapter 3 and
“Screening, survival and mortality from breast cancer” in
Chapter 6).
References
Prostate cancer has become the most commonly diagnosed
cancer among men in almost all OECD countries, except in Ferlay, J. et al. (2014), “Cancer Incidence and Mortality
Hungary, Poland, Turkey and Greece where lung cancer is Worldwide: Sources, Methods and Major Patterns in
still predominant, and in Japan and Korea where colorectal GLOBOCAN 2012”, International Journal of Cancer, Vol. 136,
cancer is the main cancer among men. On average across No. 5, pp. E359-E386.
64 HEALTH AT A GLANCE 2017 © OECD 2017
3. HEALTH STATUS
Cancer incidence
3.20. All cancers incidence by gender, 2012
Total Men Women
Age-standardised rates per 100 000 population
400
350
300
250
200
338.1
323.0
321.1
318.3
318.0
307.8
307.9
304.8
303.5
296.3
295.7
295.0
293.8
285.4
284.3
287.0
283.8
283.2
280.3
278.6
276.9
272.9
270.5
270.0
256.8
254.1
249.0
246.2
242.8
150
229.6
217.1
205.1
175.7
163.0
100
131.5
50
Gr o
Ze c
ia
ic
th nce
ce
Tu e
ey
Po n
Es d
Po i a
Sp l
A u in
F i ia
Sw d
Un O de n
m y
g
Ge ael
Ic y
Hu d
C z i t z ar y
Ca d
Re m
Sl da
U n Ir e a
St d
No es
Be ay
A u ium
De a li a
k
Sl d K i D 3 4
ga
w bli
an
il
xe t al
re
nd
ar
ic
h lan
pa
an
an
ur
an
ite lan
bl
en
r
n
ak do
rk
rw
ee
Ch
at
na
la
st
r
ex
to
r tu
Ko
nm
S w ng
bo
r
rm
pu
Ne epu
Ne r a
e
I
nl
Ja
el
al
la
Is
lg
ov
ite EC
ec er
st
ov ng
M
er
F
d
R
Lu
Source: International Agency for Research on Cancer (IARC), GLOBOCAN 2012.
12 http://dx.doi.org/10.1787/888933602595
3.21. Breast cancer incidence in women, 2012 3.22. Prostate cancer incidence in men, 2012
Belgium 111.9 Norway 129.7
Denmark 105.0 Sweden 119.0
Netherlands 99.0 Australia 115.2
Iceland 96.3 Ireland 114.2
United Kingdom 95.0 Switzerland 107.2
United States 92.9 Iceland 106.6
Ireland 92.3 United States 98.2
Germany 91.6 France 98.0
Italy 91.3 Finland 96.6
France 89.7 Estonia 94.4
Finland 89.4 New Zealand 92.2
Luxembourg 89.1 Denmark 91.3
Australia 86.0 Belgium 90.9
New Zealand 85.0 Canada 88.9
Switzerland 83.1 Israel 84.3
Israel 80.5 Netherlands 83.4
Sweden 80.4 Slovenia 82.9
Canada 79.8 Luxembourg 78.8
OECD34 74.2 Germany 77.3
Norway 73.1 OECD34 76.2
Czech Republic 70.3 Austria 74.7
Austria 68.0 United Kingdom 73.2
Portugal 67.6 Czech Republic 72.2
Spain 67.3 Italy 67.6
Slovenia 66.5 Spain 65.2
Slovak Republic 57.5 Portugal 63.6
Hungary 54.5 Chile 52.4
Korea 52.1 Slovak Republic 50.0
Poland 51.9 Turkey 40.6
Estonia 51.6 Hungary 37.5
Japan 51.5 Poland 35.9
Greece 43.9 Japan 30.4
Turkey 39.1 Korea 30.3
Mexico 35.4 Mexico 27.3
Chile 34.8 Greece 20.2
0 25 50 75 100 0 25 50 75 100 125 150
Age-standardised rates per 100 000 population Age-standardised rates per 100 000 population
Source: International Agency for Research on Cancer (IARC), GLOBOCAN Source: International Agency for Research on Cancer (IARC), GLOBOCAN
2012. 2012.
12 http://dx.doi.org/10.1787/888933602614 12 http://dx.doi.org/10.1787/888933602633
HEALTH AT A GLANCE 2017 © OECD 2017 65
3. HEALTH STATUS
Diabetes prevalence
Diabetes is a chronic disease, characterised by high levels of countries and these countries spent an average of about
glucose in the blood. It occurs either because the pancreas USD 4 600 per diabetic adult in 2015 (IDF, 2015). These
stops producing the hormone insulin (Type 1 diabetes), or burdens highlight the need for effective management of
because the cells of the body do not respond properly to the diabetes and its complications (see indicator on “Diabetes
insulin produced (Type 2 diabetes). People with diabetes are care” in Chapter 6), as well as appropriate preventive
more likely to suffer from cardiovascular diseases such as actions (see Chapter 4).
heart attack and stroke, sight loss, foot and leg amputation
and renal failure.
Across the OECD, over 93 million adults – or 7% of all adults –
were diabetics in 2015 (Figure 3.23). The International Definition and comparability
Diabetes Federation estimates that a further 33 million
The sources and methods of the NCD Risk Factor
adults have undiagnosed diabetes in OECD countries.
Collaboration is described in the Lancet article and
Diabetes prevalence is highest in Mexico, where more than
appendix (Lancet, 2016). Sources were selected among
15% of adults have diabetes. Diabetes prevalence is also
population-based studies that had collected data on
high in Turkey, the United States and Chile, where 10%
measurement of diabetes biomarkers for Type 1 or
or more of adults were diabetics. In contrast, less than
Type 2 diabetics. Prevalence in sources were converted
5% of adults suffered from diabetes in Estonia, Ireland,
to meet the definition of diagnosed diabetes as defined
Luxembourg, Sweden and the United Kingdom. Among
in the Global Monitoring Framework for NCDs. Then,
partner countries, diabetes prevalence is relatively high in
Bayesian hierarchical models were applied to estimate
Brazil and Colombia, at about 10% of the adult population,
trends in prevalence. Adult’s population covers those
and low in Lithuania.
aged 18 years and over.
Diabetes prevalence has risen slowly or stabilised in the
The sources and methods used by the International
majority of OECD countries, especially in Western Europe,
Diabetes Federation are outlined in their Diabetes
but it has increased markedly in Turkey and most partner
Atlas, 7 th edition (IDF, 2015). Sources were only
countries (Figure 3.24) These trends mirror partly trends
included if they met several criteria for reliability. Age-
in population ageing, as well as the rise of obesity and
standardised rates were calculated using the world
physical inactivity, and their interactions (NCD Risk Factor
population based on the distribution provided by the
Collaboration, 2016). The share of obese people has been
World Health Organization. Adult’s population covers
increasing strongly all around the world, and especially in
those aged between 20 and 79 years old with Type 1
the BRIICS (see indicators on obesity in Chapter 4).
or Type 2 diagnosed diabetes.
Diabetes is slightly more common among men than women
and the prevalence increases substantially with age. For
example, in the United States the estimated share of
diagnosed diabetics was about 3% for those aged 20-44, 12%
References
for those aged 45-64 and 21% for those aged 65 years and
over (Menke et al., 2015). Diabetes also disproportionately IDF – International Diabetes Federation (2015), IDF Diabetes
affects those in lower socio-economic groups and people Atlas, 7th edition, International Diabetes Federation,
from certain ethnicities. Brussels.
Diabetes prevalence among children is much lower Menke, A. et al. (2015), “Prevalence of and Trends in Diabetes
than among adults (Figure 3.25). Nevertheless, almost Among Adults in the United States, 1988-2012”, Journal
230 000 children suffered from Type 1 diabetes in OECD of American Medical Association, Vol. 314, No. 10,
countries in 2015. In Finland, almost five children per 1 000 pp. 1021-1029, http://dx.doi.org/10.1001/jama.2015.10029.
were Type 1 diabetics. Prevalence rates were next highest NCD Risk Factor Collaboration (2016), “Worldwide Trends
in Sweden (2.6) and Norway (2). Korea and Japan had the in Diabetes Since 1980: A Pooled Analysis of 751
lowest rates amongst OECD countries. Population-based Studies with 4·4 Million Participants”,
Diabetes bears heavy consequences on communities. Over The Lancet, Vol. 387, pp. 1513-1530, http://dx.doi.org/10.1016/
700 000 people died partly because of diabetes in OECD S0140-6736(16)00618-8.
66 HEALTH AT A GLANCE 2017 © OECD 2017
0
1
2
3
4
5
Li
2
4
6
8
10
12
14
%
K
0
2
4
6
8
10
12
14
16
18
%
th
C o or e ua
lo a
1980
m 0.1 Es ni a
to 4.0
b
Ja ia 0.1 L u Ir n i a
pa xe ela 4.4
m nd
Ch n 0.2 Un bo 4.4
in
a ite Sw urg
Per 1 000 children
In 0.2 d e
K i de 4.7
di ng n
La a
t 0.3 A u do 4.7
M via st m
ex 0.3 4.7
i B e r a li
Tu co 0.4 lg a 5.1
rk iu
ey m
Ch 0.5 It 5.1
Source: IDF Atlas, 7th Edition, 2015.
il e Gr a l y
5.1
HEALTH AT A GLANCE 2017 © OECD 2017
Sl 0.6 ee
Fr c e
ov G It al y
ak r e 0.8 an 5.2
Ne L ce
Re e c 5.3
Turkey
pu e 0.8 th at v
b er ia
1990
la 5.4
Source: NCD Risk Factor Collaboration (2016).
S w Fr a lic 0.8 n
Note: Data cover those aged under 14 years old.
it z nc Ja d s 5.5
e e 0.9 F i p an
Li rlan
th d
nl 5.7
u 0.9 No an d
Po ani rw 6.0
r tu a 0.9 S w Ic e ay
6.0
g i t z lan
B al 0.9 er d 6.1
Sl r a z la
ov il 1.0 P nd
e In o l a 6.1
Lu B e l nia do nd
x e gi u 1.0 ne 6.2
N e mb m A u si a
1.0 6.5
BRIICS
w ou OE stri
Ru Ze rg
ss al 1.1 C a 6.9
ia a De D 3
n Po n d nm 5 7.0
F e la 1.1 a
de nd Ko r k 7.2
2000
ra
t 1.1 N e Hu r e a
A ion 7.2
Cz O ustr 1.1 w nga
Ze r y
e c E C ia
1.2
Cz al 7.3
h ec Ca and
Note: Data cover those aged 18 years and over with Type 1 or Type 2 diagnosed diabetes.
Re D 3 h n 7.3
pu 5 1.2 Re a d
bl p a 7.4
S ic 1.2 Ge ubl
rm ic
Hu pain 7.4
ng 1.3 So I any
ar 7.4
Sweden
Un u t sr a
Note: Data cover those aged between 20 and 79 years old with Type 1 or Type 2 diagnosed diabetes.
i t e Isr y 1.3 h e
d a Sl
ov
Af l
r 7.5
3.23. Share of adults with diabetes, 2015
Ne S t el
th a te 1.3 ak S i c a 7.6
er s Re pai
l
Source: OECD estimates based on IDF Atlas, 7th Edition, 2015 and the United Nations population statistics.
1.4 p n 7.7
Au and Ru S ub
st s
r 1.4 s s C lov lic 7.8
E s a li a i a o s en
n t a ia
1.4
3.24. Trends in share of adults with diabetes, 1980-2014
Fe R 7.8
2010
Ge ton de ic
rm ia 1.4 ra a 8.5
a tio
Ic n y
e 1.5 In n 9.2
De lan d
C ia 9.3
3.25. Share of children with Type 1 diabetes per 1000 population, 2015
Un nm d 1.5 Po hin
i te C ar k r tu a 9.8
d an 1.6
OECD35
Ki ad ga
ng a 1.6 Co Ch l 9.9
do lo il e
Ir e m m 10.0
l 1.7 Un b
No and i t e Br i a 10.0
rw 1.7 d az
S t il
S w ay a 10.4
ed 2.0 Tu tes
F i en rk 10.8
nl 2.6 M ey
an
d
ex 12.8
4.6 ic
o
2014
15.8
67
12 http://dx.doi.org/10.1787/888933602690
12 http://dx.doi.org/10.1787/888933602671
12 http://dx.doi.org/10.1787/888933602652
Diabetes prevalence
3. HEALTH STATUS
4. RISK FACTORS FOR HEALTH
Smoking among adults
Alcohol consumption among adults
Smoking and alcohol consumption among children
Healthy lifestyles among adults
Healthy lifestyles among children
Overweight and obesity among adults
Overweight and obesity among children
Air pollution
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.
Health at a Glance 2017 © OECD 2017 69
4. RISK FACTORS FOR HEALTH
Smoking among adults
The health consequences of tobacco smoking are stringent policies led to strong reductions in smoking
numerous, and include cancers, stroke, and coronary heart rates between 1996 and 2011 in many OECD countries
disease, among others. It is also an important contributory (OECD, 2015). In 2014, 29 OECD countries applied tobacco
factor for respiratory diseases, such as chronic obstructive advertising bans on at least national television, print media
pulmonary disease (COPD), while smoking among pregnant and radio, while 26 countries applied taxation rates of at
women can lead to low birth weight and illnesses among least 70% (WHO, 2015). In all OECD countries, packages
infants. Smoking causes the largest share of overall years displayed at least a medium-sized a health warning. Every
of healthy life lost in 15 OECD countries, and ranks second year on May 31st, “World No Tobacco Day” advocates for
in another 16 OECD countries (Forouzanfar et al., 2016). The effective policies to reduce tobacco consumption, and
WHO has estimated that tobacco smoking kills 7 million highlights the health and additional risks associated with
people per year across the world, of which 890,000 are due tobacco use. The theme for 2017 was “Tobacco – a threat to
to second-hand smoke. It is the leading cause of death, development”, and focused on the threats of the industry
illness and impoverishment. to the sustainable development of countries.
Across the OECD, just over 18% of adults smoke tobacco
daily (14% of women and 23% of men) (Figure 4.1). Rates are
highest in Greece, Hungary and Turkey, as well as Indonesia Definition and comparability
(over 25%), and lowest in Mexico as well as Brazil (under 10%).
Women smoke the most in Austria, Greece and Hungary, The proportion of daily smokers is defined as the
where rates exceed 20%, while they smoke the least in Korea percentage of the population aged 15 years and over
and Mexico, as well as China, India, and Indonesia, where who report tobacco smoking every day. Other forms
rates are below 5%. In men, rates are highest in Turkey of smokeless tobacco products, such as snuff in
as well as China, Indonesia, and the Russian Federation Sweden, are not taken into account. This indicator
(exceeding 40%), while they are below 10% in Iceland as well is more representative of the smoking population
as Brazil. Men smoke more than women in all countries than the average number of cigarettes smoked per
except Denmark and Iceland, where the gender gap is about day, as the act of smoking is more determining than
one percentage point. In other countries, the gender gap the quantity. Most countries report data for the
ranges from below 2 points in Sweden and the United States, population aged 15 +, but there are some exceptions
to over 30 points in China (46 points), Indonesia (73 points) as highlighted in the data source of the OECD Health
and the Russian Federation (34 points). Statistics database.
Daily smoking has decreased in most OECD countries
since 2000, although rates have slightly risen in the Slovak
Republic (+0.8 points), have greatly increased in Indonesia References
(+8.8 points), and have not changed in Austria (Figure 4.2).
In 2015, an average of 18% of adults smoked daily in the Forouzanfar, M.H. et al. (2016), “Global, Regional, and
OECD, as opposed to 26% in 2000, equivalent to a 28% drop. National Comparative Risk Assessment of 79 Behavioural,
The strongest decreases occurred in Denmark, Iceland, Environmental and Occupational, and Metabolic Risks or
Luxembourg, the Netherlands, New Zealand and the United Clusters of Risks, 1990–2015: A Systematic Analysis for
Kingdom, as well as Brazil, India, Lithuania and the Russian the Global Burden of Disease Study 2015”, The Lancet,
Federation, where they exceeded 10 percentage points, and Vol. 388, pp. 1659-1724.
in Norway, where the drop was 20 points. In 2015, rates were OECD (2015), Cardiovascular Disease and Diabetes: Policies for
highest in Greece, Hungary, Turkey and Indonesia (over 25%), Better Health and Quality of Care, OECD Publishing, Paris,
while they were lowest in Mexico and Brazil (under 10%). http://dx.doi.org/10.1787/9789264233010-en.
Raising taxes on tobacco is the most effective way to reduce WHO (2015), Report on the Global Tobacco Epidemic – Raising
tobacco use (WHO, 2015). High levels of taxes as well as Taxes on Tobacco, WHO, Geneva.
70 Health at a Glance 2017 © OECD 2017
0
5
10
25
30
15
20
35
40
45
0
10
20
30
40
50
60
70
80
Br Br
M a z il M a z il
ex 7.2 ex
Ic i c o Ic i c o
Co elan 7.6 Co elan
lo d lo d
m 10.2 m
bi bi
In a 11.1 In a
Un S w di a Un S w di a
ite e 11.2 ite e
d de n d de n
St St
a 11.2 a
No t es No t es
Au r wa 11.4 Au r wa
C o s tr y 12.0 C o s tr y
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
s t a li a s t a li a
12.4
Health at a Glance 2017 © OECD 2017
aR aR
% of population aged 15 years and over
% of population aged 15 years and over
Ne Ca ic a Ne Ca ic a
w na 13.4 w na
L Z d L Z d
Un ux e a l a 14.0 Un ux e a l a
i t e em a n d i t e em a n d
d bo 14.2 d bo
K i ur K i ur
ng g ng g
14.9
Total
Po dom Po dom
r r
2000
De tug 16.1 De tug
nm al nm al
a 16.8 a
Cz Ko r k 17.0 Cz Ko r k
ec Fi rea ec Fi rea
h nla
Re n 17.3 h nla
Re n
pu d 17.4 pu d
b b
J lic 18.2 J lic
OE apa OE apa
C n 18.2 C n
Be D 3 4 Be D 3 4
lg 18.4 lg
S l ium S l ium
ov ov
e 18.9 e
N e Ir e n i a N e Ir e n i a
18.9
Men
th lan th lan
So erl d So erl d
ut an 19.0 ut an
h ds h ds
Af
ric 19.0 Af
ric
Is a 19.0 Is a
ra ra
el el
L It 19.6 L It
S w ithu al y S w ithu al y
i t z ani 20.0 i t z ani
e a 20.4 e a
Ge rlan Ge rlan
rm d rm d
20.4
Es any Es any
to 20.9 to
Fr nia Fr nia
Sl 21.3 Sl
2015
ov P a n c ov P a n c
ak o e 22.4 ak o e
Ru Re lan Ru Re lan
Women
ss pu d 22.7 ss pu d
ia bl ia bl
n i n i
F e Sp c
de ain 22.9 F e Sp c
de ain
4.2. Adult population smoking daily, 2000 and 2015 (or nearest year)
ra 23.0 ra
4.1. Adult population smoking daily by gender, 2015 (or nearest year)
ti ti
L a on 23.1 L a on
t t
Au via Au via
st 24.1 st
r r
C ia 24.3 C ia
Hu hin Hu hin
ng a 24.7 ng a
Gr a r y Gr a r y
ee 25.8 ee
T ce T ce
In u r k 27.3 In u r k
do ey do ey
ne 27.3 ne
si
si
a 39.9 a
71
12 http://dx.doi.org/10.1787/888933602728
12 http://dx.doi.org/10.1787/888933602709
Smoking among adults
4. RISK FACTORS FOR HEALTH
4. RISK FACTORS FOR HEALTH
Alcohol consumption among adults
Harmful alcohol use is a leading cause of death and drivers, but the enforcement of these regulations may be
disability worldwide, particularly in those of working age haphazard and varies widely across and within countries.
(OECD, 2015). Alcohol use is among the top ten leading risk Less stringent policies include health promotion messages,
factors in terms of years of healthy life lost in 32 OECD school-based and worksite interventions and interventions
countries (Forouzanfar et al., 2016), and consumption in in primary health care settings. Comprehensive policy
OECD countries remains well above the world average. packages including fiscal measures, regulations and less
In 2015, alcohol use lead to 2.3 million deaths, caused by stringent policies are shown to be the most effective to
cancers, heart diseases and liver diseases, among others. reduce harmful use of alcohol (OECD, 2015).
Most alcohol is drunk by the heaviest-drinking 20% of
the population. Heavy drinking is associated with a lower
probability of employment, more absence from work, and Definition and comparability
lower productivity and wages.
On average, recorded alcohol consumption has decreased Recorded alcohol consumption is defined as annual
in the OECD since 2000 (Figure 4.3), from 9.5 litres per sales of pure alcohol in litres per person aged 15 years
capita per year to 9 litres of pure alcohol per capita each and over. Most countries report data for the population
year, equivalent to 96 bottles of wine. The extent of the aged 15 +, but there are some exceptions as highlighted
decrease varies greatly by country, and consumption has in the data source of the OECD Health Statistics database.
in fact increased in thirteen OECD countries, as well as The methodology to convert alcohol drinks to pure
in China, India, Lithuania and South Africa. Consumption alcohol may differ across countries. Official statistics
increased by 0.1 to 1 litre in Canada, Chile, Israel, Korea, do not include unrecorded alcohol consumption, such
Mexico, Norway, Slovenia, Sweden and the United States, as home production. Unrecorded alcohol consumption
as well as in South Africa. The increase was stronger in and low quality of alcohol consumed (beverages
Belgium, Iceland, Latvia and Poland, as well as China, produced informally or illegally) remain a problem,
India and Lithuania (1.1 to 5.3 litres per capita). In all other especially when estimating alcohol-related burden of
countries, alcohol consumption decreased between 2000 disease among low income groups. The WHO reports
and 2015. The largest drops occurred in Denmark, Ireland, unrecorded alcohol consumption in their Global Health
Italy and the Netherlands (more than 2 litres per capita). Observatory data repository. In some countries (e.g.
Luxembourg), national sales do not accurately reflect
Although adult alcohol consumption per capita is a useful
actual consumption by residents, since purchases by
measure to assess long-term trends, it does not identify
non-residents may create a significant gap between
sub-populations at risk from harmful drinking patterns.
national sales and consumption. Alcohol consumption
Heavy drinking and alcohol dependence account for an
in Luxembourg is thus the mean of alcohol consumption
important share of the burden of diseases associated with
in France and Germany as recorded in the WHO-GISAH
alcohol. Across the OECD, an average of 12% of women
database.
and 30% of men take part in regular binge-drinking (at
least once per month) (Figure 4.4). Rates range from 8% Regular binge drinking is derived from self-reports of
in Hungary to 37% in Denmark, and display large gender the European Health Interview Survey 2014. Regular
gaps, with men exhibiting higher rates in virtually all binge drinking is defined as having six or more
countries. These gaps are lowest in Spain and Greece (8- alcoholic drinks per single occasion at least once a
10 points), and are highest in Estonia, Finland and Latvia month over the past 12 months.
(over 25 points).
Many policies addressing harmful use of alcohol already
exist: some target heavy drinkers only, while others are References
more broadly based. While all OECD countries apply taxes
to alcoholic beverages, the level of taxes may greatly vary Forouzanfar, M.H. et al. (2016), “Global, Regional, and
across countries. New forms of fiscal policies have been National Comparative Risk Assessment of 79 Behavioural,
implemented like minimum pricing of one unit of alcohol in Environmental and Occupational, and Metabolic Risks or
Scotland. Regulations on advertising alcoholic products have Clusters of Risks, 1990–2015: A Systematic Analysis for
been set up in many OECD countries, but the forms of media the Global Burden of Disease Study 2015”, The Lancet,
included in these regulations (e.g. printed newspapers, Vol. 388, pp. 1659-1724.
billboards, the internet) and the enforcement of the law OECD (2015), Tackling Harmful Alcohol Use: Economics and
vary a lot across countries. All OECD countries have legally Public Health Policy, OECD Publishing, Paris, http://dx.doi.
set maximum levels of blood alcohol concentration for org/10.1787/9789264181069-en.
72 Health at a Glance 2017 © OECD 2017
In
0
5
10
15
20
25
30
35
40
45
50
0
2
4
6
8
10
12
14
16
do
Hu ne
ng
Tu sia
ar
y rk 0.1
e
8.3
Is y 1.4
ra
Sp C o In e l 2.6
ai
n s t di a
a
9.3
Co Ri 3.1
Po lo c a
r tu m 3.5
ga M bia
l ex 4.4
Source: Eurostat EHIS 2014.
10.2
ic
Gr Ch o 5.2
ee No in a
Sl
ov ce rw 5.8
a
Source: OECD Health Statistics 2017.
10.3
ak
Litres per capita (15 years and older)
6.0
Health at a Glance 2017 © OECD 2017
Ch y
% of population aged 15 years and over
Re
pu i
Cz bl
Ja le 7.2
ec ic S w p an
e 7.2
12.8
h S o B de n
Re
pu ut r a 7.2
bl h zi
ic Af l
r 7.3
Total
14.9
Gr i c a
e 7.4
2000
Po Ic e c e
la el 7.5
nd an
d
17.4
Ne
th It a 7.5
Au er l y
st la 7.6
r ia Ca nds
n 8.0
18.7
Un F a d
Sl i t e inl a 8.1
ov
en N e d S an d
ia w ta 8.5
19.0
Ze tes
a 8.8
La OE lan
tv CD d 8.9
ia 3
19.2
Ko 5 9.0
Men
re
Sw S a 9.1
ed De pa
en Un S w nm in 9.3
20.4
i t e i t z ar
OE d erl k
K i an 9.4
CD ng d
Un
ite 18 Ru A u do 9.5
d s m
20.7
ss 9.5
Ki i a P tr a l
ng
do Sl n F e or t u i a 9.7
m ov d g
ak er a a l 9.9
22.1
Re tio
Es pu n 10.1
to Es blic
ni
a to 10.2
23.3
Po nia
10.3
2015
Be la
lg L nd
iu 10.5
m Hu a t v i
Women
27.5
Ge
ng a 10.8
rm Ir e a r y
10.9
G l
4.4. Regular binge-drinking (at least once a month) by gender, 2014
an L e an
y 10.9
33.0
C z uxe rma d
e c mb n y
Fi
nl
h o
Re ur 11.0
an
d p g 11.1
4.3. Recorded alcohol consumption among adults, 2000 and 2015 (or nearest year)
Lu
33.8
Sl ubli
xe ov c 11.5
m e
bo Fr nia
ur an 11.5
g Au ce
34.5
De 11.9
Be stri
nm a 12.3
L i l gi u
ar
k th m
ua
37.4 ni
12.6 15.2
a
73
12 http://dx.doi.org/10.1787/888933602766
12 http://dx.doi.org/10.1787/888933602747
Alcohol consumption among adults
4. RISK FACTORS FOR HEALTH
4. RISK FACTORS FOR HEALTH
Smoking and alcohol consumption among children
Smoking and excessive drinking during adolescence have Sweden. Rates have increased since 1997-98 for girls in the
both immediate and long-term health consequences. Czech Republic, Estonia, Hungary, Latvia and Poland.
Establishing smoking habits early on increases the risk Worldwide, one third of youth experimentation with
of cardiovascular diseases, respiratory illnesses, and tobacco occurs as a result of exposure to tobacco advertising,
cancer (Currie et al., 2012). Smoking during adolescence promotion and sponsorship (WHO, 2013). To reduce youth
has immediate adverse health consequences, including tobacco use, its use in the general population must be
addiction, reduced lung function and impaired lung growth, denormalised. Young smokers are responsive to policies
and asthma (Inchley et al., 2016). It is also associated with aiming to reduce tobacco consumption, including excise
an increased likelihood of experimenting with other drugs, taxes to increase prices, clean indoor-air laws, restrictions
as well as engaging in other risky behaviours (O’Cathail et on youth access to tobacco, and greater education about
al., 2011). Early and frequent drinking and drunkenness the effects of tobacco (Forster et al., 2007).
is associated with detrimental psychological, social and
physical effects, such as dropping out of high school
without graduating (Chatterji and DeSimone, 2005).
Definition and comparability
Results from the Health Behaviour in School-aged Children
(HBSC) surveys, a series of collaborative cross-national Estimates for smoking refer to the proportion of
studies, allow for monitoring of smoking and drinking 15-year-old children who self-report smoking at least
behaviours among adolescents. Other national surveys, once a week. Estimates for drunkenness refer to the
such as the Youth Risk Behavior Surveillance System in proportions of 15-year-old children who report that
the United States, or the Escapad survey in France, also they have been drunk twice or more in their lives.
monitor risky behaviours. The Health Behaviour in School-aged Children (HBSC)
Over 15% of 15-year-olds smoke at least once a week surveys were undertaken every four years between
in France, Hungary, Italy, Luxembourg, and the Slovak 1993-94 and 2013-14, and include up to 29 OECD
Republic, as well as Lithuania (Figure 4.5). At the other countries, Lithuania and the Russian Federation.
end of the scale, fewer than 5% report weekly smoking Data are drawn from school-based samples of 1,500
in Iceland and Norway. Across the OECD, the average is in each age group (11-, 13- and 15-year-olds) in most
12%. On average, boys smoke slightly more than girls, but countries.
girls smoke more than boys in twelve countries (Australia,
the Czech Republic, Denmark, France, Germany, Hungary,
Italy, Luxembourg, the Slovak Republic, Spain, Sweden and
the United Kingdom). Gender gaps are particularly high References
in Israel, as well as Lithuania and the Russian Federation. Chatterji, P. and J. DeSimone (2005), “Adolescent Drinking
Over 30% of 15-year-olds have been drunk at least twice and High School Dropout”, NBER Working Paper,
in the Czech Republic, Denmark, Hungary, Slovenia and No. w11337, Cambridge, United States.
the United Kingdom, as well as Lithuania (Figure 4.6). In Currie, C. et al. (eds.) (2012), “Social Determinants of Health
Iceland, Israel, Luxembourg, Switzerland as well as the and Well-being Among Young People”, Health Behaviour
Russian Federation, rates drop below 15%. Across the OECD, in School-aged Children (HBSC) Study: International
the average is 22.3%, with a small gap between boys (23.5%) Report from the 2009/2010 Survey, WHO Regional Office
and girls (21.2%). Gender disparities, with boys more prone for Europe, Copenhagen.
to drink than girls, are especially high in Austria, Hungary,
Forster, J. et al. (2007), “Policy Interventions and Surveillance
Israel, as well as Lithuania and the Russian Federation
as Strategies to Prevent Tobacco Use in Adolescents and
(over 5 points). Only in Canada, Sweden and the United
Young Adults”, American Journal of Preventive Medicine,
Kingdom do girls report repeated drunkenness more often
Vol. 33, No. 6 (Suppl.), pp. S335-S339.
than boys.
Inchley, J. et al. (eds.) (2016), “Growing Up Unequal: Gender
Trends for repeated drunkenness and regular smoking
and Socioeconomic Differences in Young People’s
in 15-year-olds display similar patterns (Figure 4.7). Both
Health and Well-being”, Health Behaviour in School-
health behaviours are now at their lowest since 1993-94.
aged Children (HBSC) Study: International Report from
Regular smoking displays the strongest decrease, as rates
the 2013/2014 Survey, WHO Regional Office for Europe,
in boys and girls more than halved between 1997-98 and
Copenhagen.
2013-14. The gender gap for drunkenness has also shrunk
since the 1990s. All countries present a decrease in regular O’Cathail, S.M. et al. (2011), “Association of Cigarette
smoking since 1997-98, exceeding 60% for both boys and Smoking with Drug Use and Risk Taking Behaviour
girls in Belgium, Canada, Denmark, Ireland, Norway, Sweden in Irish Teenagers”, Addictive Behaviors, Vol. 36, No. 5,
and the United Kingdom, and for girls in Austria, Finland pp. 547-550.
and Switzerland. The decreases are weaker for repeated WHO (2013), Report on the Global Tobacco Epidemic, WHO,
drunkenness, and reach 60% only for boys in Ireland and Geneva.
74 Health at a Glance 2017 © OECD 2017
4. RISK FACTORS FOR HEALTH
Smoking and alcohol consumption among children
4.5. Smoking among 15-year-olds, 2013-14
Smoking at least once a week
% Total Boys Girls
25
20
15
10 20.5 21.0
19.0
17.0
15.0 15.5 16.0
13.5 13.5 13.5 14.0 14.0 14.5 14.5
11.7 12.0 13.0
5 9.0 10.0 10.8 11.0 11.0
7.5 8.0 8.5
5.5 6.5 11.5
5.0
3.0 4.0
0
i a Slo ic
de nia
No d
Ca y
Au da
S w li a
D e de n
i t e Ir e k
K i nd
i t z ain
B e nd
t h um
Po d s
l
OE el
ia
Fr c
Re n d
Ge ion
Gr ia
ce
m d
Sl L i t u r g
Re ni a
Hu e
y
ly
Es 8
ga
an
i
a
c
ar
ar
xe an
an
bl
bl
ra
r
tv
It a
do
rw
an
ee
ra
na
n
la
la
st
CD
e
Sw p
to
r tu
a
t
nm
N e l gi
ng
bo
rm
pu
pu
e
e c F inl
Lu Pol
el
la
La
Is
ra
v
ov hu
S
er
st
ng
Au
Ic
er
Fe
h
ak
d
n
Cz
Un
ss
Ru
Source: Inchley et al. (2016); Cancer Council Victoria (2016) for Australia.
12 http://dx.doi.org/10.1787/888933602785
4.6. Drunkenness among 15-year-olds, 2013-14
Drunk at least twice in life
% Total Boys Girls
45
40
35
30
25
20 37.0 37.5 38.5
15 27.5 28.0 29.5 30.0
30.5 30.5 30.5
23.5 24.5 26.0
10 19.0 20.8 21.0 21.5 22.0 22.3
16.0 16.5 16.5 16.5 16.5 16.5
13.5 14.0 14.5
5
6.0 10.0
0
ic
Sl lic
Li nia
ov Po y
Re nd
Fi ia
d
C z K in i a
d
i t z ael
d nd
m n
Ir e g
Sw d
F n
er e
Po s
No l y
Be ay
m
Gr n
Ca e
OE a
G e r ia
Re om
Hu ni a
De ar y
k
Au 7
ga
an
c
ar
nd
xe io
ai
an
an
ur
bl
on
tv
It a
iu
Ne r an
rw
ee
b
ed
na
n erla
ak la
la
st
CD
S w Isr
e
Sp
r tu
ua
Lu er a t
nm
e c gd
ng
bo
rm
pu
pu
nl
el
La
la
lg
ov
ite st
th
Ic
E
th
Fe
h
d
ia
Un
Sl
ss
Ru
Source: Inchley et al. (2016).
12 http://dx.doi.org/10.1787/888933602804
4.7. Trends in regular smoking and repeated drunkenness among 15-year-olds for selected OECD countries,
1994 to 2014
% Smoking - girls Smoking - boys Drunkenness - girls Drunkenness - boys
45
40
35
30
25
20
15
10
5
0
1993-94 1997-98 2001-02 2005-06 2009-10 2013-14
Note: Average for 1993-94 includes 19 countries; average for 1997-98 includes 22 countries; average for 2001-02 includes 25 countries; average for
2005-06 includes 28 countries; averages for 2009-10 and 2013-14 include 27 countries.
Source: WHO (1996); Currie et al. (2000, 2004, 2008, 2012); Inchley et al. (2016).
12 http://dx.doi.org/10.1787/888933602823
Health at a Glance 2017 © OECD 2017 75
4. RISK FACTORS FOR HEALTH
Healthy lifestyles among adults
Low fruit consumption, low vegetable consumption, and Sweden (Figure 4.10). In Portugal, Italy and Spain, fewer
low levels of physical activity are among the ten leading than 60% meet the WHO recommendation. Across the
risk factors in terms of years of healthy life lost in 24, 6, OECD, an average of 66.5% of people perform 150 minutes
and 16 OECD countries respectively (Forouzanfar et al., of moderate physical activity per week, with 70.5% of
2016). Worldwide, diets low in fruit were the cause of nearly men and 63% of women. Men are more physically active
3 million deaths in 2015, while low vegetable consumption than women in all countries but Denmark. The gap is
caused nearly 2 million deaths, and low physical activity particularly high (over 15 points) in the Czech Republic,
caused 1.6 million deaths. Including fruit and vegetables in Latvia, Turkey and Spain.
the daily diet reduces the risk of coronary heart disease,
stroke, as well as certain types of cancer (WHO, 2014).
They include dietary fibre which lowers blood pressure
and regulates insulin, possibly impacting the risk of Definition and comparability
type 2 diabetes (InterAct Consortium, 2015). Regular physical
Fruit and vegetable consumption is defined as the
activity improves muscular and cardiorespiratory fitness,
proportion of individuals consuming at least one fruit
and reduces the risk of hypertension, coronary heart disease,
or vegetable per day. Data rely on self-reporting and
stroke, diabetes, and various cancers (WHO, 2017). It has also
are subject to errors in recall.
been shown to positively impact mental health (Lindwall
et al., 2012). In adults, the WHO recommends at least 150 Data for Australia, Korea and New Zealand are derived
minutes of moderate-intensity physical activity per week, at from quantity-type questions. Data from the United
least 75 minutes of vigorous-intensity physical activity per States include juice made from concentrate. In these
week, or an equivalent combination of the two (WHO, 2017). countries, values may be overestimated as compared
with other countries. Most countries report data for the
Fifty-seven per cent of adults across the OECD consume fruit
population aged 15 +, but there are some exceptions
daily, with values ranging from 30-35% in Finland and Latvia,
as highlighted in the data source of the OECD Health
to over 70% in Australia, Italy, New Zealand and Portugal
Statistics database.
(Figure 4.8). Women consume more fruit than men in all
countries, and display the highest rates of consumption The indicator of moderate physical activity is defined
in Australia, Canada, Italy and New Zealand (over 75%). as doing at least 150 minutes of moderate physical
Meanwhile, they display the lowest rates in Finland, Latvia, activity per week. Estimates of moderate physical
Mexico, the Netherlands and Turkey (under 50%). Levels of activity are based on self-reports from the European
consumption for men are highest in Australia, Canada, Italy, Health Interview Survey 2014, combining work-related
Korea, New Zealand, Portugal and Spain (over 60%), while physical activity with leisure-time physical activity
they are lowest in Finland, and Latvia (below 30%). Gender (bicycling for transportation and sport). Walking for
gaps are largest in Austria, the Czech Republic, Denmark, transportation is not included.
Finland, Germany, Iceland, Latvia, Norway, Slovenia, Sweden
and Switzerland (15-20 points), and lowest in Australia,
Mexico and Turkey (under 5 points). Overall, 63% of women
in the OECD consume fruit daily, while 50% of men do. References
Vegetable consumption is higher than fruit consumption Forouzanfar, M.H. et al. (2016), “Global, Regional, and
(Figure 4.9). On average, 60% of people in the OECD consume National Comparative Risk Assessment of 79 Behavioural,
vegetables daily (65% of women, and 55% of men). Rates are Environmental and Occupational, and Metabolic Risks or
highest in Australia, Korea, New Zealand and the United Clusters of Risks, 1990–2015: A Systematic Analysis for
States, with over 90% of people reporting eating vegetables the Global Burden of Disease Study 2015”, The Lancet,
daily, although the methodology differs across countries Vol. 388, pp. 1659-1724.
(see Definition and comparability). On the other end of
Lindwall, M. et al. (2012), “Self-Reported Physical Activity
the spectrum, fewer than 40% report doing so in Finland,
and Aerobic Fitness are Differently Related to Mental
Germany and the Netherlands. In the United States, men
Health”, Mental Health and Physical Activity, Vol. 5, No. 1,
consume slightly more vegetables than women, and in
pp. 28-34.
Korea and Mexico they consume the same amount; in all
other countries, women consume more vegetables than The InterAct Consortium (2015), “Dietary Fibre and Incidence
men. Gender gaps are large in Austria, Denmark, Finland, of Type 2 Diabetes in Eight European Countries: The
Germany, Luxembourg, Norway, Sweden and Switzerland EPIC-InterAct Study and a Meta-analysis of Prospective
(15-19 points). Studies”, Diabetologia, Vol. 58, pp. 1394–1408.
Over 70% of adults perform at least 150 minutes of WHO (2017), “Fact Sheet on Physical Activity”.
moderate physical activity weekly in Austria, Denmark, WHO (2014), “Increasing Fruit and Vegetable Consumption
Finland, France, Iceland, Ireland, Norway, Slovenia and to Reduce the Risk of Noncommunicable Diseases”.
76 Health at a Glance 2017 © OECD 2017
A
0
10
60
90
20
30
40
50
70
80
0
10
60
90
20
30
40
50
70
80
100
0
10
60
20
30
40
50
70
80
90
100
Sw
concentrate.
concentrate.
Ko Ne ust
ed w r al
en A re
79.5 Ne ust a 99.1
Ze ia
al 95.0
Ic w r al an
el d
an Un Z e a i a 99.0 80.5
d 79.4 ite lan It a
No d Po l y
76.2
rw St d 95.3 r tu
ay a g
Source: Eurostat EHIS 2014.
78.3 Be tes 70.9
92.4
De Ca al
nm lg na
iu da
ar
k C m 78.2 68.8
Sl 77.6 Sw an Sp
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
ov i t z ad a Un ai
Health at a Glance 2017 © OECD 2017
en er 74.7 ite K n 66.7
ia
73.6
la d o
Ki rea
66.4
S w nd
68.5
Au
st e S w ngd
% of population aged 15 years and over
% of population aged 15 years and over
de
% of population aged 15 years and over
r ia Un i t z om
73.3 i t e Ir e n 67.5 er 62.8
Ir e d la l
la K i nd S l a nd
nd ng 67.0 ov 61.5
72.7 do e
Fr Hu nia
an Gr m 65.5 ng 60.7
ee
Total
ce a
Total
Total
71.2 ce
59.1
Fi 62.2
Sw r y
nl It a ed
an
d Sl l y e
Ge 70.7 ov 61.3 Un P o n 58.6
rm en ite lan
d 58.5
an Tu ia 60.9 St d
y 69.8 rk a
Un Gr OE tes
ite ee Ic e y
el 60.3 CD 57.9
d ce a 3
68.0 OE nd
56.8
60.0
Ki Au 2
ng CD st
do 3 r
m Fr 2 59.8 F r ia 56.2
OE 67.6 an an
Sl CD M ce c
55.1
ov 57.6 Gr e
ak 23 ex
Men
ee
66.5
Men
Men
Re ic ce
pu Po o 57.5 Ir e 55.0
bl la la
ic Po nd
Hu 64.5 r tu 55.7
No nd 55.0
rw
Lu
ng g Be ay
ar
y
No al 55.2 lg 54.0
xe
m 63.4 rw
bo a D e ium
ur Lu E s t y 55.0 nm 53.9
g 62.6
xe o n a
Tu m ia 54.8 Es r k 53.0
Cz bo to
ec r ke u ni
h y 62.5 Au rg 52.2 Lu Ic e a 52.2
Re st
pu xe lan
bl H u r ia
47.5
m d 52.0
ic ng bo
61.8 ar Sl u
Po y 46.3 ov T r g 49.8
Women
Women
Women
la Sp a k ur k
nd
60.8
Sl D ain Re e y
ov en 44.6 47.5
4.8. Daily fruit eating among adults, 2015 (or nearest year)
pu
4.10. Moderate weekly physical activity among adults, 2014
La ak m
tv Re ar k C z Ge blic
ia 44.1 47.4
4.9. Daily vegetable eating among adults, 2015 (or nearest year)
60.7 pu ec rm
h
Es bl
to Cz i R e a ny
ni ec L a c 44.0 pu 47.3
Po a 60.1
h t bl
Re v ia
43.3 M ic 46.8
r tu pu Ne ex
ga b th ic
l 57.1 F i lic 41.4 er o 43.1
nl la
It a Ge an d n
ly 39.7 Fi ds 42.8
47.2 Ne rm nl
an
th an
Sp er y
la 34.1 La d 32.3
ai
n 47.1
nd tv
s 29.4
ia
29.7
77
12 http://dx.doi.org/10.1787/888933602880
12 http://dx.doi.org/10.1787/888933602861
12 http://dx.doi.org/10.1787/888933602842
Healthy lifestyles among adults
4. RISK FACTORS FOR HEALTH
Note: Data for Australia, Korea and New Zealand are derived from quantity-type questions. Data for the United States include juice made from
Note: Data for Australia, Korea and New Zealand are derived from quantity-type questions. Data for the United States include juice made from
4. RISK FACTORS FOR HEALTH
Healthy lifestyles among children
Consuming a healthy diet and performing regular physical habits have evolved, due to new leisure patterns (TV,
activity when young can be habit forming, promoting a internet, smartphones) which have led to a decrease in
healthy lifestyle in adult life. Daily consumption of fruit physical activity (Inchley et al., 2016). Age-specific policies
and vegetables can help reduce the risk of coronary heart should promote a decrease in screen time and an increase
diseases, strokes, and certain types of cancer (Hartley et al., in physical activity levels. Furthermore, the gender gap
2013; World Cancer Research Fund, 2007). The most common between boys and girls has not decreased with time,
guideline recommends consuming at least five portions of suggesting that girls should be targeted with gender-
fruit and vegetables daily. Moderate-to-vigorous physical sensitive approaches and interventions.
activity is beneficial to adolescents’ physical, mental and
psycho-social health, as it helps build and maintain healthy
bones and muscles, reduces feelings of depression and
anxiety, and improves academic achievement (Janssen and Definition and comparability
LeBlanc, 2010; Singh et al., 2012). The WHO recommends
60 minutes of moderate-to-vigorous daily physical activity Dietary habits are measured here in terms of the
for those aged 5-17 years. proportions of children who report eating fruit and
vegetables at least every day or more than once a
Over 40% of 15-year-olds consume fruit daily in Canada,
day, no matter the quantity. No reference to exclude
Denmark, Iceland and Switzerland, while less than 25%
juice, soup or potatoes was mentioned in the survey
do so in Finland, Greece, Latvia and Sweden (Figure 4.11).
questions. In addition to fruit and vegetables, healthy
Rates exceed 50% for girls in Denmark and Switzerland,
nutrition also involves other types of foods.
while only boys in Canada reach 40%. Rates are under 30%
for girls in Greece, Hungary, Latvia, Poland and Sweden, and Data for physical activity consider the regularity of
under 20% for boys in Finland, Latvia, and Sweden. Across self-reported moderate-to-vigorous physical activity
the OECD, nearly one in three 15-year-olds consumes fruit lasting at least 60 minutes. Moderate-to-vigorous
daily, with girls at 37% and boys at 28%. Girls consume more physical activity refers to exercise undertaken for at
fruit than boys in all countries. Gender gaps are largest in least an hour each day which increases the heart rate,
Denmark, Finland and Switzerland (17-18 points). and sometimes leaves the child out of breath.
Daily vegetable consumption in 15-year-olds exceeds 50%
in Belgium and 40% in Canada, Denmark, Ireland, Israel,
the Netherlands and Switzerland (Figure 4.12). Rates
are under 25% in the Czech Republic, Estonia, Germany,
References
Portugal, the Slovak Republic, Slovenia and Spain. Overall, Hartley, L. et al. (2013), “Increased Consumption of Fruit and
the OECD average is 32%, nearly identical to the average Vegetables for the Primary Prevention of Cardiovascular
for fruit consumption. Rates are highest in girls in Belgium Diseases”, Cochrane Database of Systematic Reviews, Vol. 4,
(over 60%), and Israel and Switzerland (over 50%); they are No. 6, CD009874.
highest for boys in Belgium (over 50%) and Ireland (over
Inchley, J. et al. (eds.) (2016), “Growing Up Unequal: Gender
40%). Daily vegetable consumption is lowest for girls in
and Socioeconomic Differences in Young People’s
Estonia, Portugal and Spain, and boys in Finland, Germany
Health and Well-being”, Health Behaviour in School-
and Spain. In all countries, girls consume more vegetables
aged Children (HBSC) Study: International Report from
than boys. Gender gaps are largest in Finland, Germany,
the 2013/2014 Survey, WHO Regional Office for Europe,
Italy and Switzerland (15 points or over).
Copenhagen.
Rates of physical activity meeting the WHO guidelines reach
Janssen, I. and A.G. LeBlanc (2010), “Systematic Review of
20% in Canada and Spain, and are lower than 10% in Israel,
the Health Benefits of Physical Activity and Fitness in
Italy and Switzerland (Figure 4.13). They are consistently
School-Aged Children and Youth”, International Journal
higher in boys, and by a large margin, as gender gaps range
of Behavioral Nutrition and Physical Activity, Vol. 7, No. 40.
from 5 points (Israel, Sweden and Switzerland) to 17 points
(Luxembourg). Physical activity is lowest in girls in Austria, OECD (2017), “Obesity Update”, OECD Publishing, Paris, http://
Israel, Italy and Portugal (5%), and boys in France, Israel, Italy www.oecd.org/els/health-systems/Obesity-Update-2017.pdf.
and Switzerland (under 15%). Sufficient physical activity Singh, A. et al. (2012), “Physical Activity and Performance at
is most prevalent in girls in Canada, Iceland and Latvia School: A Systematic Review of the Literature Including
(14-15%), and boys in Canada and Spain (nearly 30%). The Methodological Quality Assessment”, Archives of Pediatrics
OECD average is just under 15%, with nearly 20% for boys and Adolescent Medicine, Vol. 166, No. 1, pp. 49-55.
and 10% for girls, resulting in a 10 point average gender gap. World Cancer Research Fund, American Institute for Cancer
Nearly all OECD countries promote fruit and vegetable Research (2007), “Food, Nutrition, and Physical Activity,
consumption: most widely known is the “5 a day” and the Prevention of Cancer: A Global Perspective”,
guideline (OECD, 2017). In recent years, children’s daily AICR, Washington DC.
78 Health at a Glance 2017 © OECD 2017
0
5
10
15
20
25
30
35
%
0
10
20
30
40
50
60
70
60
%
0
10
20
30
40
50
%
Ca Be Ca
na lg n
iu
da m D e ad a
Sp 22.0 56.3 45.0
S w Isr S w nma
ai
Sl
ov Ic n 20.0
i t z ael
45.0
it z rk
42.5
el er er
ak a la la
Re n d n n
pu 19.5 Ca d 44.5 Ic d 42.5
Ne na el
Source: Inchley et al. (2016).
Source: Inchley et al. (2016).
Source: Inchley et al. (2016).
bl an
i th da d
Po c 19.0 er 44.0 40.5
la la Is
ra
nd
nd e
Fi 18.0 Ir e s 43.5 Ir e l 40.0
nl la la
a
Health at a Glance 2017 © OECD 2017
De nd nd
17.5 43.0
Hu n d
Lu ng nm 37.0
xe ar a It a
m y 17.5 Sw rk 41.0 No l y 36.5
bo ed rw
ur en
g 17.5
Un
38.5
Po ay
La i t e Fr a r tu 36.0
Li via t d n ga
th 17.5
K i ce
ng 38.0 A l 35.0
ua Ru do Lu u s
Total
Total
Total
ni ss m xe t r i
m a
Ir e a 17.5 ia I 34.3 bo 34.5
Ne la n c ela
Fe n u
17.0 34.0
Cz the nd de d Be rg 34.5
Ru e c h r l a n ra lg
ti iu
ss Re d s m
17.0 33.5
OE on
ia Fr 33.5
n pu CD an
F e bl 2
16.5 32.4
de ic No 7 OE ce
33.0
ra rw CD
tio ay Un G e 2 7
No n 16.0 32.0 ite rm 32.3
rw L u Gr e
xe e d
OE ay 15.5 m ce 30.5
K i a ny
ng 32.0
CD bo Ru do
ur m
Sl 2 7 14.6
g 30.5
ss S
ia lo 31.5
ov n v
Boys
Boys
Boys
en It a F e en
Li l
Es ia 14.0 th y 29.0 C z de ia
ec ra 31.0
to ua
h t
n ni
Un B ia 13.5 Po a 29.0
Re ion
31.0
ite elg pu
d i la
n bl
K i um Sl ic
ng 12.8 Au d 27.0 ov 30.0
do st ak Spa
Ge m r Re in
rm 12.7 F i ia 26.0 pu 30.0
nl
a an bl
12.5 25.5
Hu d
S w ny
ng Es ic 28.5
ed to
e ar
y
ni
Au n 12.5 Cz
ec L a 25.5 Po a 28.5
st h tv la
Li nd
4.11. Daily fruit eating among 15-year-olds, 2013-14
P o r ia Sl Re ia t
Girls
Girls
Girls
r tu 11.5 ov 25.0 27.5
ak pub Ne hua
4.12. Daily vegetable eating among 15-year-olds, 2013-14
De gal Re lic th nia
nm 11.5 24.5 er 27.5
pu la
ar b n
11.5
Sl lic 24.0
Hu d s
Gr k ov ng 27.0
ee en
c Ge ia ar
F e 11.0 rm 24.0 Gr y 25.5
Sw r an a ee
ce
it z ce Po ny
10.0 23.5 24.5
4.13. Moderate-to-vigorous daily physical activity among 15-year-olds, 2013-14
er r tu La
la g tv
nd
9.5 Es al 23.0 Sw ia 23.0
It a to ed
ly ni
a e
Is 8.0 22.5 Fi n 21.5
Sp nl
ra ai an
el
7.5
n 19.5
d
21.0
79
12 http://dx.doi.org/10.1787/888933602937
12 http://dx.doi.org/10.1787/888933602918
12 http://dx.doi.org/10.1787/888933602899
Healthy lifestyles among children
4. RISK FACTORS FOR HEALTH
4. RISK FACTORS FOR HEALTH
Overweight and obesity among adults
Overweight and obesity are major risk factors for many would result in proportional reductions in consumption,
chronic diseases, including diabetes, cardiovascular especially if fixed at 20% of the retail price or more (WHO,
diseases, and cancer. High body mass index (BMI) led to 2016). Comprehensive policy packages that include health
nearly 4 million deaths in 2015, a 19.5% increase since 2005 promotion, education, interventions in primary care
worldwide. It is the leading risk factor in terms of healthy settings, and broader regulatory and fiscal policies, provide
years of life lost in Turkey, second leading in six other affordable and cost-effective solutions to tackle obesity
OECD countries, and third leading in another 24 member (OECD, 2010).
countries (Forouzanfar et al., 2016). Obesity has risen
quickly in the OECD in recent decades, and projections
show that this trend will continue (OECD, 2017). It has
affected all population groups, regardless of gender, age, Definition and comparability
race, income or education level, though to varying degrees
(Sassi, 2010). Overweight and obesity are defined as excessive
Across the OECD, 54% of the population is overweight, weight presenting health risks because of the high
including 19% who are obese (Figure 4.14). Total overweight proportion of body fat. The most frequently used
(BMI≥25) ranges from 24% in Japan and 33% in Korea to just measure is based on the body mass index (BMI), which
over 70% in Mexico and the United States. Obesity (BMI≥30) is a single number that evaluates an individual’s weight
is lowest in Italy, Japan and Korea (under 10%), and highest in relation to height (weight/height2, with weight in
in Hungary, Mexico, New Zealand and the United States kilograms and height in metres). Based on the WHO
(30% or over). In most countries, pre-obesity (25≤BMI<30) classification, adults over age 18 with a BMI greater
accounts for the largest share of overweight people. than or equal to 25 are defined as overweight, and
those with a BMI greater than or equal to 30 as obese.
On average, 20% of women and 19% of men are obese Pre-obesity defines people whose BMI is greater than
(Figure 4.15). Gender gaps are lower than 1 point in or equal to 25 and below 30. Most countries report
Canada, France, Germany, Iceland, the Slovak Republic, data for the population aged 15 +, but there are some
Spain, Sweden and the United Kingdom. Women are more exceptions as highlighted in the data source of the
obese than men in a majority of countries, with disparities OECD Health Statistics database.
10 points and over in Mexico, Turkey, as well as Colombia,
and 22 points in South Africa. In the countries where men Overweight and obesity rates can be assessed
are more obese than women (Australia, the Czech Republic, through self-reported estimates of height and weight
Japan, Korea, Ireland and Slovenia), the gender gaps are derived from population-based health interview
much lower. surveys, or measured estimates derived from health
examinations. Estimates from health examinations
Obesity has greatly risen in the past two decades, even are generally higher and more reliable than from
in countries where rates have been historically low health interviews.
(Figure 4.16). Obesity has more than doubled since the late
1990s in Korea and Norway. Rates seem to have stabilised
in recent years in Italy and Japan. OECD countries with
historically high rates of obesity are Canada, Chile, Mexico,
the United Kingdom and the United States. These countries
References
have also shown a great increase since the 1990s: +92% in Forouzanfar, M.H. et al. (2016), “Global, Regional, and
the United Kingdom, and +65% in the United States. The National Comparative Risk Assessment of 79 Behavioural,
increase has been slower in Canada, and Mexico since 2006, Environmental and Occupational, and Metabolic Risks or
and the rise in Chile is nearly imperceptible. Clusters of Risks, 1990–2015: A Systematic Analysis for
OECD countries have increased implementation of a range the Global Burden of Disease Study 2015”, The Lancet,
of public health policies to try to slow the obesity epidemic Vol. 388, pp. 1659-1724.
(OECD, 2017). Food labelling measures, such as nutrient lists, Goryakin, Y. et al. ( forthcoming), “The Role of Communication
informative logos, or traffic light schemes have been set in Public Health Policies. The Case of Obesity Prevention
up in Australia, England, France and New Zealand, among in Italy”, OECD Health Working Paper, OECD Publishing,
other countries. Social media and new technologies have Paris.
become tools for public health promotion, through mass
OECD (2017), “Obesity Update”, OECD Publishing, Paris, http://
media campaigns aiming to increase public awareness
www.oecd.org/els/health-systems/Obesity-Update-2017.pdf.
about healthier choices (Goryakin et al., forthcoming).
Taxation policies have also been increasingly implemented Sassi, F. (2010), Obesity and the Economics of Prevention:
to raise the price of potentially unhealthy products such as Fit not Fat, OECD Publishing, Paris, http://dx.doi.
foods high in salt, fat, or sugar. Taxes on sugar-sweetened org/10.1787/9789264084865-en.
beverages are amongst the most popular, and there is WHO (2016), “Fiscal Policies for Diet and Prevention of
reasonable evidence that appropriately designed taxes Noncommunicable Diseases”, WHO, Geneva.
80 Health at a Glance 2017 © OECD 2017
0
5
10
60
15
20
25
30
35
40
45
0
10
20
30
40
50
70
80
0
5
10
15
20
25
30
35
40
%
Ja Ja
pa pa
1990
Ko n n 3.7 20.1
re S w Ko
a it z rea
In In d
do ia er 5.3 28.1
ne la
nd
s 10.3 30.8
Ch i a It a
1995
in
Sw a No l y 9.8 35.3
Italy*
rw
i t z It al a
er y 12.0 34.0
Norway*
l Au y
N o a nd s
D e t r ia
rw 14.7 32.4
Men (measured)
S w ay Ne nm
2000
ed th ar k
14.9 33.3
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
er
Obesity (measured)
la
Health at a Glance 2017 © OECD 2017
Ne A u en n
% of population aged 15 years and over
% of population aged 15 years and over
th str
e r ia Sw ds 12.8 34.7
l ed
De and e
nm s Fr n 12.3 36.0
Korea
an
Japan
2005
Po ar k Be ce
la lg 17.0 32.0
n iu
Is d
r Es m 18.6 32.4
Gr a e l to
Sl ee ni
ov a 18.0 33.3
ak S c e Sp
2010
Re pai Po ain
pu n r tu 16.7 35.1
b ga
Fr lic
Is l 16.6 36.4
Sl anc ra
o e
16.6 36.5
Men (self-reported)
C z P ven Po el
2015
Switzerland*
e c or ia la
h tu n
Obesity (self-reported)
Sl O
Re ga ov E C d 16.7 36.6
pu l ak D
Note: Data in countries with a * were self-reported rather than measured.
Re 3 4
Es blic pu 19.4 34.5
to bl
Ic ni a Cz i
e ec L a c 16.3 37.9
Li lan
th d h t
u Re v i a
23.2 31.4
0
5
10
15
20
25
30
35
40
%
Be ania pu
lg bl
i i 36.7
18.7
1990
Lu O E C um Tu c
xe D rk
Sl ey
22.3 33.1
Ru m 34 ov
ss bo
ia u en
n Ir e r g
F e la Gr i a 19.2 36.4
de nd ee
r ce
17.0 39.2
1995
Ge atio Lu Ic e l
rm n xe an
m d
Women (measured)
an bo 19.0 38.6
Overweight (measured)
Br y Ge urg
rm 22.6 35.5
United States
C a a z il
C o n ad a
United Kingdom
2000
Un lo a
m
C a ny 23.6 36.4
na
i t e F bia d
d inl Ir e a 25.8 34.4
K i an l
4.15. Obesity among adults by gender, 2015 (or nearest year)
ng d Un H n d a
do i t e un 23.0 37.0
L m d ga
2005
Ki r y
Au at v ng 30.0 32.3
4.14. Overweight including obesity among adults, 2015 (or nearest year)
st ia
ra d
Chile
A u om
Tu lia st 26.9 36.1
Canada
r ra
N e Hun ke y li a
N e F in 27.9 35.5
4.16. Evolution of obesity in selected OECD countries, 1990 to 2015 (or nearest year)
w ga
w lan
2010
So Zea r y
ut la Un Z e a d 24.8 40.2
h nd ite lan
Af d 31.6 35.2
Un M r i c a St d
Women (self-reported)
ite ex at
Overweight (self-reported)
d ico
St M es 38.2 31.9
ex
2015
at ic
Mexico
es o 33.3 39.2
81
12 http://dx.doi.org/10.1787/888933602994
12 http://dx.doi.org/10.1787/888933602975
12 http://dx.doi.org/10.1787/888933602956
Overweight and obesity among adults
4. RISK FACTORS FOR HEALTH
4. RISK FACTORS FOR HEALTH
Overweight and obesity among children
Childhood obesity has become one of the most serious communities, families and individuals is necessary to
public health challenges of the 21st century. Obesity can effectively halt the epidemic and decrease prevalence
affect a child’s physical health, through cardiovascular, (Inchley et al., 2016).
endocrine, or pulmonary diseases, and psycho-social
health, through the development of poor self-esteem,
eating disorders, and depression (Inchley et al., 2016).
Obesity can also affect educational attainment (Cohen Definition and comparability
et al., 2013). Furthermore, childhood obesity is a strong
predictor of adult obesity, which has health and economic Estimates of overweight and obesity are based on body
consequences (WHO, 2016). mass index (BMI) calculations using either measured
or child self-reported height and weight, the latter
Overweight (including obesity) based on measured rather
possibly under-estimating obesity and overweight.
than self-reported height and weight ranges from 15% in
Overweight and obese children are those whose BMI
Norway to 45% in Chile (Figure 4.17). Across the OECD, the
is above a set of age- and sex-specific cut-off points
average is 25%, with 26% of overweight boys, and 24% of
(Cole et al., 2000).
overweight girls, although rates are based on different age
groups. Prevalence of overweight is higher in girls than in Measured data are gathered by the World Obesity
boys in Ireland, Mexico, New Zealand, Portugal, Sweden, Federation (WOF, former IASO) from different national
Switzerland, Turkey and the United Kingdom (England), as studies. The estimates are based on national surveys
well as South Africa. Gender gaps are largest in Denmark, of measured height and weight among children at
Greece, Korea, Poland, Sweden, as well as South Africa various ages. Caution is therefore needed in comparing
(larger than 8 points). rates across countries. Definitions of overweight and
obesity among children may sometimes vary among
Over 20% of 15-year-olds self-report overweight in Canada,
countries, although whenever possible the IOTF BMI
Greece and the United States, while prevalence drops under
cut-off points are used.
10% in Denmark (Figure 4.18). The highest rates occur for
girls in Canada, Greece, Iceland and the United States Self-reported data are from the Health Behaviour in
(15% or over), and in boys in Canada, Greece, Israel, Italy, School-aged Children (HBSC) surveys undertaken
Slovenia and the United States (over 20%). Rates are lowest between 2001-02 and 2013-14. Data are drawn from
in girls in Poland and Norway, as well as Lithuania and school-based samples of 1 500 in each age group (11-,
the Russian Federation (6-7%), and in boys in Denmark, 13- and 15-year-olds) in most countries. Self-reported
the Netherlands, France as well as Lithuania (10-14%). height and weight are subject to under-reporting,
Self-reported overweight is higher in boys than in girls in missing data and error, and require cautious
all countries, and the overall OECD average is 16% (19% in interpretation.
boys, 12% in girls). Gender gaps are large overall, but are
highest in Canada, Estonia, Greece, Italy, Norway, Poland
and the Russian Federation (10-15 points). The gaps remain
very small in Denmark, the Netherlands, and Portugal
References
(1-3 points). Cohen, A.K. et al. (2013), “Educational Attainment and
Self-reported overweight in 15-year-olds has increased Obesity: A Systematic Review”, Obesity Reviews, Vol. 14,
in most OECD countries in the past decade (Figure 4.19). No. 12, pp. 989-1005.
Overall across the OECD, overweight increased by 28%, from Cole, T.J. et al. (2000), “Establishing a Standard Definition for
12% in 2001-02 to 16% in 2013-14. The strongest increases Child Overweight and Obesity Worldwide: International
occurred in the Czech Republic, Israel, Latvia, Poland, Survey”, British Medical Journal, Vol. 320, pp. 1-6.
the Slovak Republic and Sweden, where overweight rose
Inchley, J. et al. (eds.) (2016), “Growing Up Unequal: Gender
by more than 50%, as well as Estonia and Lithuania and
and Socioeconomic Differences in Young People’s
the Russian Federation, where they more than doubled.
Health and Well-being”, Health Behaviour in School-
Overweight has dropped since 2001-02 in Denmark, as well
aged Children (HBSC) Study: International Report from
as for boys in Iceland and Spain, and girls in Norway and
the 2013/2014 Survey, WHO Regional Office for Europe,
the United Kingdom (England).
Copenhagen.
Increasingly obesogenic environments have contributed to
OECD (2017), “Obesity Update”, OECD Publishing, Paris, http://
the rise in overweight and obesity in children. Several OECD
www.oecd.org/els/health-systems/Obesity-Update-2017.pdf.
countries have implemented policies aimed at tightening
regulation of advertisements of unhealthy foods and Veugelers, P. and A. Fitzgerald (2005), “Effectiveness of
beverages, specifically targeted toward children and young School Programs in Preventing Childhood Obesity: A
adults to prevent obesity (OECD, 2017). Children have been Multilevel Comparison”, American Journal of Public Health,
found to respond well to school programmes (Veugelers and Vol. 95, No. 3, pp. 432-435.
Fitzgerald, 2005), but a systemic approach encompassing a WHO (2016), “Report of the Commission on Ending
broad spectrum of factors leading to obesity and including Childhood Obesity”, WHO, Geneva.
82 Health at a Glance 2017 © OECD 2017
In
do
0
5
10
15
20
25
30
35
%
0
5
10
15
20
25
30
35
40
%
De De ne
nm nm
0
5
10
15
20
25
30
35
40
45
50
%
ar ar No si a
Li Li k r w (6-
th k 9.5 th 9.5 a 1
S w ua ua
ni Po y (1 2) 9.5
i t z nia
er 10.5 Fr a 10.5 la 0 -
la Sw an B nd 12)
n it z c Ne elgi (13 14.5
Fr d 12.0 er e 12.0 th um -18
an la
Source: Inchley et al. (2016).
c nd er 14.9
la (10 )
Po e 12.0 Au 12.0 nd -1
la st 2
n r ia Tu s (1 ) 15.5
No d 12.5 12.5 rk 0-1
rw Ne L at ey
for Health and Welfare for Finland.
Health at a Glance 2017 © OECD 2017
2)
a th via S (1 16.0
Au y 12.5 er
la 12.5 Sw we 1-1
st n i t z de n 8 )
r ia 16.0
N La 12.5 No d s 12.5 er
la (7-
rw nd 9 )
Sl eth t v i a
o e Sl ay (6 17.1
12.5 12.5
Source: Currie et al. (2004); Inchley et al. (2016).
Ru v a r l a
Total
Ru o v a P o l Ire -13
ss k R nd ss k R and
ia s ia 17.5
n epu 12.5 n epu 12.5 Ch l a n d )
F e bl F e bl in (
d i Li a ( 7 ) 17.9
Total
de ic
2001-02
ra 13.0 UK er a c 13.0 So thu 13-
tio ( E tio ut an 15)
h i 18.7
Es n 13.0 ng n 13.0 Af a (6
to la
n ric -9
B ni B e d) a 18.9
UK e l g a 14.0 lg 14.0 Au (2 )
( i iu st -
m r 14
20.0
C z En um Cz
ec gla 14.0 ec Est 14.0 U K K ia ( )
o 2
h h on (E rea -15
R n d)
14.5
Re ia
14.0
ng ) 21.2
Lu epu Lu pu la (10
xe bli xe bli nd -1
m c 15.0 m c 15.0
) ( 9)
21.5
bo bo 14
ur ur -1
g 15.0 15.0
A u O E 5)
Ir e g st C 23.5
It a la r a D2
nd l
Boys
Ir e l y 15.5 15.5 Ca ia ( 4 24.6
la
Boys
It a na 7-1
Po n d Po l y d a 7)
rt 15.5 15.5 Sl 25.0
u r tu ov ( 5
OE gal OE gal De eni -17
CD 15.5 15.5 nm a ( ) 25.0
Note: The numbers in parentheses refer to the age of the children surveyed in each country.
CD
Hu 2 8 Ge 28 ar 7-1
ng 15.7 rm 15.6 k ( 8)
25.1
Ge ar y a Fi 14-1
n
Hu ny
rm 16.0 ng 16.0 Hu la 6 )
25.2
an ng n d
ar ar (1
y 16.0 y 16.0 y 5
Sp Sp Po (10 ) 25.5
ai ai r tu -12
16.5
Sw n 16.5 Sw n
2013-14
ed ed Br g a l ) 25.5
en e
Girls
a z (1
Is 16.5 Fi n 16.5 il ( 0 )
r
nl
an 1 28.5
Sl ael
ov 17.0 d 17.0
Gr 1-1
ee 4)
Is c 28.8
en r Sp e (
Sl ael
F i ia
nl 17.0 ov 17.0 ai 15
29.4
Girls
an en Ne Mex n (8 )
Ic d 17.0 Ic ia 17.0
w ic -17
o
Note: International Obesity Task Force cut-offs. Rates for the United States refer to survey year 2009-10 rather than 2013-14.
el el
an an Un Zea (1 ) 30.9
ite lan 2-1
18.0 18.0
Rates for the first data point for Iceland, Luxembourg and the Slovak Republic refer to survey year 2005-06 rather than 2001-02.
d d 9
4.18. Self-reported overweight (including obesity) among 15-year-olds, 2013-14
Gr d Gr d
ee
c
ee
c
St (1 ) 35.0
at 0-1
Un C a e 21.5 Un C a e 21.5 es 4
i t e n ad i t e n ad (1 ) 36.2
d d 3
St a 24.5 St a 24.5 Ch -17
at
es
at
es il e ) 39.9
30.5 30.5 (1
3)
44.5
4.17. Measured overweight (including obesity) among children at various ages, 2010 (or nearest year)
4.19. Change in self-reported overweight (including obesity) among 15-year-olds, 2001-02 and 2013-14
83
12 http://dx.doi.org/10.1787/888933603051
12 http://dx.doi.org/10.1787/888933603032
12 http://dx.doi.org/10.1787/888933603013
Source: International Association for the Study of Obesity (2013); World Obesity Federation (2016, 2017); JUNAEB (2016) for Chile; THL National Institute
Overweight and obesity among children
4. RISK FACTORS FOR HEALTH
Note: International Obesity Task Force cut-offs. Rates for the second data point for the United States refer to survey year 2009-10 rather than 2013-14.
4. RISK FACTORS FOR HEALTH
Air pollution
Air pollution is a major environment-related health threat, 1% of global GDP (OECD, 2016). Policies to limit air pollution
especially to children and the elderly, as it can cause consist of regulatory approaches, such as air quality
respiratory diseases, lung cancer, and cardiovascular standards, fuel quality standards or emission ceilings, as
diseases. It has also been linked to low birth-weight, well as economic instruments, which include fuel taxes,
dementia, and damage to DNA and the immune system road pricing or taxes on emissions.
(WHO, 2017). Outdoor air pollution in both cities and
rural areas was estimated to cause 3 million premature
deaths worldwide in 2012 (WHO, 2016), and can also have
substantial economic and social consequences, from Definition and comparability
health costs to building restoration needs and agricultural
output (OECD, 2015). Of particular concern for outdoor air The WHO has established guidelines for air pollution,
pollution are carbon monoxide, nitrogen oxide and ozone, expressed as the average level of exposure of a nation’s
but also fine particulates, or PM 2.5, whose diameter is population (urban and rural) to concentrations of
2.5 μm or smaller. These are potentially more dangerous suspended particles which must not be exceeded.
than the larger particulates (PM10), as they can penetrate The indicators presented here reflect the estimated
deeper into the respiratory tract, and cause severe health average level of exposure to concentrations of fine
effects. In 2015, particulate matter pollution was the particulates, which measure less than 2.5 microns
cause of over 4.2 million deaths worldwide (Forouzanfar in diameter. The WHO guidelines for PM2.5 are an
et al., 2016). The WHO has claimed that air pollution is annual mean of 10 microgrammes/m3, which is the
one of the most pernicious threats facing global public lower range over which adverse health effects have
health today and on a bigger scale than HIV or Ebola been observed.
(WHO, 2017). Data for PM2.5 are made available by the World Bank,
In 2015, exposure levels to PM 2.5 exceeding the WHO through the Global Burden of Disease Study. They
guidelines were higher than 90% in 21 OECD countries are generated by combining data from different
(Figure 4.20). In 19 of those countries, 100% of the population sources, including satellite observations of aerosols
was exposed. Australia, Canada, Finland, Iceland, New in the atmosphere and round-level monitoring of
Zealand and Sweden display rates of nearly 0%, followed particulates. However, pollutant concentrations
by the United States and Norway with rates below 10%. The are sensitive to local conditions, and measurement
OECD average is 68%. protocols may differ across countries. The data must
therefore serve as a general indicator of air quality,
The mean annual population exposure to PM2.5 has decreased
mostly allowing for cross-country comparison.
in the OECD, on average, from 18.2 microgrammes/m3 in
1990 to 15.1 microgrammes/m3 in 2015 (Figure 4.21). While
the overall trend since 1990 has been downward, there have
been some increases in population exposure in more recent
References
years. This is largely due to the concentration of pollution
sources in urban areas and to increasing use of private Forouzanfar, M.H. et al. (2016), “Global, Regional, and
vehicles for urban trips (OECD, 2015). National Comparative Risk Assessment of 79 Behavioural,
In 2015, population exposure was lowest in Australia, Environmental and Occupational, and Metabolic Risks or
Canada, Finland, Iceland, New Zealand and Sweden, and Clusters of Risks, 1990–2015: A Systematic Analysis for
highest in Korea and Turkey, as well as China, India and the Global Burden of Disease Study 2015”, The Lancet,
South Africa. Population exposure has decreased in most Vol. 388, pp. 1659-1724.
countries since 1990, except in Israel, Italy, Japan, Korea, OECD (2015), Environment at a Glance 2015: OECD Indicators,
Turkey, China and India where increases range from 5% OECD Publishing, Paris, http://dx.doi.org/10.1787/
in Japan to 24% in India. In countries where exposure 9789264235199-en
has dropped, the decreases range from 3-8% in Finland,
OECD (2016), The Economic Consequences of Outdoor Air
Iceland, Spain, Sweden, Switzerland as well as Costa Rica
Pollution, OECD Publishing, Paris, http://dx.doi.org/10.1787/
and Indonesia, to 30-40% in the Czech Republic, Hungary,
9789264257474-en.
Latvia, Poland, the Slovak Republic as well as Lithuania.
WHO (2017), “Healthier, Fairer, Safer: The Global Health
The WHO estimates that overall, 92% of the world’s
Journey”, WHO, Geneva.
population is breathing air above the PM 2.5 guidelines
(WHO, 2017), and indoor and outdoor air pollution cause WHO (2016), “Ambient (Outdoor) Air Quality and Health Fact
approximately 7 million premature deaths per year (WHO, Sheet”, WHO, Geneva.
2014). OECD projections estimate that outdoor air pollution WHO (2014), “7 Million Premature Deaths Annually Linked
will cause 6 to 9 million premature deaths by 2060, and cost to Air Pollution”, press release, WHO, Geneva.
84 Health at a Glance 2017 © OECD 2017
Ne
w
0
10
20
30
40
50
60
70
80
90
100
%
N e Ic
0
10
20
30
40
50
60
70
80
Ze w ela
Au alan Ze nd
st d a 0
5.5 Sw land
S w r a li a e 0
e 5.9 C a de n
C a de n na 0
na 6.2 F i da
F i da n 0
n 7.2
Microgrammes/m 3
Un A u lan
Un I lan i t e s tr d 0
ite c el d
d an 7.4 d a li
0
St a
St d
7.8 No a t es
No a t es r 9
r 8.4 Es way
Es way t 9
to
n 9.1 Po on
15
Health at a Glance 2017 © OECD 2017
r tu ia
S ia 9.4 g
Po pai Sp al 24
r tu n 9.7 Ir e a i n
Ir e g a l la 28
De lan 9.8 n
nm d B d 31
a 9.8 OE r a z
il
B rk D CD 56
Un 11.0 Ru
68
ite F r a z s s I enm 3 5
d r a il 11.4 ia nd a
K n
1990
S w ing c e Un n F e on r k 79
i t z dom 12.4 i t e de esi
er d ra a 89
la 12.4 K i tio
ng n
J a nd d 90
p 12.9 F om
G an Co r an 90
Ge reec 13.3 lo ce
Ne rm e m 92
th a 13.5 b
Source: World Bank (2017), World Development Indicators (database).
Source: World Bank (2017), World Development Indicators (database).
e r ny
l
C o Ja ia 96
14.0 S o s t a pan
98
OE and
u t Ri
In C D s 14.6 h ca
do 3 5 A 100
Ru L B n e s 15.1 Ge fric
rm a
s s ux e l g i a 100
i a em iu 15.4 an
n C y 100
F e bo m
de ur 15.6 M hil e
ra g ex 100
t 16.6 ic
Au ion In o 100
Co st 16.6 A u di a
l o r ia Cz s 100
Li mb 17.0 e c B e t r ia
th ia h l gi 100
ua
n 18.0 Re um
pu 100
Co I ia 19.1 b
Gr l i c
st t al 100
aR y 19.9 Hu e e c
M ic a ng e 100
e 20.1 a
Sl x i c
ov o Is r y
ra 100
Sl
ov e 20.2 e
100
2015
ak L a nia It a l
Re t v 20.3
pu ia Ko l y 100
bl 20.4 Lu L r e a
i xe a t 100
Cz Ch c 20.5
4.21. Mean annual population exposure to PM2.5, 1990 and 2015
ec i Ne mb v i a
th ou 100
h Isr l e
Re a 20.9 er rg
p el Sl la
n 100
21.1 ov
ak Po d s 100
Hu ubli
ng c 21.4 Re lan
p d 100
4.20. Population exposed to PM2.5 levels exceeding 10 microgrammes/m3, 2015
Po ar y
l 23.1 S ub
So K and S w lov lic 100
u t or 24.3 i t z en
er ia
h ea la 100
Af
r 28.7 Tu nd
Tu ica Li rk 100
rk 29.6 th ey
e ua 100
Ch y 36.4 n
in Ch i a 100
In a 58.4 in
a 100
di
a 74.3
85
12 http://dx.doi.org/10.1787/888933603089
12 http://dx.doi.org/10.1787/888933603070
Air pollution
4. RISK FACTORS FOR HEALTH
5. ACCESS TO CARE
Population coverage for health care
Unmet needs for health care due to cost
Out-of-pocket medical expenditure
Geographic distribution of doctors
Waiting times for elective surgery
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.
Health at a Glance 2017 © OECD 2017 87
5. ACCESS TO CARE
Population coverage for health care
Health care coverage, through government schemes and access or larger choice to providers (duplicate insurance).
private health insurance, provides financial security against Among OECD countries, nine have private coverage for
unexpected or serious illness. However, the percentage of over half of the population (Figure 5.2).
the population covered by such schemes does not provide Private health insurance offers 96% of the French
a complete indicator of accessibility, since the range of population complementary insurance to cover cost-sharing
services covered and the degree of cost-sharing applied to in the social security system. The Netherlands has the
those services also affects access to care. largest supplementary market (84% of the population),
Most OECD countries have achieved universal (or near- followed by Israel (83%), whereby private insurance pays
universal) coverage of health care costs for a core set of for prescription drugs and dental care that are not publicly
services, which usually include consultations with doctors reimbursed. Duplicate markets, providing faster private-
and specialists, tests and examinations, and surgical and sector access to medical services where there are waiting
therapeutic procedures (Figure 5.1). Generally, dental care times in public systems, are largest in Ireland (45%) and
and pharmaceutical drugs are partially covered, although Australia (56%).
there are a number of countries where these services must The population covered by private health insurance has
be purchased separately (OECD, 2015). Universal coverage increased in some OECD countries over the past decade,
has typically been achieved through government schemes particularly in Denmark, Korea, Slovenia and Belgium.
(national health systems or social health insurance), But private health insurance coverage has come down in
though a few countries (the Netherlands and Switzerland) other countries, notably Greece, Ireland, New Zealand and
use compulsory private health insurance to cover some or the United States (Figure 5.3). The importance of private
all of the population. health insurance is linked to several factors, including
Population coverage for a core set of services is below gaps in access to publicly financed services, government
95% in seven OECD countries, and lowest in Greece, the interventions directed at private health insurance markets,
United States and Poland. In Greece, the economic crisis and historical development.
continues to have a significant effect, reducing health
insurance coverage among the long-term unemployed.
Many self-employed workers have also decided not
to renew their health insurance because of reduced
Definition and comparability
disposable income. However, since 2014 uninsured people Coverage for health care is defined here as the share
are covered for prescribed pharmaceuticals, emergency of the population receiving a core set of health care
services in public hospitals, and for non-emergency goods and services under public programmes and
hospital care under certain conditions (Eurofound, 2014). through private health insurance. It includes those
Further, since 2016 new legislation has sought to close covered in their own name and their dependents.
remaining coverage gaps. In the United States, coverage is Public coverage refers to national health systems or
provided mainly through private health insurance. Publicly social health insurance. Take-up of private health
financed coverage covers the elderly, and people with low insurance is often voluntary, although it may be
income or with disabilities. The share of the population mandatory by law or compulsory for employees
uninsured decreased from 14.4% in 2013 to 9.1% in 2015. as part of their working conditions. Premiums are
This followed implementation of the Affordable Care Act, generally not income-related, although the purchase
which was designed to expand health insurance coverage of private coverage can be subsidised by government.
(Cohen and Martinez, 2015). However, this Act is under
review by the current United States administration.
In Poland, a tightening of the law in 2012 made people
lose their social health insurance coverage if they fail to References
pay their contribution. But uninsured people who need
Burke, S. et al. (2016), “From Universal Health Insurance
medical care utilise emergency hospital services, where
to Universal Healthcare? The Shifting Health Policy
they will be encouraged to obtain insurance. In Ireland,
Landscape in Ireland since the Economic Crisis”, Health
though coverage is universal, most of the population have
Policy, Vol. 120, No. 3, pp. 235-240.
to pay not insignificant user charges (upwards of EUR50)
to access primary care (Burke et al., 2016). Cohen, R.A. and M.E. Martinez (2015), “Health Insurance
Coverage: Early Release of Estimates from the National
Basic primary health coverage, whether provided through
Health Interview Survey, 2014”, National Center for
government schemes or private insurance, generally
Health Statistics, June.
covers a defined “basket” of benefits, in many cases
with cost-sharing. In some countries, additional health Eurofound (2014), Access to Healthcare in Times of Crisis,
coverage can be purchased through voluntary private Dublin.
insurance to cover any cost-sharing left after basic OECD (2015), “Measuring Health Coverage”, OECD, Paris,
coverage (complementary insurance), add additional available at: http://www.oecd.org/els/health-systems/
services (supplementary insurance) or provide faster measuring-health-coverage.htm.
88 Health at a Glance 2017 © OECD 2017
5. ACCESS TO CARE
Population coverage for health care
5.1. Population coverage for a core set of services, 2015 5.2. Private health insurance coverage, by type,
(or nearest year) 2015 (or nearest year)
Total public coverage Primary Complementary
Primary private health coverage Supplementary Duplicate
Australia 100 France 96
Canada 100
Czech Republic 100 Slovenia 71 16
Denmark 100 Netherlands 84
Finland 100
Ireland 100 Israel 83
Israel 100 Belgium 82
Italy 100
Japan 100 Canada 67
Korea 100 Korea 67
New Zealand 100
Norway 100 United States 55 8
Portugal 100 Australia 56
Russian Federation 100
Luxembourg 50
Slovenia 100
Sweden 100 Chile 19 30
Switzerland 100
Ireland 45
United Kingdom 100
Germany 89.2 10.8 Denmark 37
Iceland 99.7
Austria 36
Austria 99.9
France 99.9 Germany 11 23
Netherlands 99.9
New Zealand 29
Spain 99.1 0.8
Belgium 99.0 Switzerland 28
Turkey 98.4 Portugal 26
Colombia 96.6
Luxembourg 95.2 Spain 1 15
Hungary 95.0 Finland 15
Estonia 94.0
Slovak Republic 93.8 Greece 12
Lithuania 92.4 United Kingdom 11
Mexico 92.3
Chile 73.2 18.8 Mexico 8
Poland 91.0 Iceland 0
United States 35.6 55.3
Greece 86.0 Sweden 0
0 20 40 60 80 100 0 20 40 60 80 100
Percentage of total population Percentage of total population
Source: OECD Health Statistics 2017. Note: Private health insurance can be both duplicate and supplementary
12 http://dx.doi.org/10.1787/888933603108 in Australia; both complementary and supplementary in Denmark and
Korea; and duplicate, complementary and supplementary in Israel and
Slovenia.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603127
5.3. Trends in private health insurance coverage, 2005 and 2015 (or nearest year)
2005 2015
Percentage of total population
100 96
87
82 83
80 84
67 67
60 56
63
40 36 37 45
34
25 26
28 29
20 15 16 16
5 8
0 11 12
0 7
o
ia
d
r ia
nd
li a
es
da
m
d
ia
ce
il
nd
el
ce
ga
ric
an
ke
re
ar
az
nd
ic
ai
an
an
an
en
ra
tv
iu
do
at
an
ee
ra
na
la
la
st
Sp
ex
r tu
nm
Ko
Br
r
rm
nl
el
al
La
la
lg
Is
Af
ov
St
Tu
er
Ir e
st
ng
Au
Gr
Fr
Ca
M
Ic
Ze
er
Fi
Be
Po
Au
it z
Ge
De
h
Sl
d
Ki
th
ut
ite
w
Sw
Ne
d
So
Ne
Un
ite
Un
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603146
Health at a Glance 2017 © OECD 2017 89
5. ACCESS TO CARE
Unmet needs for health care due to cost
Access to health care may be prevented for a number of States, where 43% of low income adults reported having
reasons. These can be due to the functioning of the health unmet care needs due to cost in 2016. There were also large
care system (such as the cost of health care, distance to the gaps in unmet care needs between high and low income
closest health care facility, or waiting lists) or to personal people in France and Canada.
reasons (including fear of not being understood by the Self-reported unmet care needs should be assessed together
doctor or not having the time to seek care). People who with other indicators of potential barriers to access, such
forgo health care when they need it may jeopardise their as the extent of health insurance coverage and the amount
health status. of out-of-pocket payments. Strategies to improve access to
Unmet needs due to cost is a particularly pressing problem, care for disadvantaged or underserved populations need
especially among lower-income groups. Consequently, an to tackle both financial and non-financial barriers, as well
increasing number of countries collect data to measure the as promoting an adequate supply and distribution of the
extent to which health care is foregone due to cost (Fujisawa health workforce.
and Klazinga, 2017). This includes whether people skipped
consultations or prescribed medicines due to cost.
On average across OECD countries, just over one in ten
people reported having skipped a consultation due to cost
Definition and comparability
in 2016, based on 17 OECD countries (Figure 5.4). Relatively The OECD collects data on unmet care needs due
high numbers of people reporting to forego consultations to cost reported by populations from national and
is somewhat surprising, as in most OECD countries international sources and a number of countries
consultations are free of charge or with a small co-payment reporting these measures are increasing over time.
(Paris et al., 2016). The share of the population foregoing These use questions that are similar to those asked
consultations due to cost is high in Poland (33%), and also in the Commonwealth Fund International Health
in the United States (22.3%) and Switzerland (20.9%). Less Policy Survey. Rates for Figures 5.4 and 5.5 refer to
than 5% of the population in Germany, Spain, Sweden, both primary and secondary care and they are age-sex
the United Kingdom, Israel and Italy reported skipping standardised to the 2010 OECD population structure,
consultations due to cost. to remove the effect of different population structures
In most countries, the share of the population who skipped across countries. Due to the change of data source for
a consultation due to cost has not changed much in recent this indicator, data cannot be compared directly with
years, but there are some exceptions. A large increase was those presented in the previous editions of Health at
observed in Switzerland, with people who have foregone a Glance.
consultations concentrated among those younger than The 2016 Commonwealth Fund’s International Health
50 years of age and those with low income (OFSP, 2016). Policy Survey asks whether people did not visit a doctor
In Germany and Estonia, the share of the population who when they had a medical problem, skipped a medical
skipped consultation due to cost has decreased. test, treatment, or follow-up that was recommended
In terms of prescribed medicines, on average 7.1% of by a doctor, or did not fill prescription for medicines or
people reported having skipped prescribed medicines skipped doses because of cost in the past year and as
due to cost, based on 15 OECD countries (Figure 5.5). it also collects socio-economic background including
Most OECD countries have co-payments for prescribed income level, it allows analysis on unmet care needs
medicines, though often with exemptions for specific by income group. This survey was carried out in
population groups (Paris et al., 2016). Population shares 11 countries.
reporting foregone prescribed medicines were highest
in the United States (18%) and Switzerland (11.6%); and
lowest in Germany (3.2%) and the United Kingdom (2.3%).
In most countries, the share of the population who skipped References
prescribed medicine due to cost has slightly decreased in
recent years. Large improvements were reported in Israel, Fujisawa, R. and N. Klazinga (2017), “Measuring Patient
Estonia and Australia. In Israel, this may be due in part Experiences (PREMs): Progress Made by the OECD and
to policies to improve accessibility and affordability of its Member Countries 2006-2015”, OECD Health Working
medicines for chronic patients and the elderly. Papers, Paris.
Unmet needs for health care due to cost are consistently OFSP (2016), “Prise en charge médicale : la population suisse
higher among people in low income groups compared est satisfaite”, Communiqué de presse, Berne, https://
with those in high income groups, across OECD countries www.bag.admin.ch/bag/fr/home/aktuell/medienmitteilungen.
(Figure 5.6). An exception is in the United Kingdom, where msg-id-64545.html.
unmet care needs due to cost are similar for low income Paris, V. et al. (2016), “Health Care Coverage in OECD Countries
adults and the rest of the population. Unmet needs are in 2012”, OECD Health Working Papers, No. 88, OECD
particularly large among the low income in the United Publishing, Paris, http://dx.doi.org/10.1787/5jlz3kbf7pzv-en.
90 Health at a Glance 2017 © OECD 2017
5. ACCESS TO CARE
Unmet needs for health care due to cost
5.4. Consultations skipped due to cost, 2016 (or nearest year)
% (Age-sex standardised rates per 100 population)
35
33.0
30
22.3
25
20.9
20
16.2
14.5
15
10.5
10.3
9.7
8.5
8.3
10
6.6
5.9
4.8
4.7
4.2
3.9
5
2.8
2.6
ce
y
l¹
a¹
s¹
17
d¹
n¹
en
e l¹
ly¹
ay
da
nd
es
¹
li a
nd
an
ga
ni
nd
ai
do
an
rw
an
at
ed
na
It a
ra
CD
la
ra
la
rm
Sp
to
r tu
St
er
la
al
ng
Is
Fr
Sw
Ca
No
st
Po
OE
Es
er
Ze
itz
Ge
Po
Au
d
Ki
th
ite
Sw
w
d
Ne
Un
Ne
ite
Un
1. National sources.
Source: Commonwealth Fund International Health Policy Survey 2016 and other national sources.
12 http://dx.doi.org/10.1787/888933603165
5.5. Prescribed medicines skipped due to cost, 2016 (or nearest year)
% (Age-sex standardised rates per 100 population)
20
18.0
15
11.6
10.5
10.1
10
7.8
7.1
6.8
6.7
6.4
5.7
5.4
4.1
4.0
3.6
5
3.2
2.3
0
ce
l¹
da
nd
m
ay
n¹
e l¹
d¹
a¹
s¹
en
15
es
li a
an
ga
ni
nd
do
ai
an
rw
an
at
ed
na
ra
CD
la
ra
rm
Sp
to
r tu
St
er
la
al
ng
Is
Fr
Sw
Ca
No
st
OE
Es
er
Ze
it z
Ge
Po
Au
d
Ki
th
ite
Sw
w
d
Ne
Un
Ne
ite
Un
1. National sources.
Source: Commonwealth Fund International Health Policy Survey 2016 and other national sources.
12 http://dx.doi.org/10.1787/888933603184
5.6. Unmet care needs due to cost, by income level, 2016
% Low income adults All other adults
50
43
45
40
35 30 30 31
28
30 25
24 32
25 20
20 16 16
22
15 18
8
10 13 14 14 13
5 8
7 6 7
0
m
en
ce
ay
li a
10
da
nd
es
an
an
do
rw
an
at
ra
ed
na
CD
la
rm
al
St
er
st
ng
Fr
Sw
Ca
No
Ze
OE
Au
itz
Ge
d
Ki
ite
w
Sw
d
Ne
Un
ite
Un
Note: Either did not consult with/visit a doctor because of the cost, skipped a medical test, treatment, or follow-up that was recommended by a doctor
because of the cost, did not fill/collect a prescription for medicine, or skipped doses of medicine because of the cost. “Low income” is defined as
household income less than 50% of the country median. Sample sizes are small (n < 100) in the Netherlands and the United Kingdom.
Source: Commonwealth Fund International Health Policy Survey 2016.
12 http://dx.doi.org/10.1787/888933603203
Health at a Glance 2017 © OECD 2017 91
5. ACCESS TO CARE
Out-of-pocket medical expenditure
Financial protection through compulsory or voluntary greater share of household spending (20-32%) on inpatient
health coverage can substantially reduce the amount that care than the OECD average of less than 10%.
people need to pay directly for medical care. Yet in some In some Central and Eastern European countries such as
countries the burden of out-of-pocket spending can still Poland, the Czech Republic and Hungary, as well as Canada
create barriers to health care access and use: households and Mexico, expenditure on pharmaceuticals accounts for
that face difficulties paying medical bills may delay or half or more of all out-of-pocket payments. This may be due
even forgo needed health care. On average across OECD not only to co-payments for prescribed pharmaceuticals,
countries, a fifth of all spending on health care comes but also high levels of spending on over-the-counter
directly from patients (see indicator “Financing of health medicines for self-medication. Therapeutic goods, covering
care”). among other items, corrective eye products and hearing
Out-of-pocket payments rely on the ability to pay. If the aids, can also account for a significant proportion of
financing of health care becomes more dependent on out- household spending. In the case of spectacles, compulsory
of-pocket payments, the burden shifts, in theory, towards coverage is often limited to paying a contribution for the
those who use services more, and possibly from high to cost of the lenses, while private households are left to
low-income earners, where health care needs are usually bear the full cost of the frames if they are not covered
higher. In practice, many countries have safety-nets in by complementary private insurance. Overall, therapeutic
place to protect vulnerable groups of the population (such goods account for more than 20% of household spending in
as the poor, the elderly, or people with chronic diseases the Netherlands, the United Kingdom, Slovenia, Germany
or disabilities) from excessive out-of-pocket payments. and the Slovak Republic.
These may be partial or total exemptions or a cap on direct Coverage for dental treatment is typically limited and
payments, either in absolute terms or as a share of income as such dental care plays a significant part in outpatient
(Paris et al., 2016). and overall household spending, accounting for 20% of
The burden of out-of-pocket medical spending (that is, all out-of-pocket expenditure across OECD countries. In
excluding long-term care services) can be measured either Spain, Norway and Estonia, this figure reaches 30% or
as a share of total household income or consumption. The more. This can at least partly be explained by the limited
share of household consumption allocated to medical care compulsory coverage for dental care in these countries
varied considerably across OECD countries in 2015, ranging compared with a more comprehensive coverage for other
from lows of around 1.5% of total household consumption categories of care.
in France, Luxembourg and the United Kingdom, to more
than 5% in Korea and Switzerland (Figure 5.7). On average,
across OECD countries, 3% of household spending goes on
medical goods and services.
Definition and comparability
Health systems in OECD countries differ in the degree of Out-of-pocket payments are expenditures borne
coverage for different health services and goods. In most directly by a patient where neither compulsory nor
countries, a higher proportion of the cost is paid directly voluntary insurance cover the full cost of the health
for pharmaceuticals, dental care and eye care than for good or service. They include cost-sharing and other
hospital care and doctor consultations (Paris et al., 2016). expenditure paid directly by private households and
Taking into account these differences and also the relative should also include estimations of informal payments
importance of these different spending categories, it is not to health care providers. Only expenditure for medical
surprising that there are significant variations between spending (i.e. current health spending less expenditure
OECD countries in the breakdown of the medical costs that for the health part of long-term care) is presented here,
households have to bear themselves. because the capacity of countries to estimate private
In most OECD countries, spending on pharmaceuticals and long-term care expenditure varies widely.
outpatient care (including dental care) are the two main
spending items for out-of-pocket expenditure (Figure 5.8).
These two components typically account for almost four-
References
fifths of all medical spending by households. Co-payments
and additional services can result in a larger proportion Paris, V. et al. (2016), “Health Care Coverage in OECD Countries
of the cost of inpatient care being taken on directly by in 2012”, OECD Health Working Papers, No. 88, OECD
households –Greece, Belgium and the Netherlands report a Publishing, Paris, http://dx.doi.org/10.1787/5jlz3kbf7pzv-en.
92 Health at a Glance 2017 © OECD 2017
5. ACCESS TO CARE
Out-of-pocket medical expenditure
5.7. Out-of-pocket medical spending as a share of final household consumption, 2015 (or nearest year)
%
6
5.3
5.1
5
4.4
4.4
4.1
3.9
3.8
4
3.7
3.6
3.4
3.3
3.2
3.1
3.1
3.1
3.0
3.0
3.0
2.9
3
2.7
2.6
2.6
2.5
2.5
2.5
2.4
2.4
2.3
2.2
2.1
2.0
1.8
2
1.5
1.4
1.4
1
0
o
Ne C a ic
i t e er ia
Re lic
ce
Ko d
Gr a
Hu e c e
L a le
Po t v ia
Sp l
M in
S w el
B e de n
A u ium
Ic li a
d
Fi ly
Au nd
O E r ia
No 3 4
Es ay
nm a
Ja k
Po n
U n Ir n d
N S d
ov er s
e c ep s
Ze a
Sl and
Lu ing y
m m
Fr r g
ga
n
De o ni
Sl eth t a t e
w n ad
re
ar
Cz k R nd
ar
n
ic
pa
an
d n
bl
Un G ven
i
ra
It a
a
xe do
u
rw
an
h ub
Ch
ra
d ma
la
la
ite ela
st
CD
r tu
ex
ng
bo
e
pu
e
nl
el
al
a la
Is
lg
er
t
st
o
it z
K
Sw
Note: This indicator relates to current health spending excluding long-term care (health) expenditure.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603222
5.8. Out-of-pocket medical spending by services and goods, 2015 (or nearest year)
Pharmaceuticals Therapeutic goods¹ Dental
% Outpatient² Inpatient³ Other
100 1 1
3 3 7 7 2 3 4 1 6 7 6 2 5
9 8 12 10 9 8 3 10
16 16 16 5 13 15 16 9
16 14 26 23 20 19
14 27 25 28 32
13 3
80 30 26
30 30 10 28 21 14 30
7 16 21 29
23 40 42 27 42
6 19 51 35
31 55 22
22 56
6 8 27 31
60 25 18
10 28 29 19 21 26
13 12 12 10
34
20 8
19
3 9 8 8 21 20 29
1 10 14 12 11 13
18 11
40 27 20
10 14
1 6 8 15 17 20
64 7 15
59 10
6
48 48 46 15
44 43 42 42 42 41
20 38 37 36 36 36 36 35
33 33 32 31 30 29
28 27 27 27 26 25 25
16
0
Po o
Hu blic
ia
La c
OE ce
nd
Re da
y
Re an
ite Est a
K i ni a
Au and
Fr i a
Sw 31
Sl en
Gr l
ce
Fi ly
d
n
Ir e a
De and
B e ar k
No m
er ay
Ge nd s
Au y
P o r ia
i t z gal
m d
g
Ic m
e
i
an
vi
re
ar
xe a n
ic
ai
an
ur
en
bl
ra
It a
iu
do
an
Ne r w
ee
ra
ed
a
ak ap
la
st
CD
t
Sp
ex
S w r tu
Ko
nm
ng
bo
rm
ec an
pu
pu
Lu e r l
nl
el
la
lg
Is
ov
st
ng
J
M
th
h
d
ov
Cz
Un
Sl
Note: This indicator relates to current health spending excluding long-term care (health) expenditure.
1. Including eye care products, hearing aids, wheelchairs, etc.
2. Includes home care and ancillary services (and dental if not shown separately).
3. Including day care.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603241
Health at a Glance 2017 © OECD 2017 93
5. ACCESS TO CARE
Geographic distribution of doctors
Access to medical care requires an adequate number and Many OECD countries provide different types of financial
proper distribution of doctors in all parts of the country. incentives to attract and retain doctors in underserved
Concentration of doctors in one region and shortages in areas, including one-time subsidies to help them set up
others can lead to inequities in access such as longer their practice and recurrent payments such as income
travel or waiting times. The uneven distribution of doctors guarantees and bonus payments (Ono et al., 2014). A
and the difficulties in recruiting and retaining doctors in number of countries have also introduced measures to
certain regions is an important policy issue in most OECD encourage students from under-served regions to enrol
countries, especially those with remote and sparsely in medical schools. Japan established in 1973 the Jichi
populated areas, and those with deprived rural and urban Medical University specifically to educate physicians
regions. for service in rural communities, which contributed to
The overall number of doctors per capita varies across improving access to care in underserved rural regions
OECD countries from around two per 1 000 population in (Ikegami, 2014).
Turkey, Chile and Korea, to above five per 1 000 population in The effectiveness and cost of different policies to promote
Greece and Austria (see indicators on doctors in Chapter 8). a better distribution of doctors can vary significantly,
Beyond these cross-country differences, the number of with the impact depending on the characteristics of each
doctors per capita also varies widely across regions within health system, the geography of the country, physician
the same country (Figure 5.9). In many countries there is behaviours, and the specific policy and programme design.
a high concentration of physicians in capital cities; this is Policies should be designed with a clear understanding
particularly evident in Austria, the Czech Republic, Greece, of the interests of the target group in order to have any
Mexico, Portugal, the Slovak Republic, and the United States. significant and lasting impact (Ono et al., 2014).
Between regions, the United States shows nearly a five-fold
difference in physician density, while Australia, Belgium
and Korea show only around a 20 percent difference in
physician densities between regions. Definition and comparability
The density of physicians is also consistently greater in Regions are classified in two territorial levels. The
urban regions, reflecting the concentration of specialised higher level (Territorial Level 2) consists of large regions
services such as surgery and physicians’ preferences to corresponding generally to national administrative
practice in urban settings. There are large differences in regions. These broad regions may contain a mix
the density of doctors between predominantly urban and of urban, intermediate and rural areas. The lower
rural regions in Canada, the Slovak Republic and Hungary, level is composed of smaller regions classified as
although the definition of urban and rural regions varies predominantly urban, intermediate or rural regions,
across countries. The distribution of physicians between although there are variations across countries in the
urban and rural regions was more equal in Japan and Korea, classification of these regions. The data on geographic
but there are generally fewer doctors in these two countries distributions are from the OECD Regional Database.
(Figure 5.10).
Doctors may be reluctant to practice in rural regions
due to concerns about their professional life (including
their income, working hours, opportunities for career References
development, isolation from peers) and social amenities
Ikegami, N. (2014), “Factors Determining the Distribution
(such as educational options for their children and
of Physicians in Japan”, Chapter 7 in Universal Health
professional opportunities for their spouse). A range of
Coverage for Inclusive and Sustainable Development: Lessons
policy levers can be used to influence the choice of practice
from Japan, World Bank, Washington, DC, available at:
location of physicians. These include 1) the provision of
http://dx.doi.org/10.1596/978-1-4648-0408-3.
financial incentives for doctors to work in underserved
areas; 2) increasing enrolments in medical education OECD (2016), Health Workforce Policies in OECD Countries:
programmes of students coming from specific social or Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
geographic backgrounds or decentralising the location of http://dx.doi.org/10.1787/9789264239517-en.
medical schools; 3) regulating the choice of practice location Ono, T., M. Schoenstein and J. Buchan (2014), “Geographic
of doctors (for new medical graduates or foreign-trained Imbalances in Doctor Supply and Policy Responses”,
doctors); and 4) re-organising service delivery to improve OECD Health Working Papers, No. 69, OECD Publishing,
the working conditions of doctors in underserved areas. Paris, http://dx.doi.org/10.1787/5jz5sq5ls1wl-en.
94 Health at a Glance 2017 © OECD 2017
5. ACCESS TO CARE
Geographic distribution of doctors
5.9. Physician density, by level 2 regions, 2015 (or nearest year)
Australia
Austria Vienna
Belgium Brussels
Canada
Chile
Czech Republic Prague
Denmark Copenhagen
Estonia
Finland Helsinki
France Mayotte
Germany Hambourg
Greece Athens Region
Hungary
Israel
Italy
Japan
Korea
Latvia
Luxembourg
Mexico Federal District
Netherlands
New Zealand
Norway Oslo Region
Poland
Portugal Lisbon
Slovak Republic Bratislava
Slovenia
Spain
Sweden
Switzerland
Turkey
United Kingdom
United States Massachusetts Washington, DC
China Beijing
Lithuania
Russian Federation St. Petersbourg
0 1 2 3 4 5 6 7 8 9 10
Density per 1 000 population
Source: OECD Statistics Database 2017.
12 http://dx.doi.org/10.1787/888933603260
5.10. Physician density, rural vs urban areas, 2015 (or nearest year)
Predominantly urban Predominantly rural
Density per 1 000 population
7
6.8
6
5.7
5.4
5.1
5.0
5
4.7
4.4
4.4
4.2
4.2
4.2
4.2
3.9
3.9
3.8
3.8
4
3.6
3.2
3.2
2.8
2.7
2.7
2.7
3
2.6
2.4
2.4
2.3
2.3
2.2
2.2
2.0
2.0
2
1.3
1.0
0
da
ic
ic
ce
l
ay
nd
en
16
ia
li a
y
ga
ni
ke
re
ar
pa
an
bl
bl
tv
rw
an
ra
ed
na
CD
la
r tu
to
Ko
ng
r
pu
pu
nl
Ja
La
Tu
er
st
Fr
Sw
Ca
No
Es
OE
Fi
Hu
Po
Re
Re
Au
it z
Sw
h
ak
ec
ov
Cz
Sl
Source: OECD Statistics Database 2017.
12 http://dx.doi.org/10.1787/888933603279
Health at a Glance 2017 © OECD 2017 95
5. ACCESS TO CARE
Waiting times for elective surgery
Long waiting times for health services is an important by far the longest waiting times, with median waiting times
policy issue in many OECD countries (Siciliani et al., 2013), reaching over 350 days in Poland (Figure 5.13).
although less relevant in some (e.g. Belgium, France, Waiting time guarantees have become the most common
Germany, Japan, Korea, Luxembourg, Switzerland, United policy tool to tackle long waiting times in several countries.
States). Long waiting times for elective (non-emergency) This has been the case in Finland, where a National Health
surgery, such as cataract surgery, hip and knee replacement, Care Guarantee was introduced in 2005, leading to a
generates dissatisfaction for patients because the expected reduction in waiting times for elective surgery (Jonsson
benefits of treatments are postponed and the pain and et al., 2013). In England, since April 2010, the NHS Constitution
disability remain. has set out a right to access certain services within specific
Waiting times are the result of a complex interaction between maximum waiting times, or for the NHS to take all reasonable
the demand and supply of health services, with doctors steps to offer a range of alternative providers if this is not
playing a critical role on both sides. The demand for health possible (Smith and Sutton, 2013). Such guarantees are
services and elective surgeries is determined by the health only effective if they are enforced. There are two main
status of the population, progress in medical technologies approaches to enforcement: setting waiting time targets and
(including the simplification of many procedures, such as holding providers accountable for achieving these targets;
cataract surgery), patient preferences, and the burden of or allowing patients to choose alternative health providers
cost-sharing for patients. However, doctors play a crucial (including the private sector) if they have to wait beyond a
role in converting the demand for better health from maximum amount of time (Siciliani et al., 2013).
patients into a demand for medical care. On the supply
side, surgical activity rates are influenced by the availability
of different categories of surgeons, anaesthetists and other Definition and comparability
staff involved in surgical procedures, as well as the supply
of the required medical equipment. There are at least two ways of measuring waiting
The measure reported refers to the waiting time from when times for elective procedures: 1) measuring the
a medical specialist adds a patient to the waiting list for the waiting times for patients treated in a given period;
procedure, to the moment the patient receives treatment. or 2) measuring waiting times for patients still on the
Both mean and median waiting times are presented. Since list at a point in time. The data reported here relate
a number of patients wait for very long times, the median to the first measure (data on the second measure
is consistently and considerably lower than the mean, and are available in the OECD health database). The data
might represent a better measure for the central tendency come from administrative databases rather than
of this indicator. The significant difference between the two surveys.
measures, especially in countries such as Chile, Estonia, Waiting times are reported both in terms of the average
and Poland, highlights the presence of problematic groups and the median. The median is the value which
of patients who wait significantly longer than others to separates a distribution in two equal parts (meaning
receive treatment. that half the patients have longer waiting times and
In 2015, the mean waiting time for cataract surgery was just the other half lower waiting times). Compared with
over 37 days in the Netherlands, but much longer in Estonia the average, the median minimises the influence of
and Poland (Figure 5.11), with average waiting times of 253 outliers (patients with very long waiting times).
and 464 days respectively. Many countries, like the United
Kingdom, Denmark, Spain and Chile have seen waiting
times remain relatively stable over recent years. Others,
References
shown in the trends graph, have had a general decrease in
the past decade, but have increased since 2013. Jonsson, P.M. et al. (2013), “Finland”, Part II, Chapter 7 in Waiting
For hip replacement, the mean waiting time was around Time Policies in the Health Sector: What Works, OECD Publishing,
42 days in the Netherlands, but 289 days in Estonia and Paris, http://dx.doi.org/10.1787/9789264179080-en.
over 400 days in Chile and Poland (Figure 5.12). The median Siciliani, L., M. Borowitz and V. Moran (2013), Waiting Time
waiting times were around 41 days in Denmark, 49 days Policies in the Health Sector: What Works?, OECD Publishing,
in Italy and 54 days in Israel. It reached between 100 and Paris, http://dx.doi.org/10.1787/9789264179080-en.
150 days in Spain, Norway, Portugal and Australia, and over Smith, P. and M. Sutton (2013), “United Kingdom”, Part II,
200 days in Estonia, Poland and Chile. Chapter 16 in Waiting Time Policies in the Health Sector:
Waiting times for knee replacement follows the patterns What Works, OECD Publishing, Paris, http://dx.doi.
of hip replacement surgery, with Estonia and Poland having org/10.1787/9789264179080-en.
96 Health at a Glance 2017 © OECD 2017
Ne Ne Ne
th th th
0
100
200
300
400
500
600
Days
0
100
200
300
400
500
Days
0
100
200
300
400
500
Days
er er er
la la la
nd
n.a.
D nd n.a.
nd
s n.a.
De s
42 42 37
nm N e enm s
w Un It a
k
ar
49 Ze ar k 41 ite a yC l 24
N 60 al 55 d na 50
an
d K d
Un ew Z I t a l
42 76 58
75 n.a.
N e ing a
75
ite ea y w do
d l It a
K i and 78
ly 49 Ze m 59
ng 78 Un
Is
ra 79 a 72
do
54
De lan d
74
83
i te Ca el
Ca m d na 91
nm 73
na 104 Ki d
ng a 92 Hu ar k 64
109 87
Median
Median
Median
F i da n.a. ng
nl n.a. do
79 43
an Fi m A u ar y
d 110 97 st 88
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
nl ra
126
Health at a Glance 2017 © OECD 2017
Is A u an d 93
ra 94 li a
No el 77
st
r 108 Ch n.a.
rw 129 Po a li a 114 F i il e 89
OE ay 158
r tu n.a. nl 103
CD 175 No gal 116
Po an d 99
r 135 r tu 103
Au 15 143 ga
st
ra 182 Hu w ay 125 89
ng 145 Sp l 104
li a 188 ar
Sp n.a. 70 No ain 83
a Sp y 146 rw 105
Po in 137
r tu 204 OE ain 130 OE ay 95
CD 150 CD 117
Hu gal 200 16
ng 208 Es 16 115 97
159 Is 128
Mean
a to
Mean
Mean
124 n r
Es r y
226 Po ia 202 Es ael 82
to 290 to 132
n la n
286 nd 243 96
393
Po ia Po ia
la Ch 405 la 253
nd 357 nd
il e 237 414
541 410 464
0
50
100
150
200
250
Days
0
100
200
300
400
500
Days
0
50
100
150
200
250
Days
20 20 20
05 05 05
20 20 20
06 06 06
20 20 20
07 07 07
20 20 20
Australia
Denmark
08
Finland
08 08
Estonia
New Zealand
New Zealand
20 20 20
09 09 09
20 20 20
10 10 10
20 20 20
11 11 11
5.11. Cataract surgery waiting times, averages and selected trends, 2015
5.12. Hip replacement waiting times, averages and selected trends, 2015
5.13. Knee replacement waiting times, averages and selected trends, 2015
20 20 20
12 12 12
20 20 20
13 13 13
Finland
Canada
20 20 20
14 14 14
Portugal
Netherlands
20 20 20
United Kingdom
United Kingdom
15 15 15
97
12 http://dx.doi.org/10.1787/888933603336
12 http://dx.doi.org/10.1787/888933603317
12 http://dx.doi.org/10.1787/888933603298
Waiting times for elective surgery
5. ACCESS TO CARE
6. QUALITY AND OUTCOMES OF CARE
Patient experiences with ambulatory care
Prescribing in primary care
Avoidable hospital admissions
Diabetes care
Mortality following ischaemic stroke
Mortality following acute myocardial infarction (AMI)
Hospital mortality rates
Waiting times for hip fracture surgery
Surgical complications
Obstetric trauma
Care for people with mental health disorders
Screening, survival & mortality for breast cancer
Survival & mortality for colorectal cancer
Survival & mortality for leukemia in children
Vaccinations
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.
Health at a Glance 2017 © OECD 2017 99
6. QUALITY AND OUTCOMES OF CARE
Patient experience with ambulatory care
Delivering health care that is responsive and patient- Japan also has a low rate for patient’s perception on time
centred is playing a greater role in health care policy across spent with doctor, which can be inferred at least partly
OECD countries. Considering the health care user as a direct by a high number of consultations both per population
source of information is becoming more prevalent. Since and doctor (see indicator “Consultations with doctors” in
the mid-1990s, there have been efforts to institutionalise Chapter 9).
measurement and monitoring of patient experiences. In several countries, the proportion of patients with
This empowers patients and the public, involves them positive experience has decreased in recent years. For
in decisions on health care delivery and governance, and example, in Poland the share of patients reporting that a
provides insight to the extent to which they are health- doctor spent enough time with them during consultation
literate and have control over the treatment they receive fell between 2010 and 2016. However, some countries such
In many countries, responsible organisations have been as Australia and Estonia have improved some aspects of
established or existing institutions have been identified patient experiences recently.
for measuring and reporting patient experiences. They
develop survey instruments for regular collection of patient
experience data and standardise procedures for analysis
and reporting. An increasing number of countries collect
Definition and comparability
not only Patient-Reported Experience Measures (PREMs) In order to measure and monitor general patient
but also Patient-Reported Outcome Measures (PROMs) experience in the health system, the OECD
which collect patients’ perception on their specific medical recommends collecting data on patient experience
conditions and general health, including mobility, pain/ with any doctor in ambulatory settings. An increasing
discomfort and anxiety/depression, before and after a number of countries have been collecting patient
specific medical intervention such as cancer and hip experience data based on this recommendation
and knee replacement. Given the importance of utilising through nationally representative population surveys
people’s voice for developing health systems and improving while Japan and Portugal collect them through
quality of care, international efforts to develop and monitor nationally-representative service user surveys. About
patient-reported measures has been intensified in recent half of the countries presented, however, collect
years (OECD, 2017a; OECD, 2017b). data on patient experience with a regular doctor. In
Countries use patient-reported data differently to drive 11 countries, the Commonwealth Fund’s International
quality improvements in health systems. To promote Health Policy Surveys 2010 and 2016 were used as a data
quality of health care through increased provider source, even though there are critiques relating to the
accountability and transparency, many countries report sample size and response rates. Data from this survey
patient experience data in periodic national health system refer to patient experience with a regular doctor
reports or on public websites, showing differences across rather than any doctor. In 2016, the Netherlands
providers, regions and over time. Norway uses patient which participates in this survey developed a national
experience measures in payment mechanisms or for fund population survey and this resulted in improved
allocations to promote quality improvement and patient- response rates and data quality.
centred care, and Australia, Canada, the Czech Republic, Rates are age-sex standardised to the 2010 OECD
Denmark, France and the United Kingdom use them to population, to remove the effect of different population
inform health care regulators for inspection, regulation structures across countries.
and/or accreditation. Patient-reported measures are also
used in some Canadian jurisdictions, Denmark, France and
the Netherlands to provide specific feedback for provider’s
quality improvement (Fujisawa and Klazinga, 2017). References
Patients generally report positive experiences when it
Fujisawa, R. and N. Klazinga (2017), “Measuring Patient
comes to communication and autonomy in the ambulatory
Experiences (PREMs): Progress Made by the OECD and
health care system. Across countries, the majority of
its Member Countries 2006-2015”, OECD Health Working
patients report that they spent enough time with a doctor
Papers, Paris.
during consultation (Figure 6.1), a doctor provided easy-to-
understand explanations (Figure 6.2), as well as involved OECD (2017a), “Ministerial Statement: The Next
them in care and treatment decisions (Figure 6.3). For Generation of Health Reforms”, OECD Health Ministerial
all three aspects of patient experience, Belgium and Meeting, Paris, http://www.oecd.org/health/ministerial/
Luxembourg score high at above 95% of patients with ministerial-statement-2017.pdf.
positive experiences while Poland has lower rates and OECD (2017b), “Patient-Reported Indicators Survey (PaRIS)”,
for instance, only one in two patients report having been OECD Publishing, Paris, http://www.oecd.org/health/paris.
involved in their care and treatment during consultation. htm.
100 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Patient experience with ambulatory care
6.1. Doctor spending enough time with patient in consultation, 2010 and 2016 (or nearest year)
Confidence Interval 2016 2010 2016
Age-sex standardised rate per 100 patients
94.1
91.7
100
86.7
97.5
97.2
78.8
95.5
89.6
88.3
86.6
85.4
86.0
84.7
84.0
75
81.3
80.8
79.3
77.9
72.9
59.6
50
39.0
25
¹ ²,
²
n¹
a¹
18
e l¹
y²
d²
l¹
a¹
g¹
ic¹
¹
s²
m
m
nd
en
es
li a
ay
da
ce
ga
re
ni
nd
an
pa
nd
ur
an
bl
ra
CD
iu
do
rw
at
an
ra
ed
na
la
to
r tu
Ko
bo
pu
rm
Ja
la
lg
Is
la
al
St
er
st
ng
OE
Fr
Sw
Ca
Es
No
Po
er
Ze
Be
m
Po
Re
Au
it z
Ge
d
Ki
th
xe
ite
w
Sw
h
Ne
d
Lu
Ne
ec
Un
ite
Cz
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas. Un
1. National sources.
2. Data refer to patient experiences with regular doctor.
Source: Commonwealth Fund International Health Policy Survey 2016 and other national sources.
12 http://dx.doi.org/10.1787/888933603355
6.2. Doctor providing easy-to-understand explanations, 2010 and 2016 (or nearest year)
Confidence Interval 2016 2010 2016
Age-sex standardised rate per 100 patients
100
87.1
85.8
97.8
97.4
96.3
96.2
96.9
93.4
93.5
90.7
90.6
89.4
89.5
89.0
88.9
89.1
69.5
85.9
75
83.7
82.9
50
25
¹ ²,
²
²
n¹
²
a¹
ic¹
l¹
g¹
y²
a¹
17
d²
e l¹
¹
es
s²
m
nd
li a
en
ay
da
ga
ni
nd
re
an
pa
do
nd
ur
an
at
bl
ra
CD
iu
rw
ra
ed
na
la
to
r tu
Ko
bo
St
pu
rm
Ja
la
lg
ng
la
al
Is
er
st
OE
Sw
Ca
Es
No
Po
er
Ze
Be
m
Po
Re
d
Ki
Au
it z
Ge
ite
th
xe
w
Sw
d
Ne
Un
Lu
ite
Ne
ec
Un
Cz
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas.
1 National sources.
2. Data refer to patient experiences with regular doctor.
Source: Commonwealth Fund International Health Policy Survey 2016 and other national sources.
12 http://dx.doi.org/10.1787/888933603374
6.3. Doctor involving patient in decisions about care and treatment, 2010 and 2016 (or nearest year)
Confidence Interval 2016 2010 2016
Age-sex standardised rate per 100 patients
100
87.9
81.8
79.5
79.0
95.6
95.1
90.9
88.9
87.6
86.5
84.3
84.6
87.8
75
83.1
87.1
82.2
81.8
79.0
50
47.9
25
²
²
a¹
ic¹
a¹
e l¹
16
y²
d²
²
²
l¹
g¹
s²
m
nd
nd
ay
en
da
es
li a
ga
ni
re
an
nd
ur
an
bl
ra
CD
iu
do
rw
at
ra
ed
na
la
la
to
r tu
Ko
bo
pu
rm
lg
Is
la
al
St
Po
er
st
ng
OE
Sw
Ca
Es
No
er
Ze
Be
m
Po
Re
Au
it z
Ge
d
Ki
th
xe
ite
w
Sw
h
Ne
Lu
Ne
ec
Un
ite
Cz
Un
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas.
1. National sources.
2. Data refer to patient experiences with regular doctor.
Source: Commonwealth Fund International Health Policy Survey 2016 and other national sources.
12 http://dx.doi.org/10.1787/888933603393
Health at a Glance 2017 © OECD 2017 101
6. QUALITY AND OUTCOMES OF CARE
Prescribing in primary care
Prescribing can be used as an indicator of health care quality chronic benzodiazepine users (>365 defined daily doses
supplementing consumption and expenditure information in one year), and 64 per 1 000 older adults have received
(see Chapter 10). Antibiotics, for example, should be at least one prescription for a long-acting benzodiazepine
prescribed only where there is an evidence based need to or related drugs within the year. The large variation can
reduce the risk of resistant strains. Likewise, quinolones and be explained by different reimbursement and prescribing
cephalosporins are considered second-line antibiotics in policies for benzodiazepines as well as differences in
most prescribing guidelines. They should generally be used disease prevalence and treatment guidelines.
only when first line antibiotics are ineffective. Total volume
of antibiotics prescribed, and second-line antibiotics as a
proportion of total volume have been validated as markers Definition and comparability
of quality in the primary care setting.
Figure 6.4 shows volume of all antibiotics prescribed in Defined daily dose (DDD) is the assumed average
primary care in 2015, with volumes of second-line antibiotics maintenance dose per day for a drug used for its
embedded within the total amount. Total volumes vary more main indication in adults. DDDs are assigned to each
than three-fold across countries, with the Netherlands, active ingredient(s) in a given therapeutic class by
Estonia and Sweden reporting the lowest volumes, and international expert consensus. For instance, the DDD
Greece and France reporting volumes much higher than for oral aspirin equals 3 grams, which is the assumed
the OECD average. Volumes of second-line antibiotics maintenance daily dose to treat pain in adults. DDDs
vary almost 16-fold across countries. The Scandinavian do not necessarily reflect the average daily dose
countries and the Netherlands report the lowest volumes actually used in a given country. For more detail, see
of second line antibiotics, whereas Korea, Italy and Turkey http://www.whocc.no/atcddd.
reported the highest. Variation is likely to be explained, on Data for Spain, Estonia, United Kingdom, Portugal
the supply side, by differences in the regulation, guidelines and Sweden include data for primary care physicians
and incentives that govern primary care prescribers and, only. Data for Canada, Ireland, Slovenia and New
on the demand side, by cultural differences in attitudes Zealand include only those dispensed by community
and expectations regarding the natural history and optimal pharmacies. Data for Finland, Italy, and Korea include
treatment of infective illness. There has been some growth outpatients only. Data for Belgium, Denmark and
in the overall volume of antibiotics between 2010 and 2015. the Netherlands include outpatients and nursing
The highest growth was seen in Ireland and Poland and the homes. Data for Turkey include primary care, nursing
largest decline in Sweden and Iceland. and residential facilities. Data for Australia include
Antibiotic consumption is consistently higher among prescriptions dispensed at community pharmacies,
children and young adults and older adults. Volumes of private hospital pharmacies and public hospital
antibiotics dispensed to children aged 0-9 years varies outpatients and admitted day patients. Results for
by 15-fold across countries but only 5-fold across young Canada only include data from the provinces of British
adults aged 10-19 years of age (Figure 6.5). Consumption Columbia, Manitoba and Saskatchewan.
data subdivided by age groups can allow identification of Denominators comprise the population held in the
specific age groups that are prescribed high proportion of national prescribing database, rather than the general
certain antibiotics and provide detailed information for population.
campaigns or interventions aimed at more prudent use of
antibiotics in these sub-groups of population.
Benzodiazepines are often prescribed for older adults for
anxiety and sleep disorders, despite the risk of adverse References
side effects such as fatigue, dizziness and confusion.
Cecchini, M. (2016), “Tackling Antimicrobial Resistance”,
Long-term use of benzodiazepines can lead to adverse
on OECD Insights blog, June, see http://oe.cd/1JI.
events (falls, road accidents and overdose), tolerance,
dependence and dose escalation. Beside the period of OECD (2017), Tackling Wasteful Spending on Health, OECD
use, there is concern about the type of benzodiazepine Publishing, Paris, http://dx.doi.org/10.1787/9789264266414-en.
prescribed, with long-acting types not recommended OECD (2015), “Antimicrobial Resistance in G7 Countries”,
for older adults because they take longer for the body OECD Policy Brief, October, see www.oecd.org/els/
to eliminate. Figures 6.6 and 6.7 indicate that, across the health-systems/Antimicrobial-Resistance-in-G7-Countries-
OECD, on average around 25 per 1 000 older adults are and-Beyond-Policy-Brief.pdf.
102 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Prescribing in primary care
6.4. Overall volume of antibiotics prescribed, 2015 (or nearest year)
2015 - All 2015 - 2nd line
DDDs per 1 000 population, per day
40
35
30
25
20
15
10
5
0
No ia
Ic lic
ic
m d
s
a
en
Au a
G e r ia
S l ny
De ay
Li ar k
Hu ia¹
Fi y
d
Re y
K i n d¹
OE m
Po da
l
el
A u ain
li a
Re e a
w and
Be y
m
Gr e
¹
Ca 0
ga
ce
ni
ke
c
l
nd
ar
n
an
an
ur
3
en
bl
ra
tv
It a
iu
do
rw
an
b
ak or
ra
ed
na
Lu ola
st
CD
ee
to
Sp
r tu
nm
ng
ite ela
ec Tur
bo
rm
pu
pu
ua
nl
Ne el
al
La
la
Is
lg
ov
st
K
ng
Fr
Sw
Es
r
Ze
er
P
th
I
th
xe
h
Ne
ov
Cz
Un
Sl
1. Data refer to all sectors (not only primary care).
Source: European Centre for Disease Prevention and Control and OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603412
6.5. Volume of antibiotics prescribed in young people, 2015 (or nearest year)
0-9 years of age 10-19 years of age
DDDs per 1 000 population (0-9 years, 10-19 years), per day
30
25
20
15
10
0
ia
ly
li a
el
en
ay
da
ke
ar
nd
an
en
ra
It a
iu
rw
ra
ed
na
nm
r
nl
la
Is
lg
ov
Tu
st
Sw
Ca
No
er
Fi
Be
Au
De
Sl
th
Ne
Source: European Centre for Disease Prevention and Control and OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603431
6.6. Chronic Benzodiazepine Use: Number of 6.7. Long-Acting Benzodiazepine use: Number
patients per 1000, aged 65 years and over who have of patients per 1000, aged 65 years and over who
prescriptions for benzodiazepines for more than have at least one prescription for long-acting
365 days, 2015 (or nearest year) benzodiazepines, 2015 (or nearest year)
Per 1 000 persons aged 65 years and over Per 1 000 persons aged 65 years and over
80 250
70 67
200 192
60 54 52
150
50 44 150
40 34 113
102 95
30 27 25 100 85 82
18 18 18 18 17 64 57 55
20 36 31
11 10 50 24 20 18
10 5 3 13 9
0 1
0 0
Ne epu a
er c
ak Sp a
Po blic
Po an d
Is l
No el
S w ay
F i en
OE an d
De nd s
Ne Ca r k
Ze a
Ko d
Es e a
Au onia
Tu ia
Ne ust y
er ia
nm s
Is k
ey
Es rea
Sl o ni a
Re ain
Ir e a l
OE and
Ne S w el
Ze en
Ca nd
F i da
Tu d
ey
ov Slo 16
No 17
ga
R ni
th bli
A wa
w n ad
ar
De and
an
an
en
th r al
ra
a
g
a
rk
rk
rw
ra
ed
w ed
na
CD
CD
a
r
ak ve
r tu
r tu
nm
Ko
pu
nl
l
nl
l
al
al
la
r
ov
Ir e
t
t
st
l
ov
Sl
Sl
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603450 12 http://dx.doi.org/10.1787/888933603469
Health at a Glance 2017 © OECD 2017 103
6. QUALITY AND OUTCOMES OF CARE
Avoidable hospital admissions
Most health systems have developed a ‘primary level’ in recent years, whereas in Belgium rates have remained
of care whose functions include health promotion and relatively stable and in Spain rates have increased. While
disease prevention, managing new health complaints, observed improvements may represent advances in the
managing long-term conditions and referring patients to quality of primary care for these countries, recent reviews
hospital-based services when appropriate. A key aim is to undertaken by OECD indicate that investment in primary
keep people well, by providing a consistent point of care care may not be happening fast enough (OECD, 2017b),
over the longer-term, tailoring and co-ordinating care for potentially resulting in wasteful spending on health care
those with multiple health care needs and supporting the (OECD, 2017a)
patient in self-education and self-management.
Asthma, chronic obstructive pulmonary disease (COPD)
and congestive heart failure (CHF) are three widely Definition and comparability
prevalent long-term conditions. Both asthma and COPD
limit the ability to breathe: asthma symptoms are usually The indicators are defined as the number of
intermittent and reversible with treatment, whilst COPD is hospital admissions with a primary diagnosis of
a progressive disease that almost exclusively affects current asthma, COPD or CHF among people aged 15 years
or prior smokers. Asthma may affect up to 334 million and over per 100 000 population. Rates are age-sex
people worldwide (Global Asthma Network, 2014). About standardised to the 2010 OECD population aged 15
3 million people died of COPD in 2015, which is equal to and over. Admissions resulting from a transfer from
5% of all deaths globally that year (WHO, 2016). CHF is a another hospital and where the patient dies during
serious medical condition in which the heart is unable to the admission are excluded from the calculation
pump enough blood to meet the body’s needs. CHF is often as these admissions are considered unlikely to be
caused by hypertension, diabetes or coronary heart disease. avoidable.
Heart failure is estimated to affect over 26 million people Disease prevalence and availability of hospital care
worldwide resulting in more than 1 million hospitalisations may explain some, not all, variations in cross-country
annually in both the United States and Europe (Ponikowski rates. Differences in coding practices among countries
et al., 2014). may also affect the comparability of data. For example,
Common to all three conditions is the fact that the the exclusion of “transfers” cannot be fully complied
evidence base for effective treatment is well established with by some countries. Differences in data coverage
and much of it can be delivered at a primary care level. of the national hospital sector across countries may
A high-performing primary care system, where accessible also influence indicator rates.
and high quality services are provided, can reduce acute
deterioration in people living with asthma, COPD or CHF
and reduce unnecessary admissions to hospital. References
Figure 6.8 shows hospital admission rates for asthma
Global Asthma Network (2014), The Global Asthma
and COPD together, given the physiological relationship
Report 2014, Auckland, New Zealand, access at http://
between the two conditions. Admission rates for asthma
www.globalasthmareport.org/resources/Global_Asthma_
vary 15-fold across countries with Italy, Mexico and
Report_2014.pdf.
Colombia reporting the lowest rates and Latvia, Turkey,
and Korea reporting rates over twice the OECD average. OECD (2017a), Tackling Wasteful Spending on Health, OECD
International variation in admissions for COPD is 25-fold Publishing, Paris, http://dx.doi.org/10.1787/9789264266414-en.
across OECD countries, with Japan and Italy reporting the OECD (2017b), Caring for Quality in Health, Lessons Learnt
lowest rates and Hungary and Ireland the highest rates. from 15 Reviews of Health Care Quality Publishing, Paris,
Combined, there is a lower 7-fold variation across countries http://www.oecd.org/els/health-systems/Caring-for-Quality-
for the two respiratory conditions. in-Health-Final-report.pdf.
Hospital admission rates for CHF vary 12-fold, as shown Ponikowski, P. et al (2014), “Heart Failure: Preventing Disease
in Figure 6.9 Colombia, Costa Rica and Mexico, have the and Death Worldwide”, ESC Heart Failure, No. 1, pp. 4–25,
lowest rates, while Hungary, Poland and Lithuania report http://dx.doi.org/10.1002/ehf2.12005.
rates about 2 times the OECD average. WHO (2016), “Chronic Obstructive Pulmonary Disease
Figure 6.10 reveals that in Austria, Israel and Ireland a (COPD)”, November 2016 http://www.who.int/mediacentre/
reduction in admission rates for CHF has been achieved factsheets/fs315/en/.
104 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Avoidable hospital admissions
6.8. Asthma and COPD hospital admission in adults, 2015 (or nearest year)
COPD Asthma
Age-sex standardised rates per 100 000 population
450
414
411
428
400
371
363
341
333
330
350
309
303
286
284
282
300
262
261
259
247
238
237
234
234
250
223
202
193
186
184
200 184
150
146
138
137
129
150
99
96
89
100
74
64
58
50
0
o
Fr ia
Sw ce
M ia
er c
nd
C a lic
n
Po l y
lo l
s t il e
E ic a
i t z ni a
S l a nd
Lu F in en
Cz m nd
Ne epu ¹
Ic d s
P o d¹
ov OE in
Re 3 4
da
Un No el
d ay
th s
Ge ania
i t e B e l ny
Ki m
Au ea
D e s t r ia
Ne L a k
Ze ia
Au land
Ir e l i a
Tu d
Hu ey
y
Co uga
R urg
th bli
Li ate
ar
ar
ic
pa
n
en
b
ra
w tv
It a
d gi u
do
an
rk
an
ite r w
b
C o Ch
ra
ed
na
a
n
la
xe la
la
ak CD
aR
m
Sp
ex
S w sto
Ko
nm
ng
rm
pu
l
Ja
la
Is
ov
e c bo
u
St
er
rt
a
st
el
ng
h
Un
Sl
1. Three-year average.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603488
6.9. Congestive heart failure (CHF) hospital admission 6.10. Trends on CHF hospital admission in adults,
in adults, 2015 (or nearest year) selected countries
Age-sex standardised rates per 100 000 population Age-sex standardised rate per 100 000 population
450
Lithuania 576
Poland 464
Hungary 441
Slovak Republic 417
400
Germany 387
Czech Republic 380
United States 347
Finland 312
350
Estonia 269
France 266 Austria
Slovenia 261
Austria 259
300
Sweden 250
Israel 248
Israel
OECD32 228 Ireland
Italy 226
250
Australia 217
New Zealand 216
Spain 196
Belgium 189
200 Belgium
Netherlands 181
Iceland¹ 179
Switzerland 174
Portugal 167
150
Canada 167 Spain
Norway 160
Ireland 159
Denmark 150
100
Japan 137
Turkey 126
United Kingdom 101
Chile 98
50
Korea 94
Mexico 62
Costa Rica 51
Colombia 47
0
0 200 400 600 2000 2003 2006 2009 2012 2015
1. Three-year average.
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603507 12 http://dx.doi.org/10.1787/888933603526
Health at a Glance 2017 © OECD 2017 105
6. QUALITY AND OUTCOMES OF CARE
Diabetes care
Diabetes is a chronic disease that occurs when the body’s
ability to regulate excessive glucose levels in the blood is Definition and comparability
diminished. It is a leading cause of cardiovascular disease,
People with diabetes who have first choice
blindness, kidney failure and lower limb amputation.
antihypertensive medication prescriptions is defined
Globally it is estimated that over 400 million adults had
as the number of people that have one or more
diabetes in 2015 and by 2040 it is projected this will grow
prescriptions of an angiotensin converting enzyme
to over 640 million adults. Diabetes caused 5 million deaths
inhibitor (ACE-I) or angiotensin receptor blocker (ARB)
in 2015 (IDF, 2015).
among people who are long term users of glucose
Ongoing management of diabetes usually involves a regulating medication (people with diabetes) who
considerable amount of self-care, and therefore, advice and also have one or more prescriptions per year from a
education are central to the primary care of people with range of medications often used in the management
diabetes. Effective control of blood glucose levels through of hypertension.
routine monitoring, dietary modification and regular
Diabetes avoidable admission is based on the sum of
exercise can reduce the onset of serious complications and
three indicators: admissions for short-term and long-
the need for hospitalisation. Management of other key risk
term complications and for uncontrolled diabetes
factors such as smoking, blood pressure and lipid levels
without complications. The indicator is defined as
are also important in reducing complications of diabetes.
the number of hospital admissions with a primary
In diabetic individuals with hypertension, angiotensin- diagnosis of diabetes among people aged 15 years and
converting enzyme inhibitors (ACE-I) or angiotensin over per 100 000 population. Major lower extremity
receptor blockers (ARB) are recommended in most national amputation in adults with diabetes is defined as the
guidelines as first-line medications to reduce blood number of discharges of people aged 15 years and over
pressure. Figure 6.12 reveals there is broad consistency per 100 000 population, for the general population and
in the proportion of diabetic patients on recommended the estimated population with diabetes. Rates for
antihypertensive medications, with only Korea, Italy, these indicators have been directly age-standardised
Finland, Belgium and the Slovak Republic with rates less to the 2010 OECD population.
than 80%.
Differences in data definition and coding practices
Figure 6.11 shows avoidable hospital admissions for between countries may affect the comparability of
diabetes. While admissions have fallen in many countries data. For example, coding of diabetes as a principal
over time, more than a 7-fold variation in the rates is still diagnosis versus a secondary diagnosis varies across
evident across countries. Italy, Iceland and Spain report the countries. This is more pronounced for diabetes than
lowest rates with Austria, Korea and Mexico reporting rates other conditions, given that in many cases admission
at least two times that of the OECD average. Prevalence is for the secondary complications of diabetes rather
of diabetes may explain some of the variation in rates. A than diabetes itself. Diabetes population estimates
positive relationship can be demonstrated between overall used to calculate the amputation indicators were self-
hospital admissions and admissions for diabetes, providing reported by countries. Differences in data coverage of
some indication that access to hospital care can also play the national hospital sector across countries may also
a role in explaining international variation (OECD, 2015). influence indicator rates.
Hospital admissions for major lower extremity amputation
reflect the long-term quality of diabetes care. Figure 6.13
shows the rates of amputation in adults with diabetes. In
the left panel the rates based on the general population References
are presented. The international variation in rates is over
Carinci, F. et al. (2016), “Lower Extremity Amputation Rates
14-fold, with Colombia, Korea, Italy, Finland, and the United
in People with Diabetes as an Indicator of Health Systems
Kingdom reporting rates lower than 3 per 100 000 general
Performance. A Critical Appraisal of the Data Collection
population and Austria, Israel and Mexico reporting rates
2000–2011 by the Organization for Economic Cooperation
above 14. In the right panel rates based on the estimated
and Development (OECD)”, Acta Diabetologica, Vol. 53,
diabetic population are presented. The rates based on
pp. 825-832.
the diabetic population are 9-fold higher than for the
general population and display differences in the ranking IDF – International Diabetes Federation (2015), IDF Diabetes
of countries, providing an indication that differences in Atlas Seventh Edition 2015, www.diabetesatlas.org.
disease prevalence across countries may explain some, but OECD (2015), Cardiovascular Disease and Diabetes: Policies
not all, cross-country variation. In OECD countries, rates of for Better Health and Quality of Care, OECD Health
amputation have declined significantly since 2000 (Carinci Policy Studies, OECD Publishing, Paris, http://dx.doi.
et al., 2016). org/10.1787/9789264233010-en.
106 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Diabetes care
6.11. Diabetes hospital admission in adults, 2010 and 2015 (or nearest year)
2010 2015
Age-sex standardised rates per 100 000 population
400
350
300
250 281 292
266
200 218 222 225
151 204
141 187 191 192 197
150 171
130 133 136 137 141 143 148 151
100
92 94 96 101 110 113
50 66 69 70 73 73 74
41 48 53
40
0
o
Hu nia
Lu L c e
Un J lic
Ge blic
Ic a l y
S ¹
lo n
Po bia
N e Is l
S rl l
ite t ze s
K i nd
No m
Ir e a y
Ca nd
Sw da
S l de n
De ar y
Es r k
a
s t il e
OE ic a
Au D 3 3
F i li a
Ne Belg d
Ze m
Fr d
C z em t v i a
pu ¹
d an
ov P t es
Re nd
Tu y
th y
Au ia
Ko a
M a
th r ae
ga
Re u r g
d
ri
an
ni
L i ke
re
Un w i a n d
ic
Co pai
an
an
n
w iu
do
a
an
rw
an
b
C o Ch
ra
na
ite ap
d rla
ak ola
la
st
aR
It
e
m
r tu
to
ua
ex
a
nm
ng
r
rm
pu
e
nl
al
x a
ov
e c bo
C
St
st
el
ng
e
Sl
1. Three-year average.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603545
6.12 People with diabetes with a prescription 6.13. Major lower extremity amputation
of recommended antihypertensive medication in adults with diabetes, 2015
in the past year, 2015 (or nearest year) (or nearest year)
Slovenia 91.1
1.0 Colombia 18
2.4 Korea 13
Portugal 89.7 2.7 Italy 30
2.8 Finland 255
Australia 89.6
2.9 United Kingdom
New Zealand 88.7 3.1 Switzerland
3.2 Ireland
Spain 86.6 3.5 Iceland¹
3.6 Luxembourg¹ 32
Canada 85.8
3.7 Turkey 17
Ireland 85.8 3.7 Sweden
4.1 Belgium
Denmark 85.5 4.2 Australia
4.3 France 44
Sweden 85.4
4.7 Netherlands 43
Estonia 84.0 4.9 Poland
5.9 New Zealand 49
Turkey 83.0
6.2 Spain 52
Norway 82.3
6.4 Norway 79
6.4 OECD27/18 67
Netherlands 81.2 6.7 Lithuania 80
7.4 Canada 52
Korea 79.9
8.1 Latvia 81
Italy 79.1 8.3 Denmark 101
8.6 Germany 70
OECD18 79.1 8.9 Estonia 65
9.1 Portugal 62
Belgium 73.8
9.3 Costa Rica
Finland 59.4 14.1 Austria
15.7 Israel 87
Slovak Republic 12.3 19.5 Mexico 67
0 20 40 60 80 100 20 10 0 0 100 200 300
% of patients with diabetes Age-sex standardised rates Age-sex standardised rates
per 100 000 population per 100 000 people with diabetes
Note: Data for Canada only includes provinces of British
Columbia, Manitoba and Saskatchewan. 1. Three-year average.
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603564 12 http://dx.doi.org/10.1787/888933603583
Health at a Glance 2017 © OECD 2017 107
6. QUALITY AND OUTCOMES OF CARE
Mortality following ischaemic stroke
Worldwide an estimated 26 million people have experienced and Latvia rates have increased over this period by more
a stroke, with over 10 million people having an initial stroke than 1% point (Figures 6.14 and 6.15). Across the OECD, case
each year. Stroke is the second leading global cause of fatalities fell from 9.2% to 8.2% when considering same
death behind heart disease and accounted for just under hospital rates and from 12.4% to 11.6% when considering
12 percent of total deaths worldwide in 2013 (American in- and out-of-hospital rates. Figure 6.16 illustrates the
Heart Association, 2017). Stroke is also the second leading evolution of stroke rates for selected countries over this
cause of disability. A stroke occurs when the blood supply period, noting the United Kingdom was able to reduce
to a part of the brain is interrupted, leading to a necrosis their rates by an average annual reduction of more than
(i.e. cell death) of the affected part. 5% compared to an OECD average of 0.8%. Better access to
Of the two types of stroke that exist, about 85% are high-quality stroke care, including timely transportation
ischaemic (caused by clotting) and 15% are haemorrhagic of patients, evidence-based medical interventions and
(caused by bleeding).Treatment for ischaemic stroke has high-quality specialised facilities such as stroke units have
advanced dramatically over the last decade with systems helped to reduce 30-day case-fatality rates (OECD, 2015b).
and processes now in place in many OECD countries to Despite the progress seen so far, there is still room to
identify suspected ischaemic stroke patients as early as improve implementation of best practice acute care for
possible and to quickly deliver acute reperfusion therapy. cardiovascular diseases including stroke across countries.
Figure 6.14 shows the case-fatality rates within 30 days of To shorten acute care treatment time, targeted strategies
admission for ischaemic stroke where the death occurred can be highly effective. Advances in technology are now
in the same hospital as the initial stroke admission. leading to models of care to deliver reperfusion therapy
Figure 6.15 shows the case-fatality rate where deaths in an even more speedy and efficient manner, whether
are recorded regardless of where they occurred (after through pre-hospital triage via telephone, administration
transfer to another hospital or after discharge). This via telemedicine, or actually administering the therapy
indicator is more robust because it captures fatalities in the ambulance (Chang and Prabhakaran, 2017). But
more comprehensively. Although more countries report to encourage the use of evidence-based advanced
the same-hospital measure using unlinked data, an technologies in acute care, wider approaches are needed.
increasing number of countries are investing in their Adequate funding and trained professionals should be
data infrastructure and using linked data to provide more made available, and health care delivery systems should
comprehensive measures. be adjusted to enable easy access (OECD, 2015b).
Across OECD countries 8.2% of patients in 2015 died within
30 days in the same hospital in which the initial admission
for ischaemic stroke occurred (Figure 6.14). The case-fatality Definition and comparability
rates were highest in Latvia (18.3%) and Mexico (19.2%).
Case-fatality rates are defined in indicator “Mortality
Rates were less than 4% in Costa Rica, Korea, and Japan.
following acute myocardial infarction” in Chapter 6.
In Japan, many efforts have been dedicated to improving
the treatment of stroke patients in hospitals, through
systematic blood pressure monitoring, major material
investment in hospitals and the establishment of stroke References
units (OECD, 2015a). With the exception of Japan, Korea
and Germany, countries that achieve better results for American Heart Association (2017), “Heart Disease and
ischaemic stroke also tend to report good case-fatality rates Stroke Statistics 2017 At-a-Glance”, https://www.heart.org/
for acute myocardial infarction (AMI). This suggests that idc/groups/ahamah-public/@wcm/@sop/@smd/documents/
certain aspects of acute care may be influencing outcomes downloadable/ucm_491265.pdf, accessed 01-09-2017.
for both stroke and AMI patients. Chang, P. and S. Prabhakaran (2017), “Recent Advances in the
Across the 22 countries that reported in- and out-of-hospital Management of Acute Ischaemic Stroke”, F1000Research, 6,
case-fatality rates, 11.6% of patients died within 30-days F1000 Faculty Rev–484, http://doi.org/10.12688/
of being admitted to hospital for stroke (Figure 6.15). This f1000research.9191.1.
figure is higher than the same-hospital based indicator OECD (2015a), OECD Reviews of Health Care Quality: Japan
because it only counts each patient once and captures 2015: Raising Standards, OECD Publishing, Paris, http://
deaths that occur not just in the same hospital but also in dx.doi.org/10.1787/9789264225817-en.
other hospitals and out-of-hospital. OECD (2015b), Cardiovascular Disease and Diabetes: Policies for
Between 2010 and 2015, case-fatality rates for ischaemic Better Health and Quality of Care, OECD Publishing, Paris,
stroke have decreased substantially, whereas in Costa Rica http://dx.doi.org/10.1787/9789264233010-en.
108 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Mortality following ischaemic stroke
6.14. Thirty-day mortality after admission to hospital for ischaemic stroke based on unlinked data,
2010 and 2015 (or nearest years)
Confidence Interval 2015 2010 2015
Age-sex standardised rate per 100 admissions of adults aged 45 years and over
30
25
20
19.2
18.3
15
15.3
10
12.1
10.6
10.1
10.1
9.9
9.7
9.7
9.6
9.6
9.2
8.6
8.5
8.4
8.3
8.2
8.1
2.7
7.7
7.3
7.1
6.8
6.8
6.7
6.3
6.2
6.1
5.4
5.0
4.8
4.6
4.2
3.9
3.1
ov Es d a
o
Ne erla e
Li enia
li a
A u el
r ia
bo d
OE rg¹
Ic 3 2
B e n d¹
Ca m
Re ni a
No r k
F ay
it z nd
S w nd
Ge en
y
Au ly
i t e un c
Ir e i c
a
n
d ea
De t e s
C z K in ar y
Re m
Po nd
Tu al
ey
n
Sl hil e
La a
M ia
Lu Z e a s
an
th nc
Un H bli
ni
w nd
ic
ic
pa
ai
xe l a n
bl
ra
g
It a
tv
iu
h do
a
rk
rw
i t e Ko r
ra
ed
na
la
S w inla
la
st
CD
u
aR
Sp
ak to
r tu
ua
ex
a
nm
d g
a
rm
Ne Fr a
pu
pu
C
Ja
Is
lg
ov
St
er
st
el
ec g
th
st
m
Co
Un
Sl
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas.
1. Three-year average.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603602
6.15. Thirty-day mortality after admission to hospital for ischaemic stroke based on linked data,
2010 and 2015 (or nearest years)
Confidence Interval 2015 2010 2015
Age-sex standardised rate per 100 patients aged 45 years and over
30
25
20
15
10
0
ia
a
nd
ay
el
ly
en
g¹
ic
n
22
il e
ia
s
ga
da
ke
re
ar
nd
ni
ar
ai
an
an
bl
en
ra
It a
tv
do
ur
rw
Ch
ed
la
CD
Sp
r tu
to
Ko
nm
na
ng
r
pu
nl
al
la
La
Is
bo
ov
er
Tu
ng
Sw
No
Es
er
Ze
Fi
OE
Ca
Hu
Po
Re
it z
De
Sl
m
Ki
th
w
xe
Sw
h
Ne
d
Ne
ec
Lu
ite
Cz
Un
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas.
1. Three-year average.
2. Results for Canada do not include deaths outside of acute care hospitals.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603621
6.16. Thirty-day mortality after admission to hospital for ischaemic stroke based on linked data for selected
countries
Canada Italy Korea Sweden United Kingdom
Age-sex standardised rate per 100 patients aged 45 years and over
20
18
16
14
12
10
8
6
4
2
0
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
12 http://dx.doi.org/10.1787/888933603640
Health at a Glance 2017 © OECD 2017 109
6. QUALITY AND OUTCOMES OF CARE
Mortality following acute myocardial infarction (AMI)
Mortality due to coronary heart disease has declined to high-quality acute care for heart attack, including
substantially since the 1970s (see indicator “Mortality from timely transportation of patients, evidence-based medical
circulatory diseases” in Chapter 3). Important advances in interventions and specialised health facilities such as
both prevention policies, such as for smoking (see indicator percutaneous catheter intervention-capable centres have
“Smoking among adults” in Chapter 4), and treatment of helped to reduce 30-day case-fatality rates (OECD, 2015a).
cardiovascular diseases have contributed to these declines For example, Korea had higher case-fatality rates for AMI
(OECD, 2015a). but in 2006 it has implemented a Comprehensive Plan
A good indicator of acute care quality is the 30-day AMI for CVD, encompassing prevention, primary care and
case-fatality rate. The measure reflects the processes of acute CVD care (OECD, 2012). Under the Plan, specialised
care, such as timely transport of patients and effective services were enhanced through a creation of regional
medical interventions. The indicator is influenced by not cardio and cerebrovascular centres throughout the country,
only the quality of care provided in hospitals but also and average waiting time from emergency room arrival
differences in hospital transfers, average length of stay to initiation of catheterisation fell from 72.3 in 2010 to
and AMI severity. 65.8 minutes in 2011, leading to a reduction in case-fatality
(OECD, 2015a).
Figure 6.17 shows the case-fatality rates within 30 days
of admission for AMI where the death occurs in the same
hospital as the initial AMI admission. The lowest rates are
found in Australia, Denmark and Norway (all 4% or less). The Definition and comparability
highest rates are in Latvia, Hungary and Mexico, suggesting
AMI patients do not always receive recommended care. In The case-fatality rate measures the percentage of
Mexico, the absence of a coordinated system of care between people aged 45 and over who die within 30 days
primary care and hospitals may have contributed to delays in following admission to hospital for a specific acute
repurfusion and low rates of angioplasty (Martínez-Sánchez, condition. Rates based on unlinked data refer to
2017). High rates of uncontrolled diabetes may also be a a situation where the death occurred in the same
contributing factor in explaining the high AMI case-fatality hospital as the initial admission. Rates based on
rates (see indicator “Diabetes care” in Chapter 6) as patients linked data refer to a situation where the death
with diabetes have worse outcomes after AMI compared to occurred in the same hospital, a different hospital, or
those without diabetes, particularly if the diabetes is poorly out of hospital. While the linked data based method
controlled. In Japan, people are less likely to die of heart is considered more robust, it requires a unique patient
disease overall, but are more likely to die once admitted into identifier to link the data across the relevant datasets
hospital for AMI compared to many other OECD countries. which is not available in all countries.
One possible explanation is that the severity of patients’ Rates are age-sex standardised to the 2010 OECD
admitted to hospital with AMI may be more advanced population aged 45+ admitted to hospital for a specific
among a smaller group of people across the population, but acute condition such as AMI (ICD-10 I21, I22) and
could also reflect underlying differences in emergency care, ischaemic stroke (ICD-10 I63-I64).
diagnosis and treatment patterns (OECD, 2015b).
Figure 6.18 shows 30-day case fatality rates where fatalities
are recorded regardless of where they occur (after transfer
to another hospital or after discharge). This is a more robust References
indicator because it records deaths more widely than the
same-hospital indicator, but it requires a unique patient Martínez-Sánchez, C. et al. (2017), “Reperfusion Therapy
identifier and linked data which is not available in all of Myocardial Infarction in Mexico: A Challenge for
countries. The AMI case-fatality rate ranges in 2015 from Modern Cardiology”, Archivos de cardiología de México,
7.1% in Canada to 18% in Latvia. Vol. 87, No. 2, pp 144-150, http://dx.doi.org/10.1016/j.
acmx.2016.12.007.
Case-fatality rates for AMI have decreased substantially
between 2005 and 2015 (Figures 6.17 and 6.18). Across OECD (2015a), Cardiovascular Disease and Diabetes: Policies
the OECD, case fatalities fell from 8.5% to 7.5% when for Better Health and Quality of Care, OECD Health
considering same hospital deaths and from 11.3% to 9.9% Policy Studies, OECD Publishing, Paris, http://dx.doi.
when considering deaths occurred in and out of hospital. org/10.1787/9789264233010-en.
The rate of decline was particularly striking in Finland, OECD (2015b), OECD Reviews of Health Care Quality: Japan
the Netherlands and Denmark, when considering deaths 2015: Raising Standards, OECD Publishing, Paris, http://
occurred in and out of hospital, with an average annual dx.doi.org/10.1787/9789264225817-en.
reduction of over 4% compared to the OECD average of 2.5%. OECD (2012), OECD Reviews of Health Care Quality: Korea
Figure 6.19 illustrates the evolution of the decline in AMI 2012: Raising Standards, OECD Publishing, Paris, p://dx.doi.
case fatality rates for selected countries. Better access org/10.1787/9789264173446-en.
110 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Mortality following acute myocardial infarction (AMI)
6.17. Thirty-day mortality after admission to hospital for AMI based on unlinked data, 2010 and 2015
(or nearest years)
Confidence Interval 2010 2015
Age-sex standardised rate per 100 admissions of adults aged 45 years and over
28.1
30
25
20
13.9
15
13.4
10
7.3
11.7
6.5
11.3
10.6
8.6
8.1
7.9
7.9
5
7.7
7.5
7.4
7.1
7.0
6.9
6.7
6.4
6.4
6.1
5.9
5.6
5.6
5.4
5.4
5.1
5.1
4.7
4.4
4.2
4.0
4.0
3.7
o
it z da
Fi e
De w ay
A u ar k
S w li a
en
Ze d
C d
t h nd
Re nia
U n Ir e c
ite Belg c
m m
Au g¹
O E r ia
Ge 3 4
y
Po in
ly
Ic d
o v S l o d¹
Ko l
Tu a
Es ey
St d
es
a
il e
n
Hu v i a
M ry
Re el
ga
i
an
i
c
ni
re
nd
ic
pa
w lan
an
i t e lan
an
bl
bl
ra
It a
a
u
xe do
ur
an
a
rk
an
at
Ch
ra
ed
Sw ana
Ne erla
st
CD
t
e
Sp
r tu
to
ex
nm
ng
rm
pu
pu
nl
Ja
al
la
La
e c Is
r
bo
v
Ne Po
st
el
Lu ing
Fr
No
er
K
h
ak
d
Cz
Un
Sl
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas.
1. Three-year average.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603659
6.18. Thirty-day mortality after admission to hospital for AMI based on linked data, 2010 and 2015
(or nearest years)
Confidence Interval 2010 2015
Age-sex standardised rate per 100 patients aged 45 years and over
30
25
20
15
10
0
²
ly
ay
en
nd
el
nd
ia
s
es
ic
a
g¹
il e
ia
y
23
ga
da
ke
re
ni
ar
nd
ar
an
ai
an
en
bl
ra
It a
tv
do
ur
rw
at
Ch
ed
la
la
CD
Sp
r tu
to
nm
Ko
na
ng
r
pu
nl
al
la
La
Is
ov
bo
St
Po
er
Tu
ng
Sw
No
Es
Ze
er
Fi
OE
Ca
Hu
Po
Re
itz
De
Sl
m
d
Ki
th
ite
w
xe
Sw
h
Ne
d
Ne
ec
Un
Lu
ite
Cz
Un
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas.
1. Three-year average.
2. Results for Canada do not include deaths outside of acute care hospitals.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603678
6.19. Thirty-day mortality after admission to hospital for AMI based on linked data for selected countries
Canada Finland Israel New Zealand Poland
Age-sex standardised rate per 100 patients aged 45 years and over
20
18
16
14
12
10
8
6
4
2
0
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
12 http://dx.doi.org/10.1787/888933603697
Health at a Glance 2017 © OECD 2017 111
6. QUALITY AND OUTCOMES OF CARE
Hospital mortality rates
Variations in acute myocardial infarction (AMI) 30-day
case fatality rates at the national level are influenced The specific methodology used to calculate the
by the level of within-country variation in rates across hospital case fatality rates presented here differs
hospitals. Most OECD countries have established national from that used for the indicator “Mortality following
hospital performance measurement and public reporting acute myocardial infarction” and is likely to vary
programmes to monitor efforts to improve the cost, quality from the methods used by participating countries
and access of hospital care. for national monitoring and reporting purposes. Key
methodological choices include: unit of measurement,
Figure 6.20 plots the AMI 30-day case fatality rates (where type of hospital, patient risk adjustment variables,
the death occurs in the same hospital as the initial AMI selection of reference population, method of
admission). Rates are presented according to the caseload standardisation and data issues.
for each hospital and identifies where the rates are higher
or lower than expected. While most hospitals have rates Different analytical methods can result in quite
no different than expected, all countries (except Norway) different rates for and rankings of organisations and
had at least one outlier hospital. countries, making direct comparison between rates
problematic. The specific analytical method used here
The total number of hospitals and proportion of hospitals is one of several valid options considered during the
by number of AMI admissions varies across countries development work of the OECD. For more details on
(Table 6.1). Countries with a large number of hospitals are the methodology used to calculate these indicators
likely to have more outlier hospitals than countries with see Brownwood et al. (forthcoming).
fewer hospitals. Figure 6.21 presents the differences in
dispersion of AMI 30-day case fatality rates across hospitals Figure 6.20 is a funnel plot and reflects that the
within countries. The interquartile range of rates within precision of indicator rates increases as the caseload
countries varies markedly. For example, the difference increases. All rates within the 99.7% control limits
between the upper and lower rates for Israel is 1.8 deaths are considered to be no different than expected,
per 100 admissions, and 4.9 deaths per 100 admissions whereas those outside the 99.7% control limits are
for Latvia (based on unlinked data). Using linked data, the considered higher or lower than expected. The
results are slightly different, with Sweden rather than Israel reference population rate was calculated from pooled
having the least within-country variation. data from selected countries and used to calculate
the standardised rates. Figure 6.21 is a turnip plot
Multiple factors contribute to variations in outcomes of that graphically represents the relative dispersion
care including hospital structure, processes of care and of rates but does not give an indication of statistical
organisational culture. Significant variation in adherence significance of the variations in rates. Countries are
to guideline recommendations for cardiac care is observed ordered according to ascending level of dispersion
across countries and within countries (OECD, 2015, p. 174). as measured by the interquartile range (between the
In Sweden, a comprehensive national programme of 25 th percentile and the 75 th percentile) of rates.
quality improvement that includes public reporting, rapid Hospitals with less than 50 AMI admissions were
diffusion of technology, use of evidence-based practice and excluded from both figures to improve data reliability.
a system of evaluating and reporting quality and outcomes
of care is likely to have contributed to a reduced variation
in hospital care of patients after an AMI (Chung et al.,
2015, p. 7). References
Brownwood, I. et al. (forthcoming), “OECD Hospital
Performance Project: Methodological Development
Definition and comparability of International Measurement of Acute Myocardial
Infraction 30-Day Mortality Rates at the Hospital Level”,
The case-fatality rate measures the percentage of
OECD Health Working Papers, OECD Publishing, Paris.
people aged 45 and over who die within 30 days
following admission to hospital for a specific acute Chung, S.C. et al. (2015), “Comparison of Hospital Variation
condition. Rates based on unlinked data refer to in Acute Myocardial Infarction Care and Outcome
situations where the death occurred in the same Between Sweden and United Kingdom: Population Based
hospital as the initial admission. Rates based on linked Cohort Study Using Nationwide Clinical Registries”,
data include all deaths irrespective of where they British Medical Journal, Vol. 351, http://doi.org/10.1136/
occur. While the linked data method is considered bmj.h3913.
more robust, it requires a unique patient identifier to OECD (2015a), Cardiovascular Disease and Diabetes: Policies
link the data across the relevant datasets, which is not for Better Health and Quality of Care, OECD Health
available in all countries. Policy Studies, OECD Publishing, Paris, http://dx.doi.
org/10.1787/9789264233010-en.
112 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Hospital mortality rates
6.20. Thirty-day mortality after admission to hospital for AMI based on linked data, 2013-2015
(or nearest years)
Within expected range Canada Denmark Finland
Israel Italy Korea Latvia
Norway Slovenia Sweden United Kingdom
Reference population rate 99.7% Control limits
Age, sex, co-morbidity standardised mortality rates per 100 admissions of adults aged 45 years and over
40
35
30
25
20
15
10
0
50 550 1 050 1 550 2 050 2 550 3 050 3 550 4 050 4 550
Number of AMI admissions, 2013-2015 (or nearest years)
Note: Each dot in the figure represents a single hospital, unless otherwise stated. Results for Canada do not include deaths outside of acute care
hospitals. UK data are limited to England and is presented at trust-level (i.e. multiple hospitals).
Source: OECD Hospital Performance Data Collection 2017.
12 http://dx.doi.org/10.1787/888933603716
Table 6.1. Number of hospitals by AMI admissions based on unlinked data, 2013-2015 (or nearest years)
AMI admissions CAN DNK FIN ISR IRE ITA KOR LVA NOR SVN SWE GBR
> 300 151 21 21 21 20 336 67 6 35 3 62 142
50-300 158 7 0 5 8 160 83 11 17 7 4 8
< 50 261 1 0 0 6 328 155 5 2 4 0 59
12 http://dx.doi.org/10.1787/888933606262
6.21. Thirty-day mortality after admission to hospital for AMI based on linked and unlinked data,
2013-2015 (or nearest years)
Based on unlinked data Based on linked data
Age, sex, co-morbidity standardised mortality rates per 100 admissions of adults aged 45 years and over
40
30
20
10
0
m
ia
el
en
ay
da
nd
ly
ia
re
ar
an
en
ra
It a
tv
do
rw
ed
na
la
nm
Ko
nl
La
Is
ov
Ir e
ng
Sw
Ca
No
Fi
De
Sl
Ki
d
ite
Un
Note: The width of each line in the figure represents the number of hospitals (frequency) with the corresponding rate. Data for Canada not linked to
death statistics. UK data are limited to England and presented at trust level (i.e. multiple hospitals). Ordered by inter quartile range of admission-based
data. Rates based on linked data are also standardised for previous AMI.
Source: OECD Hospital Performance Data Collection 2017.
12 http://dx.doi.org/10.1787/888933603735
Health at a Glance 2017 © OECD 2017 113
6. QUALITY AND OUTCOMES OF CARE
Waiting times for hip fracture surgery
The main risk factors for hip fractures are associated with Time to surgery for hip fracture patients is influenced by
ageing, including an increased risk of falling and loss of many factors, including hospitals’ surgical theatre capacity,
skeletal strength from osteoporosis. With increasing life flow and access and targeted policy interventions, including
expectancy across most OECD countries, it is anticipated public reporting and monitoring of performance (Siciliani
that hip fracture will become a more significant public et al, 2013) Improvement in timely surgery for patients with
health issue in coming years. a particular diagnosis or injury (e.g. hip fracture) may be
In most instances following hip fracture, surgical achieved at the expense of timeliness in others (e.g. hip or
intervention is required to repair or replace the hip joint. knee replacements).
There is general consensus that early surgical intervention
maximises patient outcomes and minimises the risk of
complications. General agreement is that surgery should Definition and comparability
occur within two days (48 hours) of hospitalisation.
Guidelines in some countries call for even earlier This indicator is defined as the proportion of patients
intervention. For example, the National Institute for Health aged 65 years and over admitted to hospital in a
and Care Excellence (NICE) clinical guidelines recommend specified year with a diagnosis of upper femur fracture,
hip fracture surgery to be performed on the day of hospital who had surgery initiated within two calendar days
admission or the next day (National Institute for Health of their admission to hospital. Data are also provided
and Care Excellence, 2014). for the proportion of those patients who had surgery
The time taken to initiate hip fracture surgery after within one day of their admission to hospital, and for
hospital admission is widely considered to be a clinically patients who had surgery on the same day as their
meaningful process indicator of the quality of acute care hospital admission. Some countries supplied results
received by patients with hip fracture. In 2015, on average for surgery within two calendar days only.
across the OECD over 80% of patients admitted for hip The capacity to capture time of admission and
fracture underwent surgery within two days (Figure 6.22). In surgery in hospital administrative data varies across
Norway, Denmark and the Netherlands, the proportion was countries, resulting in the inability to precisely record
greater than 95%. Countries with the lowest proportion of surgery within 48 hours. While recent research and
patients operated on within two days of admission include development data indicates that the impact of
Italy (53.2%), Spain (48.4%), Portugal (46.5%), Latvia (46.0%) measuring days rather than hours may only result
and Costa Rica (24.9%). in marginally higher rates, the impact on relative
Many patients were treated sooner than two days following performance across countries can be noticeable, given
admission, with about a quarter of patients treated on the the similarity of rates in many countries.
same day and around two thirds of patients treated by the While cases where the hip fractures occurred during
end of the next day across the OECD. Rates were higher the admission to hospital should be excluded, not
than 40% on the same day in the Netherlands, and 80% by all countries have a ‘present on admission’ flag in
the end of the next day in Denmark. their datasets to enable them to identify such cases
Figure 6.23 shows the proportion of hip-fracture repairs accurately.
occurring within two days of admission in OECD countries
between 2005 and 2015. The OECD average increased from
72% to 81% over that time. The greatest improvement was
observed in Switzerland, where the proportion increased References
from 46% to 91% and in Italy, where it increased from 28%
Canadian Institute for Health Information (2015), “Your
in 2007 to 53% in 2015. A policy of comparative public
Health System: In Depth”, [web tool], accessed on
reporting of hospital indicators, including time to surgery
01-09-2017.
following hip fracture, implemented by Italian authorities
may partly explain the improvement observed in that National Institute for Health and Care Excellence (2014),
country. In Canada, the percentage of patients operated “Hip Fracture: The Management of Hip Fracture in
on within the two day benchmark increased over time, but Adults”, NICE Clinical Guideline No. 124, issued June 2011,
there is considerable variation in this indicator between last modified March 2014.
provinces and hospitals (CIHI, 2015). Only Portugal reported Siciliani, L., M. Borowitz and V. Moran (eds.) (2013), Waiting Time
a decline of hip fracture repair within two days of admission, Policies in the Health Sector: What Works? OECD Publishing,
reducing from 57% in 2008 to 47% in 2015. Paris. http://dx.doi.org/10.1787/9789264179080-en.
114 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Waiting times for hip fracture surgery
6.22. Hip fracture surgery initiation after admission to the hospital, 2015 (or nearest year)
Day Two Next day Same day
% of patients aged 65 years and over
100
90.8
89.6
89.7
88.3
88.3
87.3
86.9
86.8
96.3
96.4
95.2
90
83.3
93.4
82.3
81.5
91.9
80.9
91.5
80
69.4
66.7
70
60
53.2
48.4
46.5
46.0
50
40
24.9
30
20
10
ic
ia
ay
da
nd
el
y¹
r ia
nd
22
ly
m
ia
a
ga
en
an
ni
ni
ar
nd
ic
ai
an
an
bl
en
ra
It a
tv
iu
ar
do
rw
na
la
la
st
CD
aR
Sp
to
ua
r tu
nm
ed
rm
pu
nl
al
la
La
Is
lg
ng
ov
er
Ir e
ng
Au
Ca
No
Es
th
er
Ze
Fi
OE
Sw
Be
Po
Re
st
it z
Ge
De
Hu
Sl
Ki
Li
th
Co
w
Sw
h
Ne
Ne
ec
ite
Cz
Un
1. Hungary only provided data for within two calendar days.
2. Sweden provided data within 12, 24 and 48 hours.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603754
6.23. Hip fracture surgery initiation after admission to hospital, 2005 and 2015 (or nearest year)
2005 2015
% of patients aged 65 years and over being operated within 2 days
100 96.4 96.3 95.2
93.4 91.9 91.5 90.8 89.7 89.6 88.3 88.0 87.3
90 86.9 86.8
83.3 82.3 81.5 80.6
80
69.4
70 66.7
60
53.2
48.4
50
46.5 46.0
40
30 24.9
20
10
0
ic
ia
21
ly
ia
a
ay
nd
el
r ia
nd
da
en
ga
an
ni
ni
ic
ar
nd
ar
ai
an
an
bl
en
ra
It a
tv
iu
do
rw
ed
na
la
la
st
CD
aR
Sp
to
ua
r tu
nm
ng
rm
pu
nl
al
La
la
Is
lg
ov
er
Ir e
ng
Au
Sw
Ca
No
Es
th
Ze
er
Fi
OE
Be
Hu
Po
Re
st
it z
Ge
De
Sl
Ki
Li
th
Co
w
Sw
h
Ne
Ne
ec
ite
Cz
Un
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603773
Health at a Glance 2017 © OECD 2017 115
6. QUALITY AND OUTCOMES OF CARE
Surgical complications
Patient safety remains one of the most prominent issues
in health policy and public debate. Evidence suggests that Definition and comparability
over 15% of hospital expenditure and activity in OECD
Two methods of calculating surgical complications are
countries can be attributed to treating safety failures, many
presented. The surgical admission-based method uses
of which are preventable (OECD, 2017a; OECD, 2017b). In the
unlinked data to calculate the number of discharges
United States an estimated USD 28 billion has been saved
with ICD codes for the complication in any secondary
between 2010 and 2015 by systematically improving safety
diagnosis field, divided by the total number of discharges
(AHRQ, 2016).
for patients aged 15 and older. The all admission-
Robust comparison of performance with peers is based method uses linked data to extend beyond the
fundamental to securing improvement. Two types of patient surgical admission to include all subsequent related
safety event can be distinguished for this purpose: sentinel re-admissions to any hospital within 30 days. While the
or “never” events that should never occur such as failure to all admission-based method is considered more robust
remove surgical foreign bodies at the end of a procedure; and is less affected by variations in the length of stay
and adverse events, such as post-operative sepsis, which and hospital transfer practices, it requires a unique
can never be fully avoided given the high-risk nature of patient identifier and linked data which is not available
some procedures, although increased incidence at an in all countries. While the all admission-based method
aggregate level may indicate a systemic failing. strengthens identification of valid complications, the
Figure 6.24 illustrates a never event, rates of foreign impact on indicator rates is unclear given only one
body left in during procedure. The most common risk admission per patient is counted when multiple
factors for this never event are emergencies, unplanned qualifying admissions are identified.
changes in procedure, patient obesity and changes in A fundamental challenge in international comparison
the surgical team; preventive measures include counting of patient safety indicators centres on differences
instruments, methodical wound exploration and effective in the underlying data. Variations in how countries
communication among the surgical team. record diagnoses and procedures and define hospital
Figure 6.25 shows rates for two related adverse events, admissions can affect calculation of rates. In some
pulmonary embolism (PE) and deep vein thrombosis (DVT) cases, higher adverse event rates may signal more
after hip or knee replacement surgery. PE and DVT cause developed patient safety monitoring systems and a
unnecessary pain and in some cases death, but can be stronger patient safety culture rather than worse care.
prevented by anticoagulants and other measures before, There is a need for greater consistency in reporting of
during and after surgery. Large variations in rates are patient safety across countries and significant scope
observed, with nearly a 20-fold variation in DVT. Variations exists for improved data capture within national
in DVT rates may be influenced by differences in diagnostic patient safety programmes.
practices across countries, with evidence that routine
ultrasound screening can significantly increase the
detection of DVT (Kodadek, 2016).
References
Figure 6.26 shows rates for another adverse event, sepsis
after abdominal surgery. Likewise, sepsis after surgery, AHRQ (2016), “National Scorecard on Rates of
which may lead to organ failure and death, can in many Hospital-Acquired Conditions 2010 to 2015: Interim
cases be prevented by prophylactic antibiotics, sterile Data From National Efforts to Make Health Care Safer”,
surgical techniques and good postoperative care. Agency for Healthcare Research and Quality, http://
www.ahrq.gov/sites/default/files/wysiwyg/professionals/
The left panel of Figures 6.24, 6.25 and 6.26 shows the
quality-patientsafety/pfp/2015-natlscorecard -hac-rates.pdf,
rate of the three respective postoperative complications
accessed 26.01.2017.
based on the “surgical admission”, the hospital admission
where the surgery took place. The right panel of these Kodadek, L.M. and E.R. Haut (2016), “Screening and Diagnosis
figures shows rates based on not only the surgical of VTE: The More You Look, The More You Find?”, Current
admission but all subsequent re-admissions to hospital Trauma Reports, Vol. 2, No. 1, pp. 29–34.
within 30 days, whether at the same hospital or in OECD (2017a), Tackling Wasteful Spending on Health, OECD
another hospital. Publishing, Paris, http://dx.doi.org/10.1787/9789264266414-en.
Caution is needed in interpreting the extent to which OECD (2017b), The Economics of Patient Safety: Strengthening a
these indicators accurately reflect international differences Value-Based Approach to Reducing Patient Harm at a National
in patient safety rather than differences in the way that Level Publishing, OECD, Paris, available at http://www.
countries report, code and calculate rates of adverse events oecd.org/els/health-systems/The-economics-of-patient-safety-
(see “Definition and comparability” box). March-2017.pdf.
116 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Surgical complications
6.24. Foreign body left in during procedure, 2015 (or nearest year)
Confidence interval
Per 100 000 surgical discharges
16
14
12
10 12.3
8
8.8 9.0
6 7.5
7.2 8.1 7.7
2.2
4 5.4 6.4 1.9 2.1 2.2
1.6 1.6 5.5
5.2 4.2
2 0.2 3.5 4.4
2.7 3.3 4.2 4.1
0
Slovenia
Slovenia
Poland
Belgium
Finland
Portugal
Israel
OECD13
Germany
United Kingdom
United States
Ireland
Sweden
Australia
Switzerland
Italy
New Zealand
Finland
Portugal
Spain
OECD10
Norway
Israel
Sweden
Canada
Surgical admission method All admission method
Note: Given very low incidence of events, 95% confidence intervals have been calculated for all countries as represented by grey areas.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603792
6.25. Postoperative pulmonary embolism (PE) or deep vein thrombosis (DVT) in hip and knee
surgeries, 2015 (or nearest year)
DVT PE
Per 100 000 hip and knee surgical discharges
1 800
1 600
267
1 400
549
1 200
69
1 000
572
800
1 328
657
600
1 113
340
443
510
1004
357 301
411
397
170
312 289
152
183
400
233 350
237 339
166
185 344
209 294
202 316
574
84
97
200
419
69
358
290
33
300
311
293
184
153
99
90
46
58
56
69
0
Slovenia
France
Poland
Portugal
Sweden
Belgium
United States
Switzerland
Ireland
OECD14
Finland
Israel
Germany
Australia
Spain
Norway
United Kingdom
Portugal
New Zealand
Sweden
OECD9
Finland
Canada
Italy
Israel
Surgical admission method All admission method
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603811
6.26. Postoperative sepsis in abdominal surgeries, 2015 (or nearest year)
Per 100 000 abdominal surgical discharges
4 000 3 673
3 500
3 000 2 658
2 500 2 371 2 386 2 453
2 201 2 269 2 316 2 224
1 989 2 117 2 129
2 000 1 717 1 862
1 500 1352 1 365 1 395
1 218
992 996 1 004
1 000 618
784
421 552
500 259 381
0
Slovenia
Poland
Sweden
Belgium
Germany
Portugal
OECD14
United States
Denmark
Israel
Switzerland
Finland
Australia
Ireland
Korea
New Zealand
Finland
Sweden
Denmark
Portugal
OECD11
Italy
Israel
Canada
Norway
Spain
United Kingdom
Surgical admission method All admission method
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603830
Health at a Glance 2017 © OECD 2017 117
6. QUALITY AND OUTCOMES OF CARE
Obstetric trauma
Patient safety during childbirth can be assessed by without instrument (1.5 per 100 vaginal deliveries without
looking at potentially avoidable tearing of the perineum instrument assistance), there is a strong relationship
during vaginal delivery (Harvey, 2015). Such tears extend between the two indicators, with Italy, Israel and Poland
to the perineal muscles and bowel wall require surgery. reporting the lowest rates and Canada, Denmark and New
They are more likely to occur in the case of first vaginal Zealand reporting amongst the highest rates for both
delivery, high baby birth weight, labour induction, occiput indicators.
posterior baby position, prolonged second stage of labour No clear trend is evident in the rates of obstetric trauma
and instrumental delivery. Possible complications include over the five year period 2010-2015, with the OECD average
continued perineal pain and incontinence. These types of remaining relative static for both vaginal deliveries with
tears are not possible to prevent in all cases, but can be and without instrument. While rates for both indicators
reduced by employing appropriate labour management indicate noticeable improvements in Denmark and Norway
and high quality obstetric care. Hence, the proportion of over this period, rates for some countries including Slovenia
deliveries involving higher degree lacerations is a useful and Spain would appear to have deteriorated.
indicator of the quality of obstetric care.
Obstetric trauma indicators are considered to be relatively
reliable and comparable across countries, particularly given
they are less sensitive to variations in coding practices across Definition and comparability
countries. Nevertheless, differences in the consistency
The two obstetric trauma indicators are defined as
with which obstetric units report these complications may
the proportion of instrument assisted/non-assisted
complicate international comparison. Fear of litigation, for
vaginal deliveries with third- and fourth-degree
example, may cause under-reporting; conversely systems
obstetric trauma codes (ICD-10 O70.2, O70.3) in any
that rely on specially trained administrative staff to identify
diagnosis and procedure field.
and code adverse events from patients’ clinical records may
produce more reliable data. Several differences in data reporting across countries
may influence the calculated rates of obstetric
While rates of obstetric trauma may be influenced by
patient safety indicators. These relate primarily to
the overall national rate of caesarean sections, assisted
differences in coding practice and data sources. Some
vaginal delivery and episiotomy, these remain issues of
countries report the obstetric trauma rates based on
ongoing research. For example, episiotomy is a surgical
administrative hospital data and others based on
incision of the perineum performed to widen the vaginal
obstetric register data. There is some evidence that
opening for the delivery of an infant. Wide variation in
registries produce higher quality data and report a
the use of episiotomy during vaginal deliveries currently
greater number of obstetric trauma events compared
exists across Europe, ranging from around 70% of births in
to administrative datasets (Baghestan et al., 2007).
Portugal and Poland in 2010 to less than 10% in Sweden,
Denmark and Iceland (Euro-Peristat, 2013). The selective Careful interpretation of obstetric trauma for
use of episiotomy to decrease severe perineal lacerations instrument assisted delivery rates over time is
during delivery remains controversial required, given the very low number of trauma cases
in some countries is likely to give rise to significant
Figure 6.27 shows rates of obstetric trauma with instrument
year on year variation.
and Figure 6.28 shows rates of obstetric trauma after
vaginal delivery without instrument. Obstetric trauma
with instrument refers to deliveries using forceps or
vacuum extraction. As the risk of a perineal laceration is
References
significantly increased when instruments are used to assist
the delivery, rates for this patient population are reported Baghestan, E. et al. (2007), “A Validation of the Diagnosis
separately. of Obstetric Sphincter Tears in Two Norwegian
High variation in rates of obstetric trauma is evident Databases, the Medical Birth Registry and the Patient
across countries. Reported rates of obstetric trauma with Administration System”, Acta Obstetricia et Gynecologica,
instrument vary from below 2% in Israel, Italy and Poland Vol. 86, pp. 205-209.
to more than 10% in Denmark, Sweden and Canada. The Euro-Peristat (2013), European Perinatal Health Report: Health
rates of obstetric trauma after vaginal delivery without and Care of Pregnant Women and Babies in Europe in 2010,
instrument vary from below 0.5 per 100 deliveries in Poland INSERM, Paris.
and Israel to over 2.5 per 100 deliveries in Denmark, United Harvey, M.A. et al. (2015), “Society of Obstetricians and
Kingdom and Canada. Gynaecologists of Canada, Obstetrical Anal Sphincter
While the average rate of obstetric trauma with instrument Injuries (OASIS): Prevention, Recognition, and Repair”,
(5.7 per 100 instrument-assisted vaginal deliveries) Journal of Obstetrics and Gynaecology Canada, Vol. 37,
across OECD countries in 2015 was nearly 4 fold the rate No. 12, pp. 1131-1148.
118 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Obstetric trauma
6.27. Obstetric trauma, vaginal delivery with instrument, 2010 and 2015 (or nearest year)
2010 2015
Crude rates per 100 instrument-assisted vaginal deliveries
18
16 16.9
14
12
11.3 9.6
10 10.9
8.5
7.4 7.2
8 6.8
6.4
5.7
6
4.8
4.2
3.7 3.4
4 3.2
3.9 2.5
2.1 1.9 1.9
2 2.5
0.7
0
s¹
da
k¹
21
nd
a¹
¹
es
nd
ia¹
el
ly
¹
li a
d¹
ga
en
ay
nd
an
ai
an
ra
It a
ni
nd
ar
iu
do
an
at
ra
en
na
la
la
CD
rw
Sp
r tu
ed
la
rm
to
al
lg
Is
nm
St
er
Ir e
st
la
nl
ng
ov
Ca
Po
Ze
Sw
OE
No
Be
Es
Po
er
Au
Fi
it z
Ge
d
Ki
De
Sl
th
ite
Sw
d
Ne
Ne
Un
ite
Un
1. Based on registry data.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603849
6.28. Obstetric trauma, vaginal delivery without instrument, 2010 and 2015 (or nearest year)
2010 2015
Crude rates per 100 vaginal deliveries without instrument assistance
4.0
3.5
3.1
3.0 2.8
2.5 2.5
2.5
2.6
2.4 2.4
2.0 2.1
1.6 1.5
1.5 1.4
1.6 1.2
1.4
1.0 0.9
0.7 0.7
0.5 0.5 0.5
0.5
0.5 0.1
0
da
k¹
li a
s¹
nd
nd
21
es
ia¹
a¹
ly
el
¹
d¹
ga
en
ay
nd
an
ai
an
ra
It a
ni
nd
ar
iu
do
an
at
ra
en
na
la
la
CD
rw
Sp
r tu
ed
la
rm
to
al
lg
Is
nm
St
er
Ir e
st
la
nl
ng
ov
Ca
Po
Ze
Sw
OE
No
Be
Es
Po
er
Au
Fi
it z
Ge
d
Ki
De
Sl
th
ite
w
Sw
d
Ne
Ne
Un
ite
Un
1. Based on registry data.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603868
Health at a Glance 2017 © OECD 2017 119
6. QUALITY AND OUTCOMES OF CARE
Care for people with mental health disorders
The burden of mental illness is substantial, affecting an can be assessed regularly. For example, Sweden monitors
estimated one in four of the OECD population at any time, the use of inpatient physical care for patients with a mental
and one in two across the life course (see indicator on disorder that could have been avoided if primary care and/
“Mental health” in Chapter 3; OECD, 2014a). High quality, or primary or secondary prevention was sufficient (OECD,
timely care has the potential to improve outcomes and may 2014a; OECD, 2014b).
help reduce suicide and excess mortality for individuals
with psychiatric disorders.
High quality care for mental disorders in inpatient settings Definition and comparability
is vital, and inpatient suicide is a ‘never event’, which
should be closely monitored as an indication of how well The inpatient suicide indicator is composed of a
inpatient settings are able to keep patients safe from harm. denominator of patients discharged with a principal
Figure 6.29 shows rates of inpatient suicide amongst all diagnosis or first two secondary diagnosis code of
psychiatric hospital admissions. Most countries report rates mental health and behavioural disorders (ICD-10
below 1 per 1 000 patients, but Costa Rica, the Netherlands, codes F10-F69 and F90-99) and a numerator of these
Denmark, and Israel are exceptions with rates of over 1. patients with a discharge code of “suicide” (ICD-10
Steps to prevent inpatient suicide include identification codes: X60‑X84). Data should be interpreted with
and removal of likely opportunities for self-harm, risk caution due to a very small number of cases. Reported
assessment of patients, monitoring and appropriate rates can vary over time, so where possible a 3-year
treatment plans. average has been calculated to give more stability to
the indicator.
Suicide rate after hospital discharge can indicate the quality
of care in the community, and co-ordination between Suicide within 30 days and within one year of
inpatient and community settings. Across countries, discharge is established by linking discharge following
suicide rate among patients who had been hospitalised hospitalisation with a principal diagnosis or first two
in the previous year was as low as 1 per 1 000 patients listed secondary diagnosis code of mental health
in the United Kingdom but it was higher than 5 in the and behavioural disorders (ICD-10 codes F10-F69 and
Netherlands and Lithuania (Figure 6.30). Denmark also has F90-99), with suicides recorded in death registries
high suicide rates, but this may reflect that hospitalised (ICD-10 codes: X60-X84). In cases with several
patients have more severe psychiatric disorders than other admissions during the reference year, the follow-up
countries. Patients with milder psychiatric disorders are period starts from the last discharge.
usually treated in ambulatory settings. For the excess mortality indicators the numerator is
Patients with a psychiatric illness are particularly at risk the overall mortality rate for persons aged between 15
immediately following discharge from hospital. In most and 74 years old diagnosed with schizophrenia or
countries, over one quarter of suicides within the first year bipolar disorder. Most countries use registry data
following discharge occurs in the first month, and in New as a data source. The denominator is the overall
Zealand and Sweden, as many as half of suicides among mortality rate for the general population in the same
patients discharged in the previous year happen in the age group. The relatively small number of people with
first month of discharge. It is known that suicide in the schizophrenia or bipolar disorder dying in any given
high-risk days following discharge can be reduced by good year can cause substantial variations from year-to-
discharge planning and follow-up, and enhanced levels of year, so three-year averages were presented.
care immediately following discharge (OECD, 2014a). The data have been age-sex standardised to the 2010
Individuals with a psychiatric illness have a higher mortality OECD population structure, to remove the effect of
rate than the general population. An “excess mortality” different population structures across countries.
value that is greater than one implies that people with
mental disorders face a higher risk of death than the rest
of the population. Figures 6.31 and 6.32 show the excess References
mortality for schizophrenia and bipolar disorder, which is
above two in most countries. In order to reduce their high OECD (2014a), Making Mental Health Count. The Social
mortality, a multifaceted approach is needed for people and Economic Costs of Neglecting Mental Health Care,
with mental disorders, including primary care prevention OECD Publishing, Paris, http://dx.doi.org/10.1787/
of physical ill health, better integration of physical and 9789264208445-en.
mental health care, behavioural interventions, and changing OECD (2014b), OECD Reviews of Health Care Quality: Norway:
professional attitudes. In view of improving quality of Raising Standards, OECD Publishing, Paris, http://dx.doi.
health care for people with mental disorders, these efforts org/10.1787/9789264208469-en.
120 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Care for people with mental health disorders
6.29. Inpatient suicide amongst patients 6.30. Suicide following hospitalisation
with a psychiatric disorder, 2014 for a psychiatric disorder, within 30 days
(or nearest year) and one year of discharge, 2015 (or nearest year)
Age-sex standardised rate Within one year of discharge
per 1 000 patients Within 30 days of discharge
Age-sex standardised rate per 1 000 patients
9
Costa Rica
Netherlands
Denmark 8
Israel
Belgium 7
Canada
Finland 6
Estonia
Portugal
5
Norway
New Zealand
4
Chile
Slovenia
Spain 3
Switzerland
Slovak Republic 2
Poland
United Kingdom 1
Lithuania
Czech Republic
0
0 1 2 3 4 5
De nia
ic
m
ia
il e
en
Sl el
L i ar k
th ni a
s
nd
an
an
bl
ra
tv
do
Ch
ed
Ne ua
nm
pu
nl
al
La
la
Is
ov
ng
Sw
th
Ze
er
Fi
Re
Note: multiple year average when data available. 95% confidence intervals
Ki
w
h
d
Ne
ec
ite
have been calculated for all countries, represented by grey areas.
Cz
Un
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603887 Note: 95% confidence intervals have been calculated for all countries,
represented by grey areas.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603906
6.31. Excess mortality from schizophrenia, 2014 6.32. Excess mortality from bipolar disorder, 2014
Male Female Male Female
Norway 5.3 6.2 Norway 4.1 4.1
Finland 4.5 6.7 Finland 2.7 4.7
New Zealand 3.5 6.5 New Zealand 3.9 4.3
Korea 4.1 5.4 Korea 4.0 4.3
Sweden 4.1 5.1 Sweden 2.7 2.7
Denmark 4.1 4.3 Denmark 2.7 3.1
Israel 3.2 3.4 Israel 2.2 2.2
Latvia 1.7 2.4 Latvia n.a
Lithuania 1.6 2.3 Lithuania 1.4 1.3
0 2 4 6 8 0 2 4 6 8
Ratio Ratio
Note: Three-year average for all countries. Note: Three-year average for all countries.
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933603925 12 http://dx.doi.org/10.1787/888933603944
Health at a Glance 2017 © OECD 2017 121
6. QUALITY AND OUTCOMES OF CARE
Screening, survival and mortality for breast cancer
Breast cancer is the cancer with both the highest incidence In Iceland the mortality is the highest in the OECD while
and prevalence for women across OECD countries. One in in Korea, it remains the lowest.
nine women will have breast cancer at some point in their
life. Risk factors that increase a person’s chance of getting
this disease include age, family history of breast cancer, Definition and comparability
genetic predisposition, reproductive factors, oestrogen
replacement therapy, and lifestyles including obesity, Screening rates are based on surveys or encounter
physical inactivity, diet and alcohol consumption. data, which may influence the results. Survey-based
Most OECD countries have adopted breast cancer screening results may be affected by recall bias. Programme data
programmes as an effective way for detecting the disease are often calculated for monitoring national screening
early (OECD, 2013). However, due to recent progress in programmes and differences in target population
treatment outcomes and concerns about false-positive and screening frequency may lead to variations in
results, over-diagnosis and overtreatment, breast cancer screening coverage across countries.
screening recommendations have been re-evaluated Five-year net survival is the cumulative probability
in recent years. Taking into account recent research that cancer patients survive their cancer for at least
findings, WHO recommends organised population-based 5 years, after controlling for the risks of death from
mammography screening if women are able to make other causes. Net survival is expressed as a percentage.
an informed decision based on the benefits and risks of Net survival for patients diagnosed during 2000-2004
mammography screening (WHO, 2014). is based on a cohort approach, since all patients had
Screening rates range from less than 20% in Mexico to been followed up for at least 5 years by the end of
over 80% in a few countries including Sweden, Portugal, 2014. For patients diagnosed during 2010-2014, the
Denmark, Finland and Slovenia (Figure 6.33). Screening period approach is used, which allows estimation of
coverage increased substantially among countries with five-year survival, though 5 years of follow-up are not
low rates a decade ago. Mexico had an increase of more available for all patients. Cancer survival estimates
than ten-fold, and Lithuania an almost four-fold increase. are age-standardised with the International Cancer
On the other hand, several countries that had the highest Survival Standard (ICSS) weights.
screening rates in the mid-2000s experienced some Data collection, quality control and analysis were
reductions, including Finland, the Netherlands, and the performed centrally as part of the CONCORD
United States. programme, the global programme for the surveillance
Breast cancer survival reflects early diagnosis as well as of cancer survival, led by the London School of Hygiene
improved treatments. All OECD countries have attained and Tropical Medicine (Allemani et al., 2015). In some
five-year net breast cancer survival of 80% except Chile, countries, not all regional registries participated, but
the Slovak Republic, Poland and Estonia (Figure 6.34). Net survival estimates from the CONCORD programme
survival of people with colon and rectal cancers is also are considered the best available data from those
low in these countries (see indicators on “Survival and countries for international comparisons.
mortality for colorectal cancer”). See indicator “Mortality from cancer” in Chapter 3 for
Over the last decade, the five-year net breast cancer definition, source and methodology underlying cancer
survival has improved in OECD countries. Net survival mortality rates.
has increased considerably in some Central and Eastern
European countries such as Estonia and the Czech
Republic, although survival after breast cancer diagnosis
is still below the OECD average. Improvements may References
be related to strengthening of cancer care governance
Allemani, C. et al. (2015), “Global Surveillance of Cancer
in these countries. For instance, the Czech Republic
Survival 1995-2009: Analysis of Individual Data for
intensified its effort to tackle the burden of breast cancer
25 676 887 Patients from 279 Population-based Registries
through the introduction of a screening programme and
in 67 Countries (CONCORD-2)”, The Lancet, Vol. 385,
a National Cancer Control Programme in the early 2000s
pp. 977-1010, http://dx.doi.org/10.1016/S0140-6736 ( 14)
(OECD, 2014).
62038-9.
With respect to mortality rates, most OECD countries
OECD (2014), OECD Reviews of Health Care Quality: Czech
showed a decline over the past decade (Figure 6.35).
Republic 2014: Raising Standards, OECD Publishing, Paris,
The reduction is a reflection of improvements in early
http://dx.doi.org/10.1787/9789264208605-en.
detection and treatment of breast cancer. Improvements
were substantial in the Czech Republic and Denmark with OECD (2013), Cancer Care: Assuring Quality to Improve
a decline of over 20% in a decade but Denmark still has Survival, OECD Publishing, Paris, http://dx.doi.org/10.1787/
one of the highest rates. On the other hand, within the 9789264181052-en.
OECD, in Iceland and Korea, the mortality rate from breast WHO (2014), “WHO Position Paper on Mammography
cancer increased by more than 10% over the past decade. Screening”, Geneva.
122 Health at a Glance 2017 © OECD 2017
6. QUALITY AND OUTCOMES OF CARE
Screening, survival and mortality for breast cancer
6.33. Mammography screening in women aged 50-69 within the past 2 years, 2005 and 2015 (or nearest years)
2005 2015
% of women screened
100 90.4
90 84.2 83.9 82.7 81.4
79.8 79.5 79.4
75.3 75.1 74.7
80 71.8 70.8
66.8
70 72.7 61.5 60.8 59.6 59.0 59.0 58.6 55.0 54.2
54.5
60 51.9 50.8
47.4 47.0 44.8
50 54.2 41.0
40 35.0 34.9
31.0
30 23.2
18.1
20
10
0
Ze ia²
ng ²
r tu ²
nm l ²
F i r k¹
Un S ia¹
Is ¹
Ne S t a ²
Ire m¹
N e A u d¹
e c K l¹
pu ¹
OE lic¹
Ch ²
M lic¹
la ²
ite No s¹
el ²
Po m¹
o¹
ng ¹
Lu us ly¹
bo ¹
rm ³
Fr y¹
S w sto ¹
er ¹
ua ¹
Ja a¹
La ¹
ov T ia¹
pu ¹
Gr 3 3
ov ¹
Be d¹ ³,
Re e a
m li a
i t z ni a
E ce
il e
K i ay
L i ar y
Re e y
d
Hu a n d
n
Po de n
Sl d
th t es
d in
nd
Ic e c e
ae
Ge rg¹
De ga
nd
an
ni
ic
n
an
an
pa
en
tv
It a
w tr
iu
do
CD
i t e pa
a
b
h or
b
an
d rw
a k ur k
x e tr a
la
an
la
r
ex
s
u
e
lg
nl
al
l
e
Sw
th
er
A
Cz
Un
Sl
1. Programme.
2. Survey.
3. Three-year average.
Source: OECD Health Statistics 2017 and EHIS Eurostat database.
12 http://dx.doi.org/10.1787/888933603963
6.34. Breast cancer five-year net survival, 2000-2004 and 2010-2014
Confidence Interval 2010-14 2000-04 2010-14
Age-standardised net survival (%)
91.2
89.1
100
90.2
88.8
89.5
89.4
88.2
88.0
88.5
87.7
87.6
87.6
86.7
86.6
86.4
86.3
86.2
86.0
86.0
86.1
85.0
85.6
85.4
84.8
83.5
83.2
82.2
82.0
81.4
75.5
66.1
76.6
76.5
75.5
73.5
72.1
70.8
80
82.1
82.1
60
40
20
n
d ca¹
st s
Ja ¹
Ic an
ed ¹
Fi en¹
er ce
a
Is a
N o a e l¹
Ne or ay¹
al ¹
N e F r d¹
lg ¹
S w Ko ¹
e ¹
nm d
y
ng l y
Sp ¹
OE in
Au 31
ov ¹
ic
Ch ¹
La a
Br ¹
de bia
Tu il
C a d¹
e c Ir e e y
pu ¹
Es lic¹
ov Po nia¹
pu ¹
hu le
ia ol ia¹
rm ¹
li a
Be nd s
m
it z rea
ia
m
S l t r ia
ia
Sw nd
Ze al
Re n d
Re nd
G e ark
Au tate
an
di
d
in
az
tio
De r l a n
bl
L i Chi
K i It a
a
an
an
th an
rk
na
en
p
ss C an
tv
w ug
iu
do
CD
i t e Ri
ra
P w
In
b
n om
a
ak la
h la
r
to
ra
s
la
el
nl
r
S
t
Un s t a
Fe
Co
d
ite
Cz
Sl
Un
Ru
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas. Expected updates in the data may reduce the survival
estimate for Costa Rica.
1. Data with 100% coverage of the national population.
Source: CONCORD programme, London School of Hygiene and Tropical Medicine.
12 http://dx.doi.org/10.1787/888933603982
6.35. Breast cancer mortality in women, 2005 and 2015 (or nearest years)
2005 2015
Age standardised rates per 100 000 women
45
40
35 32.5
30 31.4 31.5 32.9
29.1 29.5 29.5 29.9 30.6
25 26.7 26.7 26.9 27.0 27.0 28.1 28.5 28.8
25.0 25.2 24.9 25.3 25.4 25.5 26.1 29.0
20 21.6 22.1 22.2 22.4 22.5
23.3 23.4 23.5
15 18.6 19.7 19.7 21.0
17.6
10 13.2 14.1 15.0
5 8.2
0
lo o
d ica
Ne Belg ia
Un th b l i c
Lu Z e n c e
ng ic
Tu ea
Ja y
M n
a
Sp le
C o Br n
aR l
F i ic a
S w an d
N o de n
r y
ec us al
S o ep li a
Ca tes
ov l
er m
Gr d a
Sw EC ce
E s nd
Po ni a
A u nd
ss L It a
n hu y
de nia
N e Fr ion
bo d
Sl G L a t ¹
ov e v i a
i t e Rep ny
nm s
ng k
Ic ar y
Ir e d ¹
nd
m
er 5
st a zi
Sl r ae
g
ri
bi
Po w a
e
ia it al
De and
Hu ar
Co e x i c
pa
ai
m n
it z D3
en
Ki l
i
Cz A tug
t h iu
do
ur
an
rk
d ub
r
Ch
O e
h tr a
Un ak m a
na
la
la
la
xe a l a
st
ite Afr
m
to
Fe a
a
t
Ko
e
u u
w a
e
nl
Is
ra
r
St
el
l
r
R
Ru
1. Three-year average.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604001
Health at a Glance 2017 © OECD 2017 123
6. QUALITY AND OUTCOMES OF CARE
Survival and mortality for colorectal cancer
Colorectal cancer is the third most commonly diagnosed 100 000 population between 2005 and 2015 (Figure 6.38).
form of cancer after prostate and lung cancers, for men, The decline was particularly large in Austria, the Czech
and the second most common cancer after breast cancer, Republic, Denmark and Israel with a reduction of over
for women, across OECD countries (see indicator “Mortality 30%. Despite some progress, Central and Eastern European
from cancer” in Chapter 3). There are several factors that countries, particularly the Czech Republic, Slovenia and the
place certain individuals at increased risk for the disease, Slovak Republic continue to have higher mortality rates
including age, ulcerative colitis, a personal or family history than other OECD countries.
of colorectal cancer or polyps, and lifestyle factors such as However, in some OECD countries, the mortality rate from
a diet high in fat and low in fibre, lack of physical activity, colorectal cancer increased during the same period. For
obesity, and tobacco and alcohol consumption. Incidence is instance, Hungary which had the highest mortality rate a
significantly higher for men than women across countries. decade ago, reported even higher rates. In Latin American
Generally, rectal cancer is more difficult to cure than colon countries including Chile and Mexico, the increase was
cancer due to a higher probability of spreading to other particularly large, by more than 10%, over the last decade,
tissue, recurrence and postoperative complications. although the rate remains much lower than the OECD
Following screening for breast and cervical cancers, average. Despite increases, some of these countries
colorectal cancer screening has become available, and have made progress in strengthening their systems to
an increasing number of countries have introduced free reduce the burden of colorectal cancer. For example, in
population-based screening, targeting people in their 50s 2013, Chile included treatment for colorectal cancer as
and 60s (OECD, 2013). Partly because of uncertainties about part of its guaranteed health care coverage plan, which
the cost‑effectiveness of screening (Lansdorp-Vogelaar et al., assures improved access, quality, financial protection and
2010), countries are using different methods. In most timeliness of care for priority diseases, and this may lead
countries that provide faecal occult blood test, screening to improved outcomes of colorectal cancer in the future
is available every two years and the screening periodicity (OECD, 2018).
schedule is less frequent with colonoscopy and flexible
sigmoidoscopy, generally every ten years. These differences
make screening coverage difficult to compare across
Definition and comparability
countries.
Advances in diagnosis and treatment of colorectal cancer Net survival and mortality rates are defined in indicator
including improved surgical techniques, radiation therapy “Screening, survival and mortality for breast cancer”
and combined chemotherapy and their wider and timelier in Chapter 6. See indicator “Mortality from cancer”
access have contributed to increased survival over the last in Chapter 3 for definition, source and methodology
decade. In general, OECD countries showed improvement underlying cancer mortality rates. Mortality rates of
in five-year net survival for colon and rectal cancers. On colorectal cancer are based on ICD-10 codes C18-C21
average across OECD countries, five-year colon cancer (colon, rectosigmoid junction, rectum, and anus) while
survival improved from 57.0% to 62.8% for patients with survival estimates are based on C18-C19 for colon
colon cancer between 2000-04 and 2010-14 periods while cancer and C20-C21 for rectum cancer.
survival for rectal cancer also improved from 55.1% to
61.0% during the same periods (Figures 6.36 and 6.37). Some
countries show a considerable improvement including
Chile, Lithuania, Korea, Denmark and Estonia for colon References
cancer, and Latvia, Lithuania, Slovenia, Denmark, Ireland
and Korea for rectal cancer. Generally, countries with Lansdorp-Vogelaar, I., A.B. Knudsen and H. Brenner (2010),
low survival estimates for colon cancer tend to have low “Cost-effectiveness of Colorectal Cancer Screening –An
estimates also for rectal cancer. Among OECD countries, Overview”, Best Practice & Research Clinical Gastroenterology,
net survival estimates are low for both cancers in countries Vol. 24, pp. 439-449.
such as Chile, the Czech Republic, Poland, the Slovak OECD (2018, forthcoming), OECD Reviews of Public Health;
Republic and Turkey. Chile, OECD Publishing, Paris.
In terms of mortality rates, most countries experienced OECD (2013), Cancer Care: Assuring Quality to Improve
a decline in recent years, with the average rate across Survival, OECD Publishing, Paris, http://dx.doi.org/10.1787/
OECD countries falling from 26.8 to 23.9 deaths per 9789264181052-en.
124 Health at a Glance 2017 © OECD 2017
0
5
10
25
30
15
20
35
40
45
50
0
20
40
60
80
100
0
20
40
60
80
So Me
ut xi K Is
ra
h co
A 8.2
A u or e e
Co fric st a¹
ra 71.0 K o l¹ 71.7
lo a Au re
m 12.0 No lia¹
b rw 71.0 st a¹
ra 71.6
Br i a 13.3 ay Ic lia¹
S w Isr ¹ 69.2 el 70.6
Un T a z i an
i t e ur l 13.6 i t z ael d
d k er ¹ 67.9 Ja ¹ 68.2
14.9
la
1. Three-year average.
Sw St ey Be pan
estimate for Costa Rica.
67.3 67.8
it z ate C a nd S w l gi
er s 16.5
na
la i t z um
66.6 67.8
Ne Bel da er ¹
F i nd w gi u
n 17.4 Z m la
Age-standardised net survival (%)
Age-standardised net survival (%)
66.6 67.2
A u lan Ne ea ¹ C a nd
17.8 th lan na
Source: OECD Health Statistics 2017.
C o s tr a d
Canada and for the OECD average.
e r d¹ N da
65.9 67.2
Health at a Glance 2017 © OECD 2017
s t li a la
aR 17.9 nd C o or w
s s
65.3 66.6
Un ta ay¹
17.9 Ja ¹
Gr i c a i t e Ri
ee De pa d ca¹
ce nm n 64.8 65.6
Ch 18.2 St
B e il e S w ar k Sw ates
19.2 64.8 64.9
Age-sex standardised rates per 100 000 population
lg ed ¹ ed
iu
m Fi en¹ Fi en¹
Is 20.1 nl 64.7 nl 64.9
ra an G e and
Ko e l 20.5 Un A u d¹ 64.4 rm ¹ 64.8
ite str an
1. Data with 100% coverage of the national population.
1. Data with 100% coverage of the national population.
Au rea 20.5 d ia¹ y
64.2 64.8
Ne
2005
st St
L F ri
Un a w It
20.6 i t e Ic t e s Ze al y
Un uxe r a n a 64.1 al 64.1
i t e mb c e d ela
K i nd a
20.9
Confidence Interval 2010-14
Confidence Interval 2010-14
d ou ng ¹ A u n d¹
Ki rg d 63.0 st 64.0
ng ¹ 22.2 Ge om r
d rm ¹ 62.5 Fr ia¹ 63.7
C a om a an
22.2
63.7
Ir e n y
G e n ad la 62.2 Ne S ce
rm a 22.6 nd th pa
er in
Ic a ny ¹ 61.7 la 63.3
el 22.6 I n
a OE t al y OE ds¹
S w n d¹ CD 61.3 CD 63.0
22.7
3
Source: CONCORD programme, London School of Hygiene and Tropical Medicine.
Source: CONCORD programme, London School of Hygiene and Tropical Medicine.
ed Sl 3
en Fr 1 61.0 ov 1 62.8
It a 22.8 Sl anc De eni
Ja l y 22.9
ov e 60.9 nm a¹ 61.9
O p Po enia Po ar k
Ne EC an 23.5
r tu ¹ 60.3 r tu ¹ 61.6
th D 3 ga Un g
l
2000-04
i t e Ir a l ¹
2000-04
er 5 59.6 60.9
la 23.9 C o Sp ¹ d ela
nd s t ain K i nd
S s 26.1 aR 59.5
ng ¹ 60.5
Li pa ic do
th in
u 26.4 a Es m¹
Ch ¹ 59.0 to 60.0
De an ni
nm i a 26.7 E s in a a¹
2015
to 56.9 58.4
a n L i Chin
26.8
Ir e r k th a
la L a ia¹ 54.8 ua 57.6
n Cz Li t vi Cz n
La d 27.1 ec L ia¹
53.3 56.9
t ec thu a¹
Ru Es v ia h h at v
ss t 27.2 Re ania Re i a
i a Po on i pu ¹ 52.7 pu ¹ 56.4
n 28.0
bl bl
i
6.36. Colon cancer five-year net survival, 2000-04 and 2010-14
6.37. Rectal cancer five-year net survival, 2000-04 and 2010-14
F e r tu a
de ga Tu ic¹ 52.3 Tu c¹ 56.1
r
6.38. Colorectal cancer mortality, 2005 and 2015 (or nearest years)
ra l 28.2
Sl rk
ke e
2010-14
2010-14
ti ov
ak B y 52.1 B y 54.6
28.2
Po on Re r a z Sl
Cz la ov P r a z i
e c No nd Ru pu il 50.3 ak ol l 53.7
h rw 28.3 ss b Re and
ia P lic¹
n ol 48.6 Ru pu ¹ 52.8
28.5
N e Rep ay F e an ss
w ub bl
d e d¹ ia ic¹
Z e lic
29.6 r 48.4 n 51.7
Sl a Co a tio F e Ch
ov Sl l an d d e il e
ak ov 30.8
lo n
m 41.8 ra 51.5
Re eni bi tio
pu a 30.8
a 38.0 n 44.9
Ch C o In d
Hu bli il e
ng c 38.9 In 32.7 lo ia
m 38.9
ar bi
y 43.9
di
a a
30.0 34.5
estimate for Chile to 43.9, and may also reduce the estimate for Costa Rica. Updates may also lead to very small changes in the survival estimates for
125
12 http://dx.doi.org/10.1787/888933604058
12 http://dx.doi.org/10.1787/888933604039
12 http://dx.doi.org/10.1787/888933604020
Survival and mortality for colorectal cancer
6. QUALITY AND OUTCOMES OF CARE
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas. Expected updates in the data may reduce the survival
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas. Expected updates in the data may reduce the survival
6. QUALITY AND OUTCOMES OF CARE
Survival and mortality for leukaemia in children
Leukaemia is the most common childhood cancer and although a shortage of qualified professionals still exist
accounts for over 30% of all cancers diagnosed in children at specialised centres, quality of care has become similar
aged below 15 years old in the world (IARC, 2012). Causes across providers (OECD, 2018).
of leukaemia are not well known, but some known risk Across OECD countries, the mortality rate of childhood
factors include inherited factors such as Down syndrome leukaemia has also improved over time (La Vecchia
and a family history of leukaemia and non-inherited et al., 2009; Malvezzi et al., 2013) and it was less than
factors including exposure to inonising radiation. There are 1 per 100 000 children in most OECD countries in 2012
different types of leukaemia but about three-quarters of (Figure 6.41). The rate is particularly low at less than 0.3 in
cases among children are acute lymphoblastic leukaemia Australia, and Austria. However, the mortality rate is high
(ALL). The second most frequent type is acute myeloid in Turkey at 3.0 per 100 000 children and Mexico at 2.6.
leukaemia. Prognosis of leukaemia is different depending
on various factors including age, initial white blood cell
count, gender, initial reaction to induction treatment and
type of leukaemia. Children with acute leukaemia who are Definition and comparability
free of the disease for 5 years are considered to have been
cured as remission after 5 years is rare. Incidence and mortality rates come from the
On average across OECD countries, there were 4.7 new International Agency for Research on Cancer (IARC),
cases of leukaemia per 100 000 children aged between GLOBOCAN 2012, available at www.globocan.iarc.fr.
0 and 14 in 2012. Cross-country variations are large and They refer to crude rates and are not age-standardised.
incidence rates in Germany and Finland are high at around 7 GLOBOCAN estimates for 2012 may differ from
per 100 000 children while they are as low as around 3 national estimates due to differences in methods.
in Iceland and Greece. South Africa, India and China also For example, the incidence reported by the German
have low incidence rates, below 3.0 per 100 000 children Centre for Cancer Registry Data (ZfKD) and German
(Figure 6.39). Children’s Cancer Registry is about 5 per 100 000. Net
survival is defined in indicator “Screening, survival
Five-year net survival of acute lymphoblastic leukaemia
and mortality for breast cancer” in Chapter 6.
among children is on average 86.7% during the period of
2010-2014 across OECD countries. Although prognosis of
ALL is considered better among girls than among boys,
the difference in net survival is not statistically significant
for most countries with the exception of Estonia where References
survival for girls is slightly better. Allemani, C. et al. (2014), “Global Surveillance of Cancer
Over time, five-year net survival for children with ALL Survival 1995–2009: Analysis of Individual Data for
has improved across OECD countries (Allemani et al., 25 676 887 Patients from 279 Population-based Registries
2015). This improvement is mainly due to progress in in 67 Countries (CONCORD-2)”, The Lancet, Vol. 385,
chemotherapy and stem cell transplantation technology. pp. 977–1010, http://dx.doi.org/10.1016/S0140-6736(14)
However, countries have not benefited equally from 62038-9.
progress in medical technologies. Survival estimates are IARC (2012), “GLOBOCAN 2012: Estimated Cancer Incidence,
high in Finland (95.2%) and Denmark (94.0%) but they Mortality and Prevalence Worldwide in 2012”, http://
are low in Mexico (52.7%) and Chile (63.9%). Net survival globocan.iarc.fr/Pages/online.aspx.
is low also in China (57.7%), Brazil (66.0%) and Colombia
La Vecchia, C. et al. (2009), “Cancer Mortality in Europe,
(68.9%) (Figure 6.40). In these countries, survival prospect
2000–2004, and an Overview of Trends since 1975”,
of children with ALL may improve through better
Annals of Oncology, Vol. 21, No. 6, pp. 1323–1360, https://
access to effective treatment, by expanding health care
doi.org/10.1093/annonc/mdp530.
coverage and providing high quality care by accredited
professionals at specialised centres. Some of these Malvezzi, M. et al. (2013), “European Cancer Mortality
countries are making progress in improving access Predictions for the Year 2013”, Annals of Oncology, Vol. 24,
and quality of care for childhood cancer. For example, No. 3, pp. 792–800, https://doi.org/10.1093/annonc/mdt010.
Chile included access to care for childhood cancer as OECD (2018, forthcoming), OECD Reviews of Public Health;
part of its guaranteed health care coverage plan and Chile, OECD Publishing, Paris.
126 Health at a Glance 2017 © OECD 2017
So
0
1
2
3
0
20
40
60
80
100
Au Fi ut
n h
0
1
2
3
4
5
6
7
8
Au str De lan Af
s ia 0.2 ric
C z S tr a nm d¹ 95.2
e c lo li a a In a 1.0
h ve 0.3 d
94.0
C a r k¹
Re ni Un na Ch i a 2.3
p a 0.3 i t e Ic d a Ic in a
Ge ubl 92.6 el 2.6
rm ic d ela
K n a
0.4 Gr n d
Hu an N e i n g d¹ 92.4 ee 2.9
Per 100 000 children
ng y 0.4 w do
Per 100 000 children
Ze m 3.1
estimate for Costa Rica.
Ja ce
Gr a r y al ¹ 92.2 Cz
e 0.4 e c P pan
Ge and h ola 3.3
No ec e rm ¹ 91.4 Re n
rw 0.5 Be an d 3.3
In p u b
Age-standardised net survival (%)
C a ay
na 0.5 lg y 91.1 do lic
A u ium ne 3.4
Fr d a 90.8 s
Health at a Glance 2017 © OECD 2017
an 0.6 Ne st ¹
t h r a li Br i a 3.4
Ir e c e az
Un la 0.6 Sw erla a¹ 90.7 Fr il 3.5
ite n i t z nd s a
d Ja d
K i pa 0.6 er ¹ 90.4
Es nce
t 3.7
ng n P lan Hu o ni
Sl D d 0.6 U n o r t d¹ 90.3 ng a 3.8
ov e om ite ug ar
ak nm 0.6 d a l¹ y 3.9
Un Re ar k St 89.8 C o Isr a
0.7 lo el
i t e pub
d lic
Sw a tes
89.5 m 3.9
ed
1. Data with 100% coverage of the national population
St 0.7 A u bia
at Fr en¹ st 4.0
e an 89.0 r
Is s 0.7 Cz L ia 4.1
ra
88.6
el ec Ire ce¹ De a t v
Ne 0.8 h la nm ia
t It Re nd 4.1
Ne her al y pu ¹ 88.3 Ne ar
w lan 0.8 bl w Sp k 4.2
Ze d s i Ru Ze ain
al 0.8 Is c¹ 88.2 ss a 4.2
a ra
e l¹ ia B lan
87.9
Po nd n el d
0.8 I F e gi 4.3
Po lan Es t al y d e um
r tu d 0.8 to 87.8 ra 4.4
g Sl ni Sw tion
Source: International Agency for Research on Cancer (IARC), GLOBOCAN 2012.
Source: International Agency for Research on Cancer (IARC), GLOBOCAN 2012.
ov
Confidence Interval 2010-14
Sp al 0.8 ak Ja a¹ 87.7 ed 4.5
S
Source: CONCORD programme, London School of Hygiene and Tropical Medicine.
S o w ain Re pa Tu en 4.7
ut ed 0.8 pu n 87.6 rk
h en
A b Ne OEC ey
0.8 Po lic¹ 4.7
OE fric la 87.0 th D 3
er 4
CD a 0.8 O E n d¹ l 4.7
Es 3 4 CD 86.9 S and
to 0.8 3 Un S w love s 4.9
n Sp 1 86.7 i t e i t z nia
Sw L a ia 0.9 ai
d erl
K i an 4.9
it z t vi Ko n 84.7 ng d
4.9
er a
la 0.9 r d
M om
84.4
F i nd Gr ea¹
nl 0.9 ee Sl L e x 4.9
B e an d c ov i t i c
ak hu o
l 1.0 L a e¹ 84.2 5.0
L i gi u tv Re ani
2010-14
th m N ia pu a 5.1
ua 1.2 C o or w ¹ 84.1 bl
n Un
1.2
s t ay
83.0
i t e It ic 5.2
6.41 Leukaemia mortality in children aged 0-14, 2012
Ko i a Ru aR ¹
6.39. Leukaemia incidence in children aged 0-14, 2012
re ss d al
Ru
ia ic Co Sta y 5.3
ss Ch a 1.3 n Tu a¹ 81.1
ia il Fe rk st tes
n aR 5.3
F e In e 1.4 de ey
r 80.9 ic
d e di Li atio Ko a 5.5
1.4
6.40. Acute lymphoblastic leukaemia five-year net survival, 2010-14
Co r at a th n r
ua 76.9 Ir e e a
st ion
l 5.6
aR 1.4 ni
a¹ No and
ic 74.7 5.7
a 1.5 C o In d Au r wa
C o Br a lo ia
lo z il m 69.9 st y
r 5.7
m 1.6 bi C a a li a
bi
a Br a 68.9 na 5.8
In C h i 2.2 az da
do na i 66.0 5.8
ne 2.3 Ch l Po Chil
M si a il r tu e 5.8
ex 2.4 Ch e 63.9 F i gal
i M in a n 5.9
Tu co 2.6 ex 57.7 Ge lan
rk rm d
ey ic
o¹ an 6.5
3.0 52.7 y 7.0
127
12 http://dx.doi.org/10.1787/888933604115
12 http://dx.doi.org/10.1787/888933604096
12 http://dx.doi.org/10.1787/888933604077
Survival and mortality for leukaemia in children
6. QUALITY AND OUTCOMES OF CARE
Note: 95% confidence intervals have been calculated for all countries, represented by grey areas. Expected updates in the data may reduce the survival
6. QUALITY AND OUTCOMES OF CARE
Vaccinations
All OECD countries have established vaccination 43%. Large decreases can be seen in Germany, Slovenia,
programmes based on their interpretation of the risks and Italy. Some countries did show increased vaccination
and benefits of each vaccine. For children, vaccination over this time period including Mexico, Israel, the United
rates for diphtheria, tetanus and pertussis (DTP), measles, States, Portugal, Denmark, Greece, and New Zealand. Only
and hepatitis B at age 1 are high across OECD countries two countries attained the 75% target: Mexico and Korea,
(Figures 6.42 and 6.43). On average, over 95% of children with the United Kingdom coming close to meeting the
receive the recommended DTP or measles vaccinations, target.
while almost 94% receive a recommended hepatitis B
vaccination. Vaccination rates for DTP are below 90%
in Indonesia, Mexico, and India. Vaccination rates for
Definition and comparability
measles are below 90% in Italy, Indonesia, and India while
vaccination rates for hepatitis B are below 90% in Mexico, Vaccination rates reflect the percentage of children
France, Indonesia, India, and Germany. that receives the respective vaccination in the
Overall rates of vaccination among children are increasing. recommended timeframe. The age of complete
Between 2005 and 2015, vaccination rates among children immunisation differs across countries due to
have increased 1 percentage point for DTP vaccination, different immunisation schedules. For those countries
more than 2 percentage points for measles, and nearly recommending the first dose of a vaccine after age
12 percentage points for hepatitis B among OECD countries. one, the indicator is calculated as the proportion of
Large increases in hepatitis B vaccination can be seen over children less than two years of age who have received
this period in a number of OECD countries including France that vaccine. Thus, these indicators are based on the
and the Netherlands, reflecting the introduction of national actual policy in a given country.
programmes. However, vaccination rates have dropped Some countries administer combination vaccines
in recent years in some countries, notably for measles (e.g. DTP for diphtheria, tetanus and pertussis) while
coverage in Australia and Italy. Even small decreases in others administer the vaccinations separately. Some
vaccination can result in large increases in disease cases countries ascertain vaccinations based on surveys and
(Lo et al. 2017). While national vaccination coverage rates others based on encounter data, which may influence
are high, some populations remain under-covered. A 2015 the results.
outbreak of measles in the United States was caused by a
Influenza vaccination rates refer to the number of
number of unvaccinated individuals, while in Europe 1020
people aged 65 and older who have received an annual
cases of measles were reported between February 2016 and
influenza vaccination, divided by the total number of
January 2017 in Italy alone. (CDC, 2017; ECDC, 2017).
people over 65 years of age. In some countries, the
Not all countries follow WHO recommendations to data are for people over 60 years of age. The main
incorporate hepatitis B into national immunisation limitation in terms of data comparability arises from
programmes, including Denmark, Finland, Sweden, and the use of different data sources, whether survey or
the United Kingdom, where vaccination is not part of the programme, which are susceptible to different types of
general infant vaccination programme, but is provided to errors and biases. For example, data from population
high-risk groups. Other OECD countries that do not include surveys may reflect some variation due to recall errors
vaccination against hepatitis B in their infant programmes and irregularity of administration.
are Iceland, Hungary, Japan, Slovenia and Switzerland. In
Canada, the Hepatitis B immunisation schedule varies by
jurisdiction.
Influenza is a common infectious disease responsible for References
3 to 5 million severe cases worldwide, including 250 000
CDC – Centers for Disease Control and Prevention (2017),
to 500 000 deaths. Hospitalisation and death occur mainly
“Measles Cases and Outbreaks”, available at: http://www.
among high-risk groups and in industrialised countries
cdc.gov/measles/cases-outbreaks.html, accessed 24/06/2017.
most deaths associated with influenza occur among people
age 65 or older (WHO, 2016). Safe and effective vaccination ECDC – European Centre for Disease Prevention and
is available for influenza and most countries recommend Control (2017), “Surveillance Report: Measles and Rubella
annual vaccination among older adults. Monitoring”, April 2015.
In 2003, countries participating in the World Health Lo, N.C. and P.J. Hotez PJ. (2017), Public Health and Economic
Assembly committed to the goal of attaining vaccination Consequences of Vaccine Hesitancy for Measles in the
coverage against influenza among the elderly of at least United States”, JAMA Pediatrics, 7 July, http://dx.doi.
75% by 2010. Figure 6.44 shows vaccination among adults org/10.1001/jamapediatrics.2017.1695.
over 65 for 2005 and 2015. Over this period, the average WHO (2016), “Influenza (Seasonal)”, Fact Sheet No. 211,
vaccination rate against influenza among the elderly available at: http://www.who.int/mediacentre/factsheets/
population decreased among OECD countries from 49% to fs211/en/, accessed 24/06/2017.
128 Health at a Glance 2017 © OECD 2017
0
10
60
90
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
70
80
100
80
82
84
86
88
90
92
94
96
98
100
Es Lu
to Lu Ch xe Ch
ni
a
xe in m in a
b
La 2 m a Cz
bo e c Hu ou r
99
tv u h ng g
% vaccinated
Tu ia 2 Au rg Re a
rk st pu r y
e r ia b
Po y 9 ¹ M lic¹
la Ko e
99 98
Sl S n Sw xico
10
1. All data estimated.
ov lo d Po r e a
98
a ed
% of children vaccinated
% of children vaccinated
ve
1. 2015 data estimated.
2. 2005 data estimated.
1. 2015 data estimated.
r tu e
C z k Re nia
10
K n
98
e c pu Be gal Ru
h
2. Measles data estimated.
Po or e
Ru lg ss
Re blic iu r tu a
pu 14 ss m ia
98
b ia I n Is g a l
n sr Fe r
16
F e ae de ae
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
A u lic
r a l¹
97
d e l¹
Health at a Glance 2017 © OECD 2017
st ra t
H u r ia
20 tio Tu ion
ng n rk
97
a Sp So G ey
21 a
No r y ut ree
rw in h c
97
Cz
Ge ay ec Tur
rm 27 h Ge Afr e
rm ica
a Re ke y
97
Ic ny pu
37 bl A u a ny
el
a ic s ²
97
D n B e t r ia
40
Br
2005
2005
Lu enm d az lg ¹
Sl il iu
96
xe ar
m k 42
ov
ak Ch Sp m
bo
u Re le i a
96
Ja in
Fi rg
nl 42 pu pa
an bl Br n
O
96
d 43
Po ic az
Sw ECD la Ch il
96
it z 3 5 Gr n d
43
er ee Po il e
la ce la
nd L nd
96
46 Ir e ² Sl
ov O a t v
It a la
Li nd Un ak E C D i a
95
Gr l y 49 th i t e Re 3 5
ee ua
Diphtheria, tetanus and pertussis
ni d pu
49 a K i bl
94
Sw ce
L ng ic
ed
en
Ne at d
th via No om
49
94
Ja er rw
la
(or nearest year)
Po p an n N e F in ay
50
94
r tu OE d s th la
g CD S w erl nd
Fr al 50 Au 3 5
94
i t z and
an st er s
ce ra
51 li a Sl lan
Ch 93 ov d
i So I Li en
th ia
Ir e l e 55 ut t al
h y
93
la A Co uan
nd lo ia
Sp ¹ 55
Co fric m
93
st a Ir e b i a
Ne a R
B e ain
56 w ic l
2015
2015
lg I an
92
Measles
iu Un Z e a a Co c el d
Ca m 58 ite lan Ne st a and
na d d w R
92
da St Ze ic a
N e Is 62 Co a t es a
r
92
th ae lo Un E l a n
66 m ite sto d
Ne erla l b d ni
S a
91
w nd Es ia
Un Z e a s 67 to Au tate
6.43. Percent of children aged 1 vaccinated for hepatitis B, 2005 and 2015
G e ni a s s
91
Un i t e d l a n rm D e tr a l
ite S d 67 an nm i a
d tat y
88
K i es Fr ar k
ng 69 In an
6.44. Percent of population aged 65 and over vaccinated for influenza, 2005 and 2015
d In d i
87
A u om do a Ca ce¹
st 71 ne na
ra s da
84
li a Fr i a
78
In In d
Ko an do ia
re ce ne
M ¹
83
M a 82 si
a
ex ex
ic ic It a
o
6.42. Percent of children aged 1 vaccinated for diphtheria, tetanus and pertussis (DTP) and measles, 2015
o ly
82
82
129
12 http://dx.doi.org/10.1787/888933604172
12 http://dx.doi.org/10.1787/888933604153
12 http://dx.doi.org/10.1787/888933604134
Vaccinations
6. QUALITY AND OUTCOMES OF CARE
7. HEALTH EXPENDITURE
Health expenditure per capita
Health expenditure in relation to GDP
Financing of health care
Sources of health care financing
Health expenditure by type of service
Health expenditure by provider
Capital expenditure in the health sector
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.
Health at a Glance 2017 © OECD 2017 131
7. HEALTH EXPENDITURE
Health expenditure per capita
The financial resources that a country devotes to health Only four countries – Iceland, Hungary, Switzerland and
care, both for individuals and for the population as a whole, Chile – have recorded higher average growth in the period
and how this changes over time is the result of a wide array since 2009 compared to the period before. Indeed, health
of social and economic factors, as well as the financing spending in Hungary together with Poland and Estonia has
and organisational structures of a country’s health system. remained relatively resilient since 2009 with annual growth
In 2016, the United States is estimated to have outspent of between 2.7-3.6%
all other OECD countries by a wide margin, spending the Away from Europe, Korea and Chile have continued to report
equivalent of USD 9 892 for each resident (Figure 7.1). This annual health spending increases above 5% in real terms
level of health spending is almost two-and-a-half times since 2009. Preliminary country estimates for 2016 suggest
the average of the 35 OECD countries (USD 4 003) and further strong spending growth of 6.3% in Korea and 4.5%
25% above Switzerland, the next highest spender (adjusted in Chile. In the United States, health spending grew by 4.1%
for the different purchasing powers – see box “Definition in real terms in 2015, the fastest rate for more than ten
and comparability”). Compared with the other G7 countries, years, with a preliminary estimate by the OECD suggesting
the United States spends almost 80% more than Germany a further increase of 2.7% in 2016. In the medium-term, the
and more than twice as much on health care per person as US Centers for Medicare & Medicaid Services (CMS) expect
Canada, France and Japan. OECD countries spending half or health spending growth above that of GDP in the United
less of the OECD average include many of the Central and States, driven on by faster growing medical prices.
Eastern European members of the OECD, such as Estonia
and Poland, together with Chile. Lowest per capita spenders
on health in the OECD were Mexico and Turkey with levels
around a quarter of the OECD average, and similar to Definition and comparability
spending in key emerging economies such as the Russian
Expenditure on health measures the final consumption
Federation, South Africa and Brazil. China spent around
of health goods and services (i.e. current health
20% of the OECD per capita spending level, while both India
expenditure). This includes spending by both public
and Indonesia spent less than 10% of the OECD average
and private sources on medical services and goods,
based on latest available figures.
public health and prevention programmes and
Figure 7.1 also shows the split of health spending based administration.
on whether it is paid from government sources or some
To compare spending levels between countries,
kind of compulsory insurance, or through voluntary means
per capita health expenditures are converted to a
such as voluntary health insurance or direct payments
common currency (US dollar) and adjusted to take
(see indicator on “Financing of health care”). In general,
account of the different purchasing power of the
the ranking of per capita expenditure of government and
national currencies, in order to compare spending
compulsory schemes is comparable to that of total spending.
levels. Economy-wide (GDP) PPPs are used as the
Even if voluntary insurance in the United States continues
most available and reliable conversion rates. For the
to play a significant role in financing health care, the level
calculation of growth rates in real terms, economy-
of spending from federal and state programmes (such as
wide GDP deflators are used for all countries. In some
Medicaid) and Medicare is still greater on a per capita basis
countries (e.g. France and Norway), health specific
in the United States than in most other OECD countries, with
deflators exist, based on national methodologies, but
the exceptions being Luxembourg, Norway and Switzerland.
these are not used in this publication due to limited
Per capita spending on health across the OECD continued comparability.
to grow in 2016 following the trend of recent years. This
Note that data for 2016 are based on preliminary
comes after the abrupt slowdown in health spending
figures either provided by the country or estimates
growth between 2009 and 2011 in the wake of the global
made by OECD Secretariat.
financial and economic crisis. On average, annual health
spending growth across the OECD since 2009 has been
1.4% compared with 3.6% in the six years up to 2009 (Figure 7.2).
In a number of countries there have been significant References
turnarounds in annual growth rates in health spending in
the years before, compared with after the financial crisis. Morgan, D., M. Gmeinder and J. Wilkens (2017), “An OECD
In Greece, strong annual growth increases were reversed analysis of health spending in Norway”, OECD Health
after 2009 (5.4% vs. -5.0%). A similar if less dramatic picture Working Papers, No. 91, OECD Publishing, Paris, http://
is also observed in Portugal (2.2% vs. -1.3%). In general, dx.doi.org/10.1787/63302bbf-en.
health spending growth slowed down in the vast majority OECD/Eurostat/WHO (2017), A System of Health Accounts
of OECD countries and preliminary figures or estimations 2011: Revised edition, OECD Publishing, Paris, http://dx.doi.
for 2016 still point to negative or near-zero growth in a few. org/10.1787/9789264270985-en.
132 Health at a Glance 2017 © OECD 2017
Un
-6
-4
-2
0
2
4
6
8
10
12
14
%
Gr ite
0
2 000
4 000
6 000
8 000
10 000
ee Sw d S
c 5.4 ta
USD PPP
Po
-5.0
r tu e Lu i t z e t e s
ga xe r l a 9 892
l 2.2 m nd
It a -1.3 bo 7 919
Cz l 1.6
No ur g
7 463
e c Ir e y
h lan -0.3 Ge r wa
Re d 6.9 rm y 6 647
pu 0.1
2. CPI used as deflator.
Ir e a n y
bl 5 551
4.9
2. Includes investments.
ic la
Sp 0.1 Ne Sw nd
5 528
th ed
De ain 3.4 er en
Ne nm 0.3 la 5 488
w
Ze ar k 3.4 Au nd s
5 385
Source: OECD Health Statistics 2017.
al 0.3 De s t r
Health at a Glance 2017 © OECD 2017
Sl and 4.5 n m ia 5 227
0.6
ov B e ar
Un en lg k 5 205
3.2
i t e Fr ia
0.7 C ium
d an Au ana 4 840
K i ce
ng 1.6 st da
do 0.9 ra 4 753
m 3.9 Fr lia¹ 4 708
Lu S w e 0.9 an
xe de
m n 1.8 Un Ja ce 4 600
bo 0.9 i t e Ic p a
1. Mainland Norway GDP price index used as deflator.
u d ela n 4 519
Au rg 1.8 Ki n
2003-09
st 0.9 ng d²
4 376
d
r ia 2.1 F i om
N e F inl 1.0 nl 4 192
th an
er d 3.4 Ne O E an d
w CD 4 033
la 1.0
Note: Expenditure excludes investments, unless otherwise stated.
n Ze 3 5
Be d s 3.2 al 4 003
lg 1.0 an
d
iu
Government/Compulsory
3 590
Ic m 2.7 It a
el 1.0 3 391
a Sp l y
Ca nd 0.4 Sl ai
1.0
ov n
e 3 248
Source: OECD Health Statistics 2017, WHO Global Health Expenditure Database.
na
M da 3.1 Is n i a 2 835
ex 1.1 Po r ae
OE ico 1.3 Cz r tu l ² 2 822
CD 1.1 ec g
3 h Ko a l 2 734
Tu 5 3.6 Re r e
rk 1.4 Sl pu a 2 729
e 4.6 ov b
Ja y 1.7
a k Gr l i c 2 544
Re e e c
2.9
Ge pan p e
rm 2 223
1.8
Sl a Hu ubli
o v N o ny 2.2 ng c 2 150
ak r w 1.8 Es ar y
2009-16
a 2.1 to 2 101
Un Rep y¹ ni
i t e ub 1.8 a 1 989
L i Ch
d lic 12.2 t h il e
St
at 1.9 ua 1 977
es 2.5 Po nia
Is 2.1 Ru la 1 970
ra
7.1. Health expenditure per capita, 2016 (or nearest year)
e 2.3 1 798
Voluntary/Out-of-pocket
ss Co L a nd
2.2
La l ia st t v
n a ia
Fe R 1 466
Hu t v i a 6.5
ng 2.4 S o d er i c a ²
1 390
u t a ti
Au ar y -0.4 h on
1. Australian expenditure estimates exclude all expenditure for residential aged care facilities in welfare (social) services.
S w s tr 2.7 Af 1 351
r
i t z a li a 2.7 Tu ica
er 2.7 r 1 149
la
n 1.4
M ke y
ex 1 088
Po d
la 2.8 ic
n 1 080
6.6
Es d C o Br a o
to 3.1 lo z il ²
m 995
ni
a 5.6 bi
Ko 3.6 In C h i a ² 964
re do na
a 8.5 ne ² 733
Ch 5.7 si
il e
² 5.5 In a ² 302
5.9 di
a²
7.2. Annual average growth rate in per capita health expenditure, real terms, 2003 to 2016 (or nearest year)
269
133
12 http://dx.doi.org/10.1787/888933604210
12 http://dx.doi.org/10.1787/888933604191
Health expenditure per capita
7. HEALTH EXPENDITURE
7. HEALTH EXPENDITURE
Health expenditure in relation to GDP
How much a country spends on health care over time and increased health coverage for the population. In 2003,
relative to spending on all other goods and services in the health spending in Korea accounted for only 4.3% whereas
economy can be down to both growth in health spending in 2016 it was estimated to have reached 7.2%. At the other
itself as well as how well the economy is performing overall. end of the scale, no discernible impact can be seen for
In 2016, health spending is estimated to have accounted Mexico which has seen its health spending to GDP ratio
for 9.0% of GDP on average across OECD countries, largely remain relatively constant throughout the period at around
unchanged in recent years. This comes after a period of 6% of GDP.
health spending growth above that of the overall economy In Europe, Germany has seen its health spending to GDP
in the 1990s and 2000s that saw health expenditure as a ratio stabilise since 2009 as health spending growth has
share of GDP rise sharply in many OECD countries. aligned with economic growth with a slow but steady
In 2016, the United States spent 17.2% of GDP on health, increase to reach 11.3% in 2016, almost one percentage
almost five percentage points above Switzerland, the next point above the level in 2003. Greece, on the other hand,
highest country, and more than eight percentage points where there have been significant cuts in health spending
above the OECD average (Figure 7.3). A group of ten high- since 2009, has seen the health spending to GDP ratio
income OECD countries, including Germany, France, Japan fluctuate – approaching close to 10% in 2010 – before
and Canada, follow with around 11% of GDP going on returning to a similar level to that in the early 2000s at
health services. Another large group of countries spanning around 8% of GDP.
Europe, as well as Australia and New Zealand (and South
Africa) fit roughly within a band of between 8-10% of GDP. A
similar sized group of mainly Central and Eastern European
countries, such as Hungary, the Czech Republic and Poland Definition and comparability
allocate between 6-8% of their GDP to health. Only Mexico,
Latvia and, notably Turkey at 4.3%, spend less than 6% of See indicator on “Health expenditure per capita” for a
GDP on health services. Turkey’s health spending as a share definition of expenditure on health.
of GDP is at a similar level to that in India. Gross Domestic Product (GDP) = final consumption +
Looking at changes over time, the average health spending gross capital formation + net exports. Final
to GDP ratio jumped sharply in 2009 as overall economic consumption of households includes goods and
conditions deteriorated rapidly in many countries while services used by households or the community
health spending growth was sustained at around 3% on to satisfy their individual needs. It includes final
average in 2008 and 2009 (Figure 7.4). While subsequent consumption expenditure of households, general
health spending growth also significantly declined – government and non-profit institutions serving
approaching zero growth on average in 2010/11 – this step households.
increase in the health spending to GDP ratio has been In countries, such as Ireland and Luxembourg, where a
largely maintained as the rate of health spending growth significant proportion of GDP refers to profits exported
has tended to closely track the growth in the overall and not available for national consumption, GNI may
economy since 2012. be a more meaningful measure than GDP.
However, behind the overall OECD average, some different Note that data for 2016 are based on preliminary
patterns emerge on a country by country basis. In the figures provided by the country or estimates made by
United States, after a number of years (2009-14) when the OECD Secretariat.
ratio of health spending to GDP has been stable at around
16.4%, 2015 and 2016 have seen this increase again to reach
the 17.2% in 2016 (Figure 7.5). This mirrors the period before
the economic crisis when health spending rose almost a
References
percentage point between 2003 and 2008. Korea has seen the
most notable increase in the share of economic resources OECD/Eurostat/WHO (2017), A System of Health Accounts
allocated to health over time with a significant progression 2011: Revised edition, OECD Publishing, Paris, http://dx.doi.
in the ratio over many years on the back of growing wealth org/10.1787/9789264270985-en.
134 Health at a Glance 2017 © OECD 2017
7. HEALTH EXPENDITURE
Health expenditure in relation to GDP
7.3. Health expenditure as a share of GDP, 2016 (or nearest year)
Government/Compulsory Voluntary/Out-of-pocket
% GDP
18
17.2
16
14
12.4
12
11.3
11.0
11.0
10.9
10.6
10.5
10.5
10.4
10.4
10.4
9.7
9.6
10
9.3
9.2
9.1
9.0
9.0
8.9
8.9
8.8
8.6
8.6
8.5
8.3
7.8
7.7
7.6
8
7.4
7.3
7.2
6.9
6.7
6.5
6.4
6.3
6.2
5.8
5.7
5.6
6
5.5
4.8
4.3
4
2.8
2
n L ico
Sl C epu l ²
Ja ce
ak om ic
Es blic
e s
rm d
S w a ny
ov ²
Ch a
Ch on
In a ²
In Tur ²
ne y
a²
Fr en
Ne C a an
er da
No nd s
l y
nm m
d us k
n ia
st m
Ne Fi lia¹
Co Zea nd
aR d
CD ²
Sp 5
r tu n
ut It al
Af y
el ²
Gr il e
Ir e e c e
a
e c I ar y
pu ²
th ia
Lu P ni a
m nd
Br r g
e x il
ra a
K d
Ic r i c a
Sl and
a
Re bia
OE ic a
i
it z ate
Be wa
do ke
de vi
Hu or e
i t e A ar
h al
az
Ge rlan
Po ai
st lan
n
3
R e
ov ol bl
Ki tr
en
Li ton
g
D e gi u
in
di
A u gdo
si
u
an
ed
th na
p
ti
h sr a
xe ola
w nla
la
Fe at
ra
ua
g
bo
e
la
r
Sw d St
M
ite
So
Un
ia
Cz
Un
ss
Ru
Note: Expenditure excludes investments, unless otherwise stated.
1. Australian expenditure estimates exclude all expenditure for residential aged care facilities in welfare (social) services.
2. Includes investments.
Source: OECD Health Statistics 2017, WHO Global Health Expenditure Database.
12 http://dx.doi.org/10.1787/888933604229
7.4. Average annual growth in per capita 7.5. Health expenditure as a share of GDP, selected
health expenditure and GDP, 2003-16 OECD countries, 2003-16
(OECD average)
OECD35 Greece Korea
Health GDP Germany Mexico United States
% % GDP
6 18
16
4
14
2
12
0
10
-2
8
-4
6
-6 4
4
14
16
12
03
05
07
09
11
13
15
/1
/0
/0
/0
/
/
20
20
20
20
20
20
20
11
09
07
15
13
05
03
20
20
20
20
20
20
20
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604267
12 http://dx.doi.org/10.1787/888933604248
Health at a Glance 2017 © OECD 2017 135
7. HEALTH EXPENDITURE
Financing of health care
Health care can be paid for through a variety of financing the share of health spending payable by households has
arrangements. In some countries, health care might be increased since 2009 due to the implementation of reforms
predominantly covered by government schemes by which to balance public budgets which shifted some financing
individuals are automatically entitled to care based on their responsibilities to patients. On the other hand, this share
residency. In other cases, compulsory health insurance has been reduced in Mexico (-6.0 pp) and Chile (-2.3 pp)
schemes (either through public or private entities) over the same time period.
finance the bulk of health spending. In addition to these,
a varying proportion of health care spending consists of
payments by households (either as standalone payments
or as part of co-payment arrangements) as well as various Definition and comparability
forms of voluntary health insurance intended to replace,
Health care financing can be analysed from the point
complement or supplement automatic or compulsory
of view of financing schemes (financing arrangements
coverage.
through which health services are paid for and
In all but one OECD country, government schemes and obtained by people, e.g. social health insurance),
compulsory health insurance constitute the main health financing agents (organisations managing the financing
care financing arrangements. Together they accounted, schemes, e.g. social insurance agency), and types of
on average, for almost three-quarters of all health care revenues (e.g. social insurance contributions). Here
spending across the OECD in 2015 (Figure 7.6). In Denmark, “financing” is used in the sense of financing schemes
Sweden and the United Kingdom, central, regional or local as defined in the System of Health Accounts (OECD,
government financed 80% or more of all health spending. Eurostat and WHO, 2011) and includes government
In Germany, Japan, France and the Slovak Republic more schemes, compulsory health insurance as well as
than 75% of all health expenditure was paid for through voluntary health insurance and private funds such
compulsory health insurance. Only in the United States as households’ out-of-pocket payments, NGOs and
was less than half of all health spending financed by private corporations. Compulsory health insurance can
government or compulsory health insurance. By contrast, be offered by private insurers, in some cases without
a large proportion of health spending (35%) was paid for an obligation to contract individuals (e.g. in Chile and
via voluntary health insurance. Germany). Out-of-pocket payments are expenditures
Governments provide a multitude of public services out of borne directly by patients and include cost-sharing
their overall budgets. Hence, health care is competing with arrangements and any informal payments to health
many other sectors such as education, defence and housing. care providers.
The size of public funds allocated to health is determined Total government expenditure is as defined in the
by a number of factors including, among others, the type of System of National Accounts and includes intermediate
system in place and the demographic composition of the consumption, compensation of employees, interest,
population. Relative budget priorities may also shift from social benefits, social transfers in kind, subsidies, other
year to year as a result of political decision-making and current expenditure and capital expenditure payable by
economic effects. In 2015, health spending by government central, regional and local governments as well as social
schemes and compulsory insurance stood at around 15% of security funds. Relating spending from government
total government expenditure across the OECD (Figure 7.7). financing schemes and compulsory insurance schemes
In Japan, Switzerland, New Zealand, the United States and to total government expenditure is overestimated to
Germany more than 20% of public spending was dedicated a certain extent for those countries with compulsory
to health care. On the other hand, less than one out of health insurance provided by private insurers.
every ten euros spent by governments or compulsory
Spending by private health insurance companies in
health insurance was allocated to health care in Latvia
the United States are considered under voluntary
and Greece.
health insurance although the Affordable Care Act
After government schemes and compulsory health (ACA) constitutes a mandate for individuals to buy
insurance, the main source of funding tends to be out- health insurance or pay a penalty since 2014.
of-pocket payments. On average across the OECD, private
households directly financed around one-fifth of all health
spending in 2015. This share is above a third of health
spending in Greece (35%), Korea (37%), Mexico (41%) and References
Latvia (42%), while in France it is below 10%. With the
implementation of universal health coverage in some OECD Mueller, M. and D. Morgan (2017), “New Insights into
countries over previous decades, there have been some Health Financing: First Results of the International Data
significant reductions in the share of health care costs Collection Under the System of Health Accounts 2011
payable by households. More recently, the share of out-of- Framework”, Health Policy, Vol. 121, No. 7, pp. 764–769.
pocket spending has been generally stable but with some OECD/Eurostat/WHO (2017), A System of Health Accounts
notable increases in some European countries (Figure 7.8). 2011: Revised edition, OECD Publishing, Paris, http://dx.doi.
In Greece (+6.2 percentage points) and Spain (+4.7 pp) org/10.1787/9789264270985-en.
136 Health at a Glance 2017 © OECD 2017
7. HEALTH EXPENDITURE
Financing of health care
7.6. Health expenditure by type of financing, 2015 (or nearest year)
Government schemes Compulsory health insurance Out-of-pocket
% Voluntary health insurance Other
100 1 2 2 1 6 6 3 5 5 2 3 4 5 2 5 7 4 1 6 5
14 5 14 6 13 4
13 14 13 15 15 17 18 15 12 10 11
11 12 15 13 17 18 23 18 23 20
7 20 24 23 29
80 11 13 15 15 20 28 28 32 35 42 35
9 23 37 41
29 13 5 1
1
60
45 36 11
70 56 59
78 75 73 71 29
84 84 75 75 65 42
40 74 80 75 69 61 56 46 28 23
71 66 69 70 67 65 46
61 59 57
52
20 36
31 30 27
22 18 22 24
12 11 11 17
7 9 9 9 4 3 9 2 10
0 4
d ico
ia
m ic
N e ingd c
Ca n
Un M a
²
Ge way
D e a ny
k
C z S an
R de n
I rg
Sl th n d
i t e ep s
Fr d
Tu ¹
Be ey
Es m
Au ia
r ia
Fi ly
OE nd
Sl 3 5
Ir e a
nd
Au and
Hu li a
Po ar y
i t z gal
nd
el
il e
La e
ia
om
ce
es
li
re
Un k R n d
ar
ai
an
xe bl
en
n
ra
tv
It a
iu
u
rk
ee
d ub
Ch
ra
na
p
la
a
la
a
st
CD
an
Sp
to
S w r tu
ite ex
at
Ko
nm
ng
bo
rm
Lu epu
ec we
nl
l
Ja
Ne c el
al
a la
lg
Is
r
ov
Po
er
st
Gr
No
St
Ze
ov er
K
w
h
1. France does not include out-of-pocket payments for inpatient LTC thus resulting in an underestimation of the out-of-pocket share.
2. Spending by private health insurance companies in the United States is reported under voluntary health insurance.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604286
7.7. Health spending by government schemes and compulsory health insurance as share of total government
expenditure, 2015 (or nearest year)
%
25 23.2 22.8
22.3 22.1
21.4
19.6 19.1
20 18.4 18.4 18.3 18.3
17.5 17.0
16.4 15.8
15.3 15.3 15.1 15.1 14.9
15 14.3
13.4 12.7
12.5 12.3 12.3 12.2 12.0 12.0
11.4 11.3 10.7
10.1 9.7
10 8.9 8.9
Po o
ic
ia
OE ce
m ic
Ne erl n
U n Z e a d¹
St d
rm s
Ne C ¹
la ¹
ite Ire s¹
K i and
C a en
No da
A u way
Ic li a
De and
Fr k
Sw m
Au 5
B e t r ia
m
g
Re in
Sl l y
Fi a
Po an d
ov Es al
Lu epu a
M el
Tu d
Hu ke y
La y
Gr i a
ce
t h hil e
y
Ge a te
R ni
re
ar
ar
ic
it z pa
ite lan
n
ur
3
bl
en
xe bl
ra
g
It a
tv
an
nd
h pa
iu
do
w an
an
ee
ra
ed
na
la
CD
ak to
r tu
ex
nm
Ko
ng
s
bo
r
pu
l
nl
Sw Ja
el
lg
Is
r
ov
ec S
st
ng
er
d
Cz
Un
Sl
Note: Relating spending from government and compulsory insurance to total government expenditure may lead to an overestimation in countries
where compulsory insurance is provided by private insurers.
1. Includes spending by private health insurers for compulsory insurance.
Source: OECD Health Statistics 2017, OECD National Accounts Database.
12 http://dx.doi.org/10.1787/888933604305
7.8. Change in out-of-pocket expenditure as a share of expenditure on health,
2009 to 2015 (or nearest years)
% 2009 (or nearest year) 2015 (or nearest year)
50
41.6
40 41.4 35.5
36.8 29.0 28.3 27.7
30 32.2 24.2 23.3 22.8 22.8
20.2 19.9 19.6
17.9
20 23.2 22.7 17.0 16.9
15.3 15.2
17.6 12.3
10.6
10 14.6 14.3 13.7 13.1
12.6 12.5 12.5 11.4 11.1 10.6
6.8
0
o
ce
ia
Gr a
ce
S w lic
er a
Hu le
Po ic
i t z ar y
P o a nd
l
Re ain
nd
Es l y
a
OE el
Fi 5
A u an d
m
A u li a
B e r ia
Ic m
ec Tu d
Re y
C a en
No da
De ay
Ne Ja k
Ze n
Ge and
S a ny
U n Ir e s
d
ite b s
Ki rg
ga
th ni
re
ni
h r ke
Un xem a t e
ar
nd
ic
w pa
an
ite lan
3
bl
i
ra
tv
It a
iu
do
d ou
an
ee
Ch
rw
ra
ed
na
la
st
CD
Ne love
ak Sp
ex
r tu
to
Ko
nm
S w ng
rm
pu
pu
l
nl
La
el
al
la
Is
lg
Lu S t
er
st
ng
Fr
M
d
ov
Cz
Sl
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604324
Health at a Glance 2017 © OECD 2017 137
7. HEALTH EXPENDITURE
Sources of health care financing
In all OECD countries, the various schemes that pay for the of revenues in more detail. In Poland, employees bear
health care goods and services rely on a mix of different the brunt of social contributions, whereas in Estonia the
sources of revenues. Government schemes, for example, financing responsibility falls on employers.
typically receive budget allocations out of the overall Some countries are planning to reduce their reliance on
government revenues (e.g. from income and corporate wage-based contributions in the face of shrinking labour
taxation, value-added tax, etc.). Social health insurance markets and financial shocks, and are increasingly looking
is usually financed out of social contributions payable by for ways to diversify their revenue base (OECD, 2015). While
employees and employers. However, these schemes may there is little year-to-year change in the health financing
also receive a varying proportion of their revenues from structure and composition of revenues, some trends can
governmental transfers. The main sources of revenue be discerned over a longer time horizon (Figure 7.11). In
for private health insurance are either compulsory or Belgium, for example, the share of social contributions in
voluntary prepayments, which typically take the form all revenues has fallen from over 50% to around 43% over
of regular premium payments as part of an insurance the last decade. At the same time, governmental transfers
contract. Out-of-pocket payments are exclusively financed have gained importance. The latter is also true for the United
from households’ own revenues. Some health financing States where the share from government transfers increased
schemes (e.g. non-profit or enterprise schemes) may also from 34% to 41% over the same time period. In Korea, on the
receive donations or additional income from investments other hand, government transfers have stagnated while the
or rental. Resident financing schemes can also receive share through social contributions has increased.
transfers from abroad as part of bilateral co-operations
with foreign governments or other development partners.
However, these transfers play no role in the vast majority
of OECD countries. Definition and comparability
The composition of revenues is strongly correlated with a Health financing schemes have to raise revenues in
country’s system of health care financing. Hence, when order to pay for health care goods and service for the
analysing the overall revenue structure in, say, Denmark population they are covering. There are different types
– where health care activities are predominantly financed of revenues which can however be closely correlated
through local government schemes (see indicator on with the financing scheme. In general, financing
“Financing of health care”) – governmental transfers are schemes can receive transfers from the government,
the most important revenue (Figure 7.9). Comparing the social insurance contributions, voluntary or
structure of financing schemes with the types of revenues compulsory prepayments (e.g. insurance premiums),
that these schemes receive can give important insights into other domestic revenues and revenues from abroad
how financing works in different health systems: in many as part of development aid.
countries, the government’s role is typically larger than as just
In reality, the revenues of a health financing scheme
a simple purchaser of health services (Mueller and Morgan,
are typically not identical to its expenses in a given
2017). In Japan, for example, the government is directly
year leading to a surplus or deficit of funds. In practice,
responsible for only 9% of all health spending but government
most countries only analyse the composition of
transfers to the different schemes existing in the country
revenues per scheme and apply the resulting shares
constitute 42% of all revenues for health care financing.
on a pro-rata basis to the expense of each financing
The role governments play as a financing source can scheme thus equating revenues with its expenses.
be highlighted more clearly when only analysing the
composition of revenues for compulsory health insurance,
which in most OECD countries consists of social health
insurance (SHI) (Figure 7.10). In the countries analysed, References
governmental transfers are a source of revenue in each
case but the importance differs significantly. In Japan, more Mueller, M. and D. Morgan (2017), “New Insights into
than 40% of the revenues of SHI stems from governmental Health Financing: First Results of the International Data
transfers. The shares are similar in Chile and Finland but Collection Under the System of Health Accounts 2011
account for less than 5% in Estonia, Poland and Slovenia. Framework”, Health Policy, Vol. 121, No. 7, pp. 764–769.
In those countries, SHI funds finance their outlays nearly OECD (2015), Fiscal Sustainability of Health Systems: Bridging
exclusively via social contributions. Yet, even here, Health and Finance Perspectives, OECD Publishing, Paris,
substantial variations exist when analysing this stream http://dx.doi.org/10.1787/9789264233386-en.
138 Health at a Glance 2017 © OECD 2017
7. HEALTH EXPENDITURE
Sources of health care financing
7.9. Health financing sources by type of revenue, 2015 (or nearest year)
Transfers from government Social insurance contributions Compulsory prepayment
domestic revenues Voluntary prepayment Other domestic revenues
%
100
14 14 16 13 18 19 16 17 14 16
20 24 21 24 25
2 2 5 3 33 36
80 2 10 15 43
10 6 5 5
1 3 34
42 43 7 6
60 46 25
35 11 1 5
84 62 10
40 84 81 80 59 20
72 66 39 28
67 6
20 42 39 45 41
36 30 30
18 24
13 11
0 5
o
ia
k
en
da
17
nd
nd
il e
es
ni
re
ar
ic
pa
ur
an
an
en
iu
do
at
Ch
ed
na
CD
la
la
to
ex
nm
Ko
bo
nl
Ja
el
lg
ov
St
er
Po
ng
Sw
Ca
Es
M
Ic
OE
Fi
Be
m
it z
De
Sl
d
Ki
xe
ite
Sw
d
Lu
Un
ite
Un
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604343
7.10. Financing sources of compulsory insurance by type of revenue, selected countries, 2015 (or nearest year)
Transfers from government SIC from employees SIC from employers
% SIC from self-employed SIC from others Other
100 3 3 1 1 1
10 10
2 19 20 14 15
1
80 15 4 42 5
53
15 35
60 30 36 39
95
23 86
40
57 34
47 45 44 38
20 35
15
0 4 3 2
United States Chile Japan Finland Belgium Korea Estonia Slovenia Poland
Note: SIC stands for social insurance contributions. “Other” includes compulsory prepayment and other domestic revenues.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604362
7.11. Share of government transfers and social insurance contributions in all revenues
of financing schemes, selected countries, 2003-15
% Government transfers Social insurance contributions
100
80
60 43
51
10
40 36 39 11
20 39 34 41
28
19 18
0
2003 2015 2003 2015 2003 2015
Korea Belgium United States
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604381
Health at a Glance 2017 © OECD 2017 139
7. HEALTH EXPENDITURE
Health expenditure by type of service
How health spending is split between the various services outlays for vaccinations in many countries around 2009
and goods reflects a variety of factors, from disease burden (Gmeinder et al., forthcoming).
and system priorities to organisational aspects and costs. While spending on inpatient, outpatient and long-term
Spending on inpatient and outpatient care combined care has continued to grow, the rates have also significantly
accounts for the major part of health expenditure across reduced since 2009. Expenditure growth for outpatient care
OECD countries – almost two-thirds of health spending nearly halved overall (4% vs 2.3%), but remained positive
on average in 2015 (Figure 7.12). A further 19% of health in the majority of OECD countries. Some governments
spending was accounted for by medical goods (mainly decided to protect expenditure for primary care and front-
pharmaceuticals), while 14% went on long-term care line services while looking for cuts elsewhere in the health
services. The remaining 6% was spent on prevention and system. The annual average growth rate for inpatient care
public health services as well as on the overall governance dropped to almost half of its previous growth rate, down
and administration of the health system. from 2%, and turned negative between 2009 and 2015 in
Greece has a particularly high share of spending on inpatient around one-quarter of OECD countries. Reducing wages
care (including day care in hospitals) – accounting for 40% in public hospitals, postponing staff replacement and
of its health spending in 2015. Inpatient care also plays delaying investment in hospital infrastructure were among
an important role in Poland, Austria and France, taking the most frequent measures taken in OECD countries to
up more than a third of total spending. Countries with a balance health budgets.
high share of outpatient spending include Portugal (48%)
and Israel (47%). The United States also consistently reports
one of the highest shares of outpatient care. However, this
includes physicians’ fees in cases where they independently Definition and comparability
bill patients for hospital care.
The third major category of health spending is on medical The System of Health Accounts (OECD, Eurostat and
goods. Variations can be due to a number of factors such as WHO, 2017) defines the boundaries of the health care
the different distribution channels in place, the extent of system from a functional perspective, with health care
generic use as well as the relative prices of pharmaceuticals. functions referring to the different types of health
In the Slovak Republic (35%) and Hungary (32%), medical care services and goods. Current health expenditure
goods represent the largest component of health spending. comprises personal health care (curative care,
The share is also high in Latvia, Mexico and Greece, at rehabilitative care, long-term care, ancillary services
around 30%. In Denmark, Luxembourg and Norway, on the and medical goods) and collective services (prevention
other hand, spending on medical goods represents only and public health services as well as administration –
10-11% of health spending. referring to governance and administration of the
overall health system rather than at the health provider
There are also differences between countries in the amount
level). Curative, rehabilitative and long-term care can
of health expenditure on long-term care services (see
also be classified by mode of provision (inpatient, day
Chapter 11). Norway, Sweden and the Netherlands, with
care, outpatient and home care).
their established formal arrangements for the elderly and
the dependent population, allocate more than a quarter of Concerning long-term care, only the health aspect
all health spending to long-term care. Whereas in many is reported as health expenditure, although it is
Southern European and Central and Eastern European difficult in certain countries to separate out clearly
countries with more informal long-term care sectors, the health and social aspects of long-term care. Thus,
spending on long-term care services accounts for a much estimations of long-term care expenditure continue
smaller share. to be one of the main factors limiting comparability
across countries.
The slowdown in health spending experienced in many
OECD countries following the economic crisis affected For the calculation of growth rates in real terms,
all parts of the health sector, but to varying degrees economy-wide GDP deflators are used.
(Figure 7.13). Expenditure for pharmaceuticals contracted
annually by 0.5% after positive annual increases of 2.3%
during the pre-crisis years and even stronger growth in
the 1990s and early 2000s. Despite initially protecting References
public health budgets, prevention spending growth also
turned negative in around half of OECD countries after Gmeinder, M., D. Morgan and M. Mueller (2017, forthcoming),
2009. On average, spending on preventive care contracted “How Much Do OECD Countries Spend on Prevention?”,
by 0.2% on an annual basis, after recording very high OECD Health Working Paper, OECD Publishing, Paris.
growth rates during the period 2003-09 (4.6%). Part of OECD, Eurostat and WHO (2017), A System of Health Accounts
the reversal in spending growth can be explained by the 2011: Revised edition, OECD Publishing, Paris, http://dx.doi.
H1N1 influenza epidemic, which led to significant one-off org/10.1787/9789264270985-en.
140 Health at a Glance 2017 © OECD 2017
7. HEALTH EXPENDITURE
Health expenditure by type of service
7.12. Health expenditure by type of service, 2015 (or nearest year)
Inpatient care* Outpatient care** Long-term care
% Medical goods Collective services
100 4 4 3
5 5 5 4 5 6 5 6 6 6 4 5 5 5 7 4 6 8 5 5 7
8 10 8 9 9 10 9
16 10 11
14 12
20 14 15
17
11
20 14 16
13
21 23 22 21 14
28 22 20 19 30 20 20 16 23
80 32 20
3 8 14 30 35
6 17 20 24 28
6 15 24 26
9 10
2 5 10 11 14 13
18 19 23 16 19 24 25
5 14 15
60 4
25
48 46 30
41 31 28
38 38 32 30 28 33 33
34 34 30 33 30 30 30 29 29 26 28 24
40 47 34 33 34 28
33
20 40
36 34 33
32 30 31
26 26 27 26 26 28 27 26 26 29 26 29 28 29 29 28 31 28 30
26
22 22 22 21
17
0 C z Re i c o
Ic ia
H blic
OE ce
Fr d
De v ia
ov M r k
Re blic
Is l
Es el
Po ia
Un G d
St e
¹
n
Fi ly
Au d
S l r ia
Un xem ar y
Ki rg
ay
Ir e n
Ge and
i t z ny
S w nd
Ca n
Be da
m
th rea
s
31
No m
ga
es
c
nd
e
n
ai
an
pa
an
en
ra
It a
iu
do
a
d ou
an
rw
ite ree
ed
Sw rma
na
la
la
st
CD
at
t
Sp
ak e x
to
r tu
nm
N e Ko
Lu ung
pu
e c pu
nl
el
Ja
l
La
la
lg
ov
er
ng
ite b
er
Po
h
Sl
Note: Countries are ranked by curative-rehabilitative care as a share of current expenditure on health.
* Refers to curative-rehabilitative care in inpatient and day care settings. ** Includes home care and ancillary services.
1. Inpatient services provided by independent billing physicians are included in outpatient care for the United States.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604400
7.13. Growth rates of health expenditure per capita for selected services, OECD average, 2003-15
2003-09 2009-15
Annual growth rate in real terms (%)
7
6.0
6
5 4.6
4.0
4
3.2
3
2.3 2.3 2.2
2.0
2
1.3
1
0
-0.2 -0.1
-0.5
-1
-2
Inpatient care Outpatient care Long-term care Pharmaceuticals Prevention Administration
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604419
Health at a Glance 2017 © OECD 2017 141
7. HEALTH EXPENDITURE
Health expenditure by provider
Across OECD countries, the delivery of health care services Many countries have seen a growing share of health spending
and goods takes place in many different organisational going to hospitals in recent years while at the same time
settings, ranging from hospitals and medical practices to there has been a tendency to shift medical services from
pharmacies and even private households caring for family inpatient to day care settings (see indicator on “Ambulatory
members. A breakdown by provider allows the tracking of surgery” in Chapter 9). The main motivation behind this
health expenditure from an organisational point of view, a is the generation of efficiency gains and a reduction of
useful complement to the functional breakdown of health waiting times. Moreover, for some interventions day
expenditure (see indicator “Health expenditure by type of care procedures are now the most appropriate treatment
service”). method. Hence, in a number of countries day care now
While the way in which health care provision is organised accounts for more than 10% of all hospital expenditure.
across OECD countries varies considerably, hospitals are Furthermore, the provision of long-term care in hospital
the main health care provider in terms of health spending makes up a sizeable share of hospital expenditure in some
(Figure 7.14). They account for nearly 40% of overall health countries (e.g. Korea, Japan and Iceland).
spending on average and represent the main spending
category for all but a handful of countries. In Turkey, Estonia
and Italy around half of all health spending is accounted
for by activities delivered in hospitals. On the other hand, Definition and comparability
hospitals in Canada, Germany and Mexico account for 30%
or less of health spending. The universe of health care providers is defined in
the System of Health Accounts (OECD, Eurostat and
Ambulatory providers are the second main category with
WHO, 2017) and encompasses primary providers, i.e.
regard to health spending. Overall, around one-quarter of
organisations and actors that deliver health care
health spending relates to ambulatory providers, ranging
goods and services as their primary activity, as well
from more than 50% in Israel to 20% or less in Ireland, the
as secondary providers for which health care provision
Slovak Republic, the Netherlands and Turkey. The category
is only one among a number of activities.
covers a wide range of facilities and depending on the country-
specific organisational set up, most spending relates either The main categories of primary providers are hospitals
to medical practices including offices of GPs and specialists (acute and psychiatric), residential long-term care
(e.g. Austria, France and Germany) or ambulatory health facilities, ambulatory providers (practices of GPs and
care centres (e.g. Finland, Ireland and Sweden). On average, specialists, dental practices, ambulatory health care
practices of GPs and specialists together with ambulatory centres, providers of home health care services),
health care centres account for around two-thirds of all providers of ancillary services (e.g. ambulance services,
spending on ambulatory providers. Around one-fifth of laboratories), retailers (e.g. pharmacies), and providers
ambulatory provider spending relates to dental practices of preventive care (e.g. public health institutes).
and about 10% to providers of home health care services. Secondary providers include residential care
Other main provider categories include retailers (mainly institutions whose main activities might be the
pharmacies selling prescription and over-the-counter provision of accommodation but provide nursing
medicines) and residential long-term care facilities (mainly supervision as secondary activity, supermarkets that
providing inpatient care to long-term dependent people). sell over-the-counter medicines, or facilities that
The activities performed by providers classified within provide health care services to a restricted group of the
the same category can differ widely across countries. population such as prison health services. Secondary
This variation is particularly pronounced in hospitals providers also include providers of health care system
(Figure 7.15). Although inpatient curative and rehabilitative administration and financing (e.g. government
care accounts for the vast majority of hospital expenditure agencies, health insurance agencies) and households
in almost all OECD countries, hospitals are also important as providers of home health care.
providers of outpatient care in most countries, for
example through accident and emergency departments,
hospital-based specialist outpatient units, or laboratory
and imaging services provided to outpatients. In Sweden,
References
Estonia, Finland and Portugal outpatient care accounts for
over 40% of hospital expenditure. On the other hand, in OECD, Eurostat and WHO (2017), A System of Health Accounts
Greece, Germany and Belgium, less than 10% of hospital 2011: Revised edition, OECD Publishing, Paris, http://dx.doi.
expenditure goes on outpatient care. org/10.1787/9789264270985-en.
142 Health at a Glance 2017 © OECD 2017
7. HEALTH EXPENDITURE
Health expenditure by provider
7.14. Health expenditure by provider, 2015 (or nearest year)
Hospitals LTC facilities Ambulatory providers
% Retailers Other
100
6 5 7 7 6 5 6 5 5 5
8 9 9 8 8 9 9 10 8
12 13 11 11 4 12 12 13 12 11 11 12
14 16 14
10 10
17 14 11 15 9 12
21 17 18 11 22 12
20 22 18 13 10
80 27 28 17 19 20 19
17 14
32 23
29 30
30 27 14 35
25 24
23 54 30 30
23 28 20
22 36 29
60 13 29 24 22 23 23 22
26
27
22 30 28 31
24 27
4 6 10 16 22
12 16 18 23
2 6 7 7 9 19
1 2 7 8 15 16 17
8 32
40 10 16
1 1 1 5
3 11
9
53
47 45 44 43 42 42 42 41
20 41 41 41 41 40 40 39 38 38 38 36 36 36 35 35 35 34 34 33 32 31 30 29
26
0
Sl blic
Ja a
o
Ic e
d lic
Es ey
a
De l y
Gr a
i t e or e
ng al
Re in
No an
Fr y
Be tes
m
m
Au d
S w r ia
OE en
m ia
Ca rg
Hu ael
G e ad a
M ny
Po r y
th 32
Fi d
d
Sl w i t lan d
Un ep d
i
a
c
ni
re
Un P e e c
ar
nd
R n
ic
an
n
an
en
d tug
It a
xe t v
h pa
iu
do
u
rk
rw
an
i t e ub
ed
a
p
ak r l a
la
st
Ne CD
r
to
ex
a
nm
Ko
ng
bo
rm
pu
n
el
nl
Lu L a
la
lg
Is
ov
St
Tu
ec S
S Ir e
ov z e
er
Ki
Cz
Note: Countries are ranked by hospitals as a share of current expenditure on health.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604438
7.15. Hospital expenditure by type of service, 2015 (or nearest year)
Inpatient care Day care Outpatient care*
% Long-term care Other
100 1 2 3 3 2 2 3 2 3 1 4
1 1 1 1
6 4 6 2 7 7 1 5 3 3
9 3 3 9 4 4 6 8
2 5
1 14 15 1
1 11
11 20 8
4 15 21 13
2 2 15 26 22 31 18
23 29
80 10 34 42 34 28 26
5 44
22 42 42
8 16 24
9
7 11 17
4 4
9
1 12
60 2 5
5 11 17
23
5 5
94 93 93 15
86 84 83 1
40
74 73
69 67 66 66 65 64 63 63 61 61 59 57 57
54
50 49 49 48
42
20 40
0
ia
nd
ic
ce
ic
th en
Es s
Fi a
Ca d
da
Po e a
r ia
ce
l
y
Sw m
27
ay
ia
li a
n
nd
ly
ga
an
ni
nd
ar
an
an
pa
ai
ur
bl
en
bl
tv
It a
iu
do
ee
an
rw
r
ra
ed
na
la
la
st
CD
to
Sp
r tu
Ko
ng
bo
rm
pu
pu
nl
el
Ja
la
La
lg
ov
Po
Ir e
st
ng
Au
Gr
Fr
No
Ic
er
OE
Be
Hu
m
Re
Re
Au
Ge
Sl
Ki
xe
h
ak
Ne
d
Lu
ec
ite
ov
Cz
Un
Sl
Note: Countries are ranked by inpatient curative-rehabilitative care as a share of hospital expenditure.
*Includes ancillary services.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604457
Health at a Glance 2017 © OECD 2017 143
7. HEALTH EXPENDITURE
Capital expenditure in the health sector
Although health systems remain a highly labour-intensive health care facilities) and investment programmes on
sector, capital has been an increasingly important factor new equipment (e.g., medical and ICT equipment) are
of production of health services over recent decades, implemented. Decisions on capital spending also tend to be
as reflected for example by the growing importance of more affected by economic cycles, with spending on health
diagnostic and therapeutic equipment or the expansion system infrastructure and equipment often a prime target
of information and communications technology (ICT) in for reduction or postponement in economic downturns.
health care (see previous indicator on eHealth adoption Figure 7.17 shows that a number of hard-hit European
in general practice and hospital). However, the level of countries have seen annual investment in the health sector
resources invested in infrastructure, equipment and ICT fall in real terms post-crisis. Greece, in particular, reported
tends to fluctuate more with economic cycles than current capital spending in the health sector at around a third of
spending on health services, as investment decisions are the level reported ten years before. Portugal and Italy have
often more discrete and can more easily be postponed or both seen investment drop by 30% or more from the peaks
brought forward depending on economic circumstances. In in 2010. The United Kingdom is also notable in seeing a
making capital investment decisions, policy-makers need significant reduction in investment: up to 2009, capital
to carefully assess not only the short-term costs, but also spending was increasing rapidly year-on-year whereas
the potential benefits in the short, medium and longer- between 2011 and 2014 it was back to 2003 levels.
term. Slowing down investment in health infrastructure Outside of Europe a number of countries reported a
and equipment may also reduce the capacity to treat continual increase in capital expenditure. Korea and Japan
patients and contribute to increases in waiting times for have seen recent investment in the health care sector
different types of services. around 50% higher, in real terms, than the levels of ten
In 2016, OECD countries allocated, on average, around 0.5% years earlier (Figure 7.18).
of their GDP for capital expenditure in the health sector
(Figure 7.16). This compares with the 9% of GDP going on
current spending, that is on medical care, pharmaceuticals,
Definition and comparability
etc. (see indicator “Health spending as a share of GDP”).
As is the case with current spending, there are significant Gross fixed capital formation in the health sector is
differences in the current levels of investment expenditure measured by the total value of the fixed assets that
between countries and in the recent trends observed health providers have acquired during the accounting
following the economic crisis. period (less the value of the disposals of assets) and
As a proportion of GDP, Japan was the highest spender that are used repeatedly or continuously for more
on capital investment in 2015 with more than 1% of its than one year in the production of health services.
GDP going on construction, equipment and technology The breakdown by assets includes infrastructure (e.g.
in the health and social sector. A number of European hospitals, clinics, etc.), machinery and equipment
countries – Belgium, Austria and Germany – were also (including diagnostic and surgical machinery,
relatively high capital spenders in 2015, with between 0.7- ambulances, and ICT equipment), as well as software
0.8% of GDP invested. For the most part, OECD countries and databases.
find themselves within a relatively narrow band of Gross fixed capital formation is reported by many
between 0.4-0.6% of GDP each year. However, either due countries under the System of Health Accounts.
to the economic conditions or the peculiarities of a small It is also reported under the National Accounts
economy (Luxembourg and Iceland) capital spending can broken down by industrial sector according to the
be significantly lower. Greece, for example, spent just under International Standard Industrial Classification (ISIC)
0.15% of its GDP on capital investment in the health sector Rev. 4 using Section Q: Human health and social work
in 2015. activities or Division 86: Human health activities. The
By its very nature, capital spending fluctuates from year former is normally broader than the SHA boundary
to year more than current spending as capital projects while the latter is narrower.
on construction (i.e. building of hospitals and other
144 Health at a Glance 2017 © OECD 2017
7. HEALTH EXPENDITURE
Capital expenditure in the health sector
7.16. Gross fixed capital formation in the health care sector as a share of GDP, 2015 (or nearest year)
% GDP
1.2
1.1
1.0
0.8
0.8
0.7
0.7
0.7
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.4
0.4
0.4
0.4
0.4
0.4
0.3
0.3
0.3
0.3
0.2
0.2
0.1
0.1
0.1
0.1
0
o
ia
Be an²
St e
m lic
Ic o n
Au ¹
rm a
Sp ¹
De ain ²
La k
Es a²
Ko a
Un Fr ²
Au tes
Hu li a
ec Sw ¹
Re en
Ze ic¹
r tu ²
O E a l¹
34
er y
Ca s¹
Fi a
Po d
Ir e ²
Sl nd
ng y¹
M d¹
il e
Lu ep l
ia Gr g¹
de e¹
m
R ae
m
Po n d
a
nd
Ge tri
ite nc
ni
th wa
d
ar
ic
an
en
an
nd
F e eec
l
ar
i
re
do
ur
an
xe ub
Ch
ra
w bl
ed
na
a
g
ti
la
tv
iu
CD
ak Isr
to
ex
a
nm
p
la
a
t
nl
ra
ng
N e N or
ov
bo
Ne pu
st
la
I
lg
el
Ja
al
d
Ki
d
h
ite
n
ov
Un
Cz
Sl
ss
Ru
1. Refers to gross fixed capital formation in ISIC 86: Human health activities (ISIC Rev. 4).
2. Refers to gross fixed capital formation in ISIC Q: Human health and social work activities (ISIC Rev. 4).
Source: OECD Health Statistics 2017, OECD National Accounts.
12 http://dx.doi.org/10.1787/888933604476
7.17. Gross fixed capital formation, constant prices, 7.18. Gross fixed capital formation, constant prices,
selected European OECD countries, 2003-15 selected non-European OECD countries, 2003-15
Greece Italy Canada Japan
Portugal United Kingdom Korea United States
Index (2003 = 100) Index (2003 = 100)
200 200
150 150
100 100
50 50
0 0
2003 2005 2007 2009 2011 2013 2015 2003 2005 2007 2009 2011 2013 2015
Source: OECD Health Statistics 2017, OECD National Accounts. Source: OECD Health Statistics 2017, OECD National Accounts.
12 http://dx.doi.org/10.1787/888933604495 12 http://dx.doi.org/10.1787/888933604514
Health at a Glance 2017 © OECD 2017 145
8. HEALTH WORKFORCE
Health and social care workforce
Doctors (overall number)
Doctors by age, sex and category
Medical graduates
Remuneration of doctors (general practitioners and specialists)
Nurses
Nursing graduates
Remuneration of nurses
Foreign-trained doctors and nurses
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.
Health at a Glance 2017 © OECD 2017 147
8. HEALTH WORKFORCE
Health and social care workforce
Employment in health and social care represents a large technologies will raise expectations on the quality and
and growing share of the labour force in many countries scope of care (OECD, 2015).
across the world (UN High-Level Commission on Health Many countries have also started to introduce new care
Employment and Economic Growth, 2016). In the OECD, delivery models that will involve greater integration of
health and social work activities constituted around 10% health and social services in order to meet the needs of
of total employment on average in 2015 (Figure 8.1). The ageing societies. These changes are expanding the roles
employment share is particularly pronounced in the of non-physician providers (such as nurse practitioners
Scandinavian countries, Finland and the Netherlands, and pharmacists and community health workers) into
where jobs in health and social work represent 15-20% of health care, aimed at maintaining access to services and
these countries’ workforces. increasing the productivity of the health workforce, as
Moreover, the percentage of workers employed in health well as improving the continuity and quality of care for
and social work has steadily risen across much of the OECD the patients. These changes will likely lead to significant
between 2000 and 2015. For the OECD overall, there was an transformations in staffing profile and skills requirements
average percentage point increase of 1.7 from 2000 to 2015. in the health and social care sector.
Some of the greatest increases have taken place in Japan,
Ireland, Korea, Luxembourg and the Netherlands. Four
countries experienced a decrease in share of employment Definition and comparability
in health and social work: Latvia, Mexico, Poland and the
Slovak Republic. Health and Social Work is one of the economic
The rapid employment growth in health and social care activities defined according to the major divisions of
contrasts markedly with the experience in other sectors the International Standard Industrial Classification
(Figure 8.2). Across the OECD, employment in health of All Economic Activities (ISIC). Health and Social
and social work grew on average by 42% (with a median Work is a sub-component of the Services sector, and
value of 34%) between 2000 and 2015. Over the same is defined as a composite of human health activities,
time period, there was an overall decline in the number residential care activities (including long-term care),
of jobs in agriculture and industry in the OECD countries. and social work activities without accommodation.
Employment growth in health and social work was also The employment data are taken from the OECD
noticeably higher than employment growth in the services National Accounts (SNA) database for the 35 OECD
sector, and was significantly above the growth in total member countries, except for Iceland and Turkey
employment. where the source is the OECD Annual Labour Force
Statistics (ALFS) database.
Past and current experiences show that employment in
the health and social sector tends to be less sensitive to
cyclical fluctuations than employment in other sectors in
the economy. While the total employment declined slightly References
in the United States during the economic recessions of the
OECD (2016), Health Workforce Policies in OECD Countries:
early 1990s and significantly in 2008-09, employment in the
Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
health and social sector continued to grow steadily over
http://dx.doi.org/10.1787/9789264239517-en.
this same period. In most OECD countries, the number of
doctors and nurses continued to rise through the recession OECD (2015), Fiscal Sustainability of Health Systems: Bridging
period (see indicators on doctors and nurses). Health and Finance Perspectives, OECD Publishing, Paris,
http://dx.doi.org/10.1787/9789264233386-en.
Looking forward, employment in health and social care
sector is likely to increase, but the type of skills and OECD (2011), Help Wanted?: Providing and Paying for
functions are expected to change. This reflects a number Long-Term Care, OECD Publishing, Paris, http://dx.doi.
of factors. Ageing populations will change the pattern of org/10.1787/9789264097759-en.
demand for health and social services. This could include UN High-Level Commission on Health Employment and
greater demand for long-term care and related social Economic Growth (2016), Working for Health and Growth:
services, which are particularly labour-intensive (OECD, Investing in the Health Workforce, Geneva, WHO. Retrieved
2011). Over time, rising incomes and availability of new from http://www.who.int/hrh/com-heeg/reports.
148 Health at a Glance 2017 © OECD 2017
8. HEALTH WORKFORCE
Health and social care workforce
8.1. Employment in health and social work as a share of total employment, 2000 and 2015 (or nearest year)
% 2000 2015
25
20.4
20
17.9
16.7
15.7
15.6
14.3
13.3
15
13.0
13.0
12.9
12.8
12.4
12.4
12.2
11.4
10.8
10.4
10.4
10.3
10.2
10.1
10
7.9
7.6
7.0
6.8
6.8
6.3
6.2
6.2
5.4
5.3
4.0
5
5.9
5.8
2.7
5.2
0
o
d ce
Re enia
Es lic
Gr l i c
nm y
Ne we k
er n
Au nd
Is a
Fi ds
Un F n d
Ir e m
N e Ic e n d
Ze d
Lu C a el
m da
OE urg
Po D 3 5
S w s tr s
i t z a li a
B e a nd
G e ium
i t e J ny
ng n
l
ly
Hu pain
K y
e c lov a
ov P ni a
Re and
ce
il e
Tu ia
M ey
ga
ri
De w a
Au ate
C z S or e
S ar
ar
ic
t h de
Ki a
w lan
ra
It a
tv
do
rk
ite r an
b
ee
Ch
xe n a
a
d ap
n
a
la
a
st
to
r tu
ex
ng
bo
rm
pu
pu
l
ak ol
nl
al
La
la
lg
r
C
St
S
er
No
h
Un
Sl
Source: OECD National Accounts; OECD Annual Labour Force Statistics for Iceland and Turkey.
12 http://dx.doi.org/10.1787/888933604533
8.2. Employment growth by sector between 2000 and 2015 (or nearest year), OECD average¹
Mean Median
Change in employment since 2000, %
50
42
40
34
30
23
20
20
12
9
10
-10 -6
-9
-20
-22 -22
-30
Total Agriculture Industry Services Healh and social work²
1. Average of 30 OECD countries (excluding Chile, Iceland, New Zealand, Switzerland and Turkey).
2. Health and social work is classified as a sub-component of the services sector.
Source: OECD National Accounts.
12 http://dx.doi.org/10.1787/888933604552
Health at a Glance 2017 © OECD 2017 149
8. HEALTH WORKFORCE
Doctors (overall number)
The number of doctors per capita varies widely across number of doctors by increasing their training efforts over
OECD countries. In 2015, Greece had the highest number the past decade to ensure that there would be enough new
with 6.3 doctors per 1 000 population, but this number is an doctors to replace those who will retire. But the impact of
over-estimation as it includes all doctors who are licensed this increase into medical education will take several years
to practice but may no longer be practising for various for the effects to be felt. The difficulties in anticipating
reasons. Greece was followed by Austria (5.1 doctors per the actual number of practicing doctors have resulted in
1 000 population). Turkey, Chile and Korea had the lowest countries continually having to revise and adjust their
number among OECD countries at around two doctors policies. However, in most OECD countries, there is a
per 1 000 population. The OECD average was 3.4 doctors shared concern on the shortages of general practitioners
per 1 000 population. Among the partner countries, the (see the indicator “Doctors by age, sex and category”) and
number of doctors per capita is significantly lower: there the undersupply of doctors in rural and remote regions (see
was less than one doctor per 1 000 population in Indonesia, the indicator on the “Geographic distribution of doctors”
India and South Africa. In China, the number of doctors per in Chapter 5).
capita is still about half the OECD average, but it has grown
by 44% since 2000 (Figure 8.3).
Since 2000, the number of doctors has increased in nearly
all OECD countries, both in absolute number and on a per Definition and comparability
capita basis. The growth rate was particularly rapid in some
countries which started with lower levels in 2000 but have The data for most countries refer to practising
grown at a significantly faster rate than the OECD average doctors, defined as the number of doctors who are
growth rate, such as Korea, Mexico and the United Kingdom providing care directly to patients. In many countries,
(Figure 8.4). the numbers include interns and residents (doctors
in training). The numbers are based on head counts.
At the same time, countries with high physician density Several countries also include doctors who are active
such as Australia and Austria have also continued to show in the health sector even though they may not
a high rate of increase over the same period. The number provide direct care to patients, adding another 5-10%
of doctors has continued to increase strongly in Australia, of doctors. Greece and Portugal report the number of
driven by a strong rise in the number of graduates from physicians entitled to practice, resulting in an even
domestic medical education programmes (see the indicator larger over-estimation of the number of practicing
on “Medical graduates”). doctors. Belgium sets a minimum threshold of
In the United Kingdom, concerns were raised in the early activities for doctors to be considered to be practising,
2000s about possible surpluses in certain categories of thereby resulting in an under-estimation compared
doctors. This resulted in policies to reduce student intakes with other countries which do not set such minimum
and to some tapering of the growth rate in the number thresholds. Data for India may be over-estimated as
of doctors. More recently, though, funding for additional they are based on medical registers which are not
student places at medical schools was announced to meet updated to account for migration, retirement or death,
the growing demand for care (Department of Health, 2016). nor do they take into account doctors registered in
The number of physicians per capita remained fairly stable multiple states.
between 2000 and 2015 in France, Israel, Poland and the
Slovak Republic. In Israel, the number of doctors increased
at nearly the same pace as the population size.
Overall, most OECD countries have shown a steady increase References
in the number of doctors, and did not show much effect of
the global recession. In countries such as Australia, there Department of Health (2016), “Up to 1,500 extra medical
were about 30% more employed doctors in 2015 than in training places announced”, Department of Health,
2008. There were some exceptions: the 2008-09 recession London, https://www.gov.uk/government/news/up-to-1500-
appears to have had an impact in Greece, where the extra-medical-training-places-announced.
number of doctors increased between 2000 and 2008, but OECD (2016), Health Workforce Policies in OECD Countries:
has stopped growing afterwards and has even shown some Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
decline since 2012. http://dx.doi.org/10.1787/9789264239517-en.
Projecting the future supply and demand of doctors is Ono, T., G. Lafortune and M. Schoenstein (2013), “Health
challenging given the high levels of uncertainty concerning Workforce Planning in OECD Countries: A Review of
their retirement and migration patterns as well as changes 26 Projection Models from 18 Countries”, OECD Health
in their demand (Ono et al., 2013). Many OECD countries Working Papers, No. 62, OECD Publishing, Paris, http://
have anticipated the upcoming retirement of a significant dx.doi.org/10.1787/5k44t787zcwb-en.
150 Health at a Glance 2017 © OECD 2017
8. HEALTH WORKFORCE
Doctors (overall number)
8.3. Practising doctors per 1 000 population, 2000 and 2015 (or nearest year)
2000 2015
Per 1 000 population
7
6.3
6
5.1
5
4.6
4.4
4.3
4.2
4.2
4.1
4.0
3.9
3.8
3.8
3.7
3.7
4
3.5
3.5
3.5
3.4
3.4
3.4
3.3
3.2
3.2
3.1
3.0
3.0
2.9
2.9
2.8
2.8
2.7
3
2.6
2.4
2.4
2.3
2.2
2.1
1.8
1.8
1.8
1.8
2
0.8
0.7
1
0.3
0
Po ico
a
ng ia
er a
S w a nd
n rm n
ra y
Sp n
nm ic
CD a
Fr 3 5
M an
Ko d
a
lo e¹
Br i a
Ch il
So Tu ina
h y²
do ia
a
Au ce¹
r t u ia
N o g a l¹
S w ithu ay
n
e c Ic e t a l y
p ²
S l N e u s tr r k
ak e r l li a
pu d s
Is ²
Es r ael
Fi ce ²
L d
ng a
Ne B e ar y
Lu Z e m
m nd
Ir e r g
i t e lo nd
Un C a om
St ²
Ja es
Re nd
d da
ic
de n
ric
i t z ani
OE oni
Hu a t v i
re
si
az
i a G e e de
tio
n
ai
an
Ki n
De ubl
In In d
Po s t r
C o Chil
w l gi u
u t r ke
A a
u
L rw
at
ov th a
Fe a
p
Re an
bl
la
a
xe a l a
ne
d ve
m
ee
ex
an
ite na
d
h la
bo
l
nl
Af
t
Gr
S
Cz
Un
ss
Ru
1. Data refer to all doctors licensed to practice, resulting in a large over-estimation of the number of practising doctors (e.g. of around 30% in Portugal).
2. Data include not only doctors providing direct care to patients, but also those working in the health sector as managers, educators, researchers, etc.
(adding another 5-10% of doctors).
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604571
8.4. Evolution in the number of doctors, selected OECD countries, 2000 to 2015 (or nearest year)
Countries above OECD average per capita Countries below OECD average per capita
Australia Austria Japan Korea
Germany Greece¹ Mexico United Kingdom
Sweden OECD30 United States OECD30
Index (2000 = 100) Index (2000 = 100)
200 200
180 180
160 160
140 140
120 120
100 100
80 80
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
1. The data for Greece refer to all doctors licensed to practice.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604590
Health at a Glance 2017 © OECD 2017 151
8. HEALTH WORKFORCE
Doctors by age, sex and category
The age and gender composition of the medical workforce share to 2005. Greece, Hungary and the United States
and the mix between different categories of doctors have showed the lowest share of generalists, while countries
important implications on the availability of medical such as France, Canada and Australia have been able to
services. The ageing of doctors in OECD countries has, for maintain a more equal balance between specialists and
many years, raised concerns that there may not be sufficient generalists. It should be noted that in Ireland and Portugal,
new recruits to replace them, although there is evidence most generalists are not general practitioners but rather
that the retirement of doctors often only occurs gradually non-specialist doctors working in hospitals or other settings.
and that their retirement age is increasing (OECD, 2016). In the United States, general internal medicine doctors are
The growing imbalance in favour of greater specialisation categorised as specialists although their practice is often
over general medicine raises concerns in many countries very similar to that of general practitioners, resulting in
about access to primary care for all the population. some underestimation of the capacity to provide generalist
In 2015, on average across OECD countries, one-third of care.
all doctors were over 55 years of age, up from one-fifth in In response to concerns about shortages of general
2000 (Figure 8.5). Between 2000 and 2015, France, Italy, Spain practitioners, many countries have taken steps to improve
and Austria more than doubled the share of doctors over the number of training places in general medicine. In
55 years of age. While these doctors might be expected to Canada, the number of post-graduate training places in
retire over the next ten years, a growing number of them family medicine more than doubled between 2000 and 2013,
will likely continue to practice after 65 years. In Israel and as part of a national effort to improve access to primary
Italy, half (or more) of all doctors were over 55 years of age care (CAPER, 2015). However, in most OECD countries,
in 2015. It should be noted that the high share in Israel may specialists earn more than general practitioners, providing
be due partly to the fact that these numbers are based on financial incentives for doctors to specialize (see indicator
all doctors licensed to practice, which may include some on the “Remuneration of doctors”).
who may no longer be practicing. At the other end, only 13-
17% of doctors in the United Kingdom and Korea were aged
over 55. This is consistent with the large numbers of new Definition and comparability
graduates entering medical practice over the past decade
(see the indicator on “Medical graduates”). The definition of doctors is provided under the
Several OECD countries have reformed their pension systems previous indicator. In some countries, the data are
and increased the retirement age to take into account based on all doctors licensed to practice, not only
longer life expectancy. While few studies have examined those practising (e.g., Greece and Portugal). Not all
the impact of these pension reforms specifically on doctors, countries are able to report all their physicians in the
it is possible that these pension reforms may prolong the two broad categories of specialists and generalists.
working lives of doctors after age 65, which could have a This may be due to the fact that specialty-specific data
significant impact on the future replacement needs. are not available for doctors in training or for those
working in private practice.
In 2015, 46% of doctors on average across OECD countries
were women, up from 39% in 2000 (Figure 8.6). At least
half of all doctors now are women in 11 countries, with
Latvia and Estonia showing the highest share at over 70%.
References
Between 2000 and 2015, the share of women doctors rose
most rapidly in the Netherlands (49%) and Belgium (47%). CAPER (2015), “Field of Post-M.D. Training by Faculty of
By contrast, only about one-in-five doctors in Japan and Medicine Providing Post-M.D. Training 2013-2014”, Database
Korea were women in 2015, although Japan showed a available at http://www.caper.ca.
significant increase of 42% over the 2000 figure. OECD (2016), Health Workforce Policies in OECD Countries:
On average across OECD countries, generalists made up Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
about 30% of all physicians in 2015 (Figure 8.7), a similar http://dx.doi.org/10.1787/9789264239517-en.
152 Health at a Glance 2017 © OECD 2017
0
10
20
30
40
50
60
70
80
20
30
40
50
60
%
0
10
%
0
20
40
60
80
100
%
Ir e La
l t It a
Po an d 3 Es v ia 74.4
ly
r tu 59 41 to Is 53.3
ga Sl ni a ra
l3 ov 73.1
51 49 Sl e Es el 49.9
Ch to
i l
ov F nia
62.4 ni
Ca e 49 51 ak inl Fr a 46.7
na Re an d an
d pu 57.7 ce
Fr a 47 53 b 46.6
a La
Au nce Po lic 57.5 tv
N e s tr 46 54 C z H lan Be ia 46.2
56.9 lg
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
t h a li e c un d
Health at a Glance 2017 © OECD 2017
er a 45 49 6 h ga G e ium
la Re r y rm 44.4
Ge nd s pu 55.9
45 55
an
rm P bli y 43.9
a N e or t c 54.5 Lu Ic e l
F i ny 41 59 th uga xe an
m d
Generalists¹
nl er l 54.1 42.9
an la bo
Be d
lg 37 58 4
nd u
iu Sp s 52.6 Hu r g
41.2
ng
M m 37 61 1 De ain a
ex nm 52.6 C z De r y
ic e c nm 40.5
36 64 h
49.0
Ne Au o No ar k Sl ar
2000
2000
w str rw ov Rep k 37.4
32 54 14
Z e ia S w ay ak ub
al ed 47.6 Re lic
an 37.2
pu
Lu Tu d 32 61 8 A u en 47.6 bl
xe ke r Un O s t r i
m y Un J c 36.0
32 68 i t e E C ia 46.5
bo d ite apa
OE urg Ki D3 4 d n 35.2
CD 30 70 ng 46.5 St
at
Un 32 G e d om OE s e
30 63 7 rm 45.9 35.1
Specialists²
ite
d sr a I an Sw CD
Ki it z 30
el
30 70
N e Fr y 45.7 er 34.5
S w ngd
o
w anc la
Ze e
it z m
28 72 al 44.3 Sw nd 34.1
er
la a ed
nd Ir e n d 43.9 e
Ko 28 58 14 la Ca n 32.7
na
Sl r e a Ca nd 43.2 da
ov 27 73 na Au 32.4
en d st
ia Is a 41.8 r ia
23 73 3 ra 32.1
It a Sp
Gr e l 41.5 ai
Es l y 23 77 ee n
to ce 31.0
ni 41.3 Ne C
77
23
It a w hil
La a B l Ze e 27.9
40.7
2015
2015
al
C z De t v ia
22 78
Sw elg y a
e c nm i t z ium Sl nd
h er 40.6 ov 27.0
Re ar k la en
pu 19 46 34 n
b Tu d 40.6 F i ia 26.9
r nl
No lic 19 80 1 A u ke y
rw st 40.3 A u and
26.7
8.6. Share of female doctors, 2000 and 2015 (or nearest year)
a ra N e s tr
Po y 19 41 40 li a
la 40.0 t h a li a
1. Generalists include general practitioners/family doctors and other generalist (non-specialist) medical practitioners.
n Ch er
la 26.2
2. Specialists include paediatricians, obstetricians/gynaecologists, psychiatrists, medical, surgical and other specialists.
17 83
Sw d n
ed Un Ic il e 39.6
e ite ela
n
No d s 25.6
15 53 32 rw
Medical doctors not further defined
37.4
Un Ic e n Lu d S t d a
8.5. Share of doctors aged 55 years and over, 2000 and 2015 (or nearest year)
ite lan xe a t 25.2
8.7. Generalists and specialists as a share of all doctors, 2015 (or nearest year)
d m es Ir e y
St d 15 58 27 bo 34.6 la
a ur Un nd
Hu t es 22.2
12 88 34.0
ite K
ng Ko g d o
a re Ki rea
Gr r y 12 58 31 Ja a 22.3
ng 16.7
ee pa do
ce n m
6 75 18 20.3 12.8
3. In Ireland and Portugal, most generalists are not GPs (“family doctors”), but rather non-specialist doctors working in hospitals or other settings.
153
12 http://dx.doi.org/10.1787/888933604647
12 http://dx.doi.org/10.1787/888933604628
12 http://dx.doi.org/10.1787/888933604609
Doctors by age, sex and category
8. HEALTH WORKFORCE
8. HEALTH WORKFORCE
Medical graduates
The number of new medical graduates in a given year few years. Spain showed a slight decline in the number
reflects to a large extent government decisions taken a of medical students until 2012, when the numbers have
few years earlier on the number of students admitted in begun to increase rapidly again, growing by 36% between
medical schools (so-called numerus clausus policies). Since 2012 and 2015.
2000, most OECD countries have increased the number In the United States, the increase in admission intakes to
of students admitted to medical education in response medical schools also took place after 2005, and the number
to concerns about current or possible future shortages of of medical graduates has shown a gradual increase over
doctors (OECD, 2016), but large variations remain across the past decade, which included a growing number of
countries. American students who study abroad (notably in Caribbean
In 2015, there were on average about 12 new medical countries), with the intention of coming back to complete
graduates per 100 000 population across OECD countries their post-graduate training and practice in the United
(Figure 8.8). This proportion was highest in Ireland at States. This is expected to create additional pressures
24 new medical graduates per 100 000. At the other end, to increase the number of residency posts to allow both
Israel and Japan had the lowest number of new medical domestic graduates and foreign-trained US national
graduates relative to their population. In Ireland, the graduates to complete their post-graduate training.
number of medical graduates increased strongly in 2013 There has also been a strong rise in the number of medical
due at least partly to the opening of new Graduate Entry graduates in the Czech Republic and Poland. This increase
Programmes a few years earlier, allowing students with since around 2009 can be explained partly by the growing
an undergraduate degree in another discipline to obtain a number of international students choosing these countries
medical degree in four years only. to purse their medical studies. International students
In Israel, the low number of domestic medical graduates accounted for about 30% of all medical graduates in the
is compensated by the high number of foreign-trained Czech Republic in recent years. The internationalisation
doctors. About one-third of foreign-trained doctors in Israel of medical education combined with migration makes it
are people who were born in Israel but have pursued their more challenging for national governments to set their own
study abroad before coming back. The situation is quite domestic policies (OECD, 2016).
different in Japan, where there are very few foreign-trained
doctors. Since 2008, the Japanese government decided
to increase intakes in medical education in response to
Definition and comparability
current and projected shortages of doctors; however, this
policy has not yet translated into an increase in the number Medical graduates are defined as the number of
of medical graduates. students who have graduated from medical schools
The expansion of the numerus clausus in many of the OECD in a given year. The data for Australia, Austria and
countries over the past fifteen years has resulted in an the Czech Republic include foreign graduates, but
increase in the number of medical graduates, although other countries may exclude them. In Denmark, the
they are occurring at varying paces (Figure 8.9). Australia data refer to the number of new doctors receiving
has shown the fastest rate of increase in the number of an authorisation to practice, which may result in an
medical graduates, growing by 2.7 times between 2000 over-estimation if these include a certain number of
and 2015. While most of this growth reflects an increase foreign-trained doctors.
in the number of domestic students, it should be noted that
this figure also reflects a growing number of international
students in medical schools in Australia.
In the United Kingdom, the number of medical graduates References
doubled between 2000 and 2015, reflecting an effort to
ACMMP (2014), The 2013 Recommendations for Medical
increase the domestic supply and rely less on foreign-
Specialist Training, Utrecht.
trained doctors. While there was a slight decrease in the
number of graduates from 2013, in 2016 the government Department of Health (2016), “Up to 1,500 Extra Medical
announced the intent to provide funding for additional Training Places Announced”, Department of Health,
1 500 students to meet the growing demand for care London, https://www.gov.uk/government/news/up-to-1500-
(Department of Health, 2016). By contrast, there has been extra-medical-training-places-announced.
a continued slow-down in the growth in number of medical OECD (2016), Health Workforce Policies in OECD Countries:
graduates in the Netherlands (ACMMP, 2014). Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
In France, the number of medical graduates increased http://dx.doi.org/10.1787/9789264239517-en.
steadily since 2006 following a large increase in the numerus UN High-Level Commission on Health Employment and
clausus between 2000 and 2006. However, the number of Economic Growth (2016), “Working for Health and Growth:
graduates is expected to stabilize in the coming years, as Investing in the Health Workforce”, WHO, Geneva,
admission quotas have remained fairly stable over the past retrieved from http://www.who.int/hrh/com-heeg/reports.
154 Health at a Glance 2017 © OECD 2017
8. HEALTH WORKFORCE
Medical graduates
8.8. Medical graduates, 2015 (or nearest year)
Per 100 000 population
25
23.7
20 19.5
17.4
16.2 15.9
15.8
15 14.7 14.6
13.6 13.5 13.4
13.0 12.8 12.7
12.4 12.3 12.1 12.1
11.3 11.2 11.0
10.7 10.6 10.5
10.2 10.0
10 9.4 9.3
8.9 8.7
7.9 7.9 7.5
6.4
5.5
5
S co
La a
Re lic
Hu lic
nm d
Po ia
Au gal
A ia
C z t h e r ia
Sl h R nd s
ng n
Fi m
d
OE ly
Be 3 4
G e ium
N o ny
Es ay
M ia
i t z en
Po d
Fr d
ce
Gr l e
ce
Ic d
Ze y
Ca d
Un K a
St a
es
n
el
Sl rk¹
i
w ke
d
i t e or e
ar
n
i
De lan
an
n
an
pa
an
en
ra
n
tv
It a
d Spa
do
rw
an
ak ub
ee
Ch
at
ra
Sw wed
a
na
la
la
t
CD
a
r tu
ex
to
ng
Ne us
Ne Tur
rm
pu
nl
el
Ja
al
ec rla
lg
Is
ov
er
Ir e
st
ov ep
d
Ki
ite
Un
1. In Denmark, the number refers to new doctors receiving an authorisation to practice, which may result in an over-estimation if these include
foreign-trained doctors.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604666
8.9. Evolution in the number of medical graduates, selected OECD countries, 2000 to 2015 (or nearest year)
Countries above OECD average per capita Countries below OECD average per capita
Australia Ireland Canada France
Spain United Kingdom Japan Poland
Netherlands United States
Index (2000 = 100) Index (2000 = 100)
300 300
250 250
200 200
150 150
100 100
50 50
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604685
Health at a Glance 2017 © OECD 2017 155
8. HEALTH WORKFORCE
Remuneration of doctors (general practitioners and specialists)
The remuneration level for different categories of doctors has In some OECD countries, the economic crisis of 2008-09
an impact on the financial attractiveness of different medical had an impact on the remuneration of doctors and other
specialties. In many countries, governments influence the health workers. Several European countries hard hit by the
level and structure of physician remuneration by being one recession either froze or reduced the wages or fees of doctors
of the main employers of physicians or purchaser of their in efforts to reduce cost while protecting access to care for
services, or by regulating their fees. With the increasing the population. This has been the case in Estonia, Ireland,
international mobility of doctors across national borders Italy and Slovenia, where doctors saw their remuneration
(see the indicator on migration of doctors and nurses), the decrease for some years after the crisis. However, in more
relative levels of remuneration across countries can play an recent years, the remuneration of doctors and other health
important role in influencing these movements. workers has started to rise again (OECD, 2016).
OECD data on physician remuneration distinguish
between salaried and self-employed physicians. In some
countries this distinction is increasingly blurred, as Definition and comparability
some salaried physicians are allowed to have a private
practice and some self-employed doctors may receive The remuneration of doctors refers to average gross
part of their remuneration through salaries. A distinction annual income, including social security contributions
is also made between general practitioners and all other and income taxes payable by the employee. It should
medical specialists combined, although there may be wide normally exclude practice expenses for self-employed
differences in the income of different medical specialties. doctors.
In the OECD countries where data are available, the A number of data limitations contribute to an under-
remuneration of doctors (both general practitioners and estimation of remuneration levels in some countries:
specialists) is much higher than that of the average worker 1) payments for overtime work, bonuses, other
(Figure 8.10). In 2015, self-employed general practitioners supplementary income or social security contributions
in Austria, Canada, France and the United Kingdom earned are excluded in some countries (Austria for GPs,
around three times the average wage in the country while Ireland for salaried specialists and Italy); 2) incomes
in Germany they earned over four times the average wage. from private practices for salaried doctors are not
In Australia, self-employed general practitioners earned included in some countries (e.g. Czech Republic,
about two times the average wage in 2015, but it should Hungary, Iceland, Ireland and Slovenia); 3) informal
be noted that this is an under-estimation since the figure payments, which may be common in certain countries
includes the remuneration of physicians in training. (e.g. Greece and Hungary), are not included; 4) data
relate only to public sector employees who tend to
In most countries, specialists earned significantly
earn less than those working in the private sector in
more than the average worker, and more than the
Chile, Denmark, Greece, Hungary, Iceland, Ireland,
general practitioners. In 2015, the income gap between
Norway, the Slovak Republic and the United Kingdom;
specialists and general practitioners was particularly
and 5) physicians in training are included in Australia.
high in Australia, Belgium and Luxemburg, where the self-
employed specialists earned over twice the remuneration The data for some countries include part-time
earned by general practitioners. In comparison with the workers, while in other countries the data refer only to
average worker, self-employed specialists in Belgium doctors working full time. In Belgium, the data for self-
and Luxembourg earned six times the average wage, and employed doctors include practice expenses, resulting
in France and Germany they earned around five times in an over-estimation.
the average wage. It should be noted that in Belgium The income of doctors is compared to the average
the remuneration included practice expenses, thereby wage of full-time employees in all sectors in the
resulting in an over-estimation. country. The source for the average wage of workers in
In many OECD countries, the income gap between general the economy is the OECD Employment Database. For the
practitioners and specialists has continued to widen over calculation of growth rates in real terms, economy-
the past decade, reducing the financial attractiveness of wide GDP deflators are used.
general practice (Figure 8.11). Since 2005, the remuneration
of specialists has risen faster than that of generalists in
Canada, Finland, France, Hungary, Israel, Luxembourg and
Reference
Mexico. On the other hand, in Austria, Belgium, Estonia
and the Netherlands, the gap has narrowed slightly, as the OECD (2016), Health Workforce Policies in OECD Countries:
income of general practitioners grew faster than that of Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
specialists. http://dx.doi.org/10.1787/9789264239517-en.
156 Health at a Glance 2017 © OECD 2017
8. HEALTH WORKFORCE
Remuneration of doctors (general practitioners and specialists)
8.10. Remuneration of doctors, ratio to average wage, 2015 (or nearest year)
Salaried Self-employed
Specialists General practitioners (GPs)
3.9 Australia¹ 1.9
4.1 Austria 2.7
6.0 Belgium² 2.4
4.5 Canada 2.8
3.3 Chile 2.6
2.6 Czech Republic n.a.
2.6 Denmark 2.7
2.2 Estonia 1.6
2.2
2.6 Finland 1.8
2.2 France
4.9 3.0
3.5 Germany
5.3 4.1
2.4 Greece n.a.
2.2 Hungary 1.9
2.4 Iceland 2.2
3.3 Ireland 2.4
3.8 Israel 2.2
2.5 Italy n.a.
1.6 Latvia 1.0
4.3 Luxembourg
6.0 2.8
3.4 Mexico 2.6
3.4 Netherlands 2.1
4.3 2.5
3.1 New Zealand 2.5
3.2
1.8 Norway n.a.
1.5 Poland 2.1
2.3 Slovak Republic n.a.
2.1 Slovenia 2.2
2.3 Spain 2.0
2.3 Sweden n.a.
3.4 United Kingdom 1.6
3.1
6 4 2 0 0 2 4 6
Ratio to average wage in each country Ratio to average wage in each country
1. Physicians in training included (resulting in an under-estimation).
2. Practice expenses included (resulting in an over-estimation).
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604704
8.11. Growth in the remuneration of GPs and specialists, 2005-15 (or nearest year)
GPs Specialists
Average annual growth rate (%, in real terms)
5
4.2
4.0 3.8
4
3.2
3 2.5
2.2
2 1.5 1.6
1.4
1 0.8 0.8 0.8
0.5 0.5 0.4 0.3
0.2 0.2 0.3
0
-0.2 -0.1
-1
-1.0
-2
o
r ia
ce
da
el
g¹
s¹
ni
ar
ic
an
ra
nd
iu
ur
an
na
st
to
ex
ng
nl
lg
Is
bo
la
Au
Fr
Ca
Es
M
Fi
Be
Hu
er
m
th
xe
Ne
Lu
1. The growth rate for the Netherlands and for Luxembourg is for self-employed GPs and specialists.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604723
Health at a Glance 2017 © OECD 2017 157
8. HEALTH WORKFORCE
Nurses
Nurses greatly outnumber physicians in most OECD In response to shortages of doctors and to ensure proper
countries, and they play a critical role in providing health access to care, some countries have developed more
care not only in traditional settings such as hospitals and advanced roles for nurses. Evaluations of nurse practitioners
long-term care institutions but increasingly in primary care from the United States, Canada and the United Kingdom
settings (especially to manage the care of the chronically show that advanced practice nurses can improve access
ill) and in home care settings. to services and reduce waiting times, while delivering the
There are growing concerns in many OECD countries about same quality of care as doctors for a range of patients,
possible future shortages of nurses, given that the demand including those with minor illnesses and those requiring
for nurses is expected to rise in a context of population routine follow-up. Existing evaluations find a high patient
ageing and the retirement of the current “baby-boom” satisfaction rate, while the impact on cost is either cost-
generation of nurses. These concerns have prompted reducing or cost-neutral. The implementation of new
actions in many countries to increase the training of advanced practice nursing roles may require changes
new nurses (see the indicator on “Nursing graduates”), to legislation and regulation to remove any barrier to
combined with efforts to increase the retention rate of extensions in their scope of practice (Delamaire and
nurses in the profession. The retention rate of nurses has Lafortune, 2010).
increased in recent years in many countries either due
to the impact of the economic crisis that have prompted
more nurses to stay or come back in the profession, or Definition and comparability
following deliberate efforts to improve their working
conditions (OECD, 2016). The number of nurses includes those employed in
public and private settings providing services directly
On average across OECD countries, the number of nurses on
to patients (“practising”) and in some cases also those
per capita basis has gone up from 7.3 per 1 000 population
working as managers, educators or researchers.
in 2000 to nine nurses per 1 000 population in 2015
(Figure 8.12). In 2015, the number of nurses per capita was In those countries where there are different levels of
highest in Switzerland, Norway, Denmark, Iceland and nurses, the data include both “professional nurses”
Finland, with more than 14 nurses per 1 000 population. who have a higher level of education and perform
The number of nurses per capita in OECD countries was higher level tasks and “associate professional
lowest in Turkey, Chile and Mexico (with less than 3 nurses” who have a lower level of education but are
per 1 000 population). With regards to OECD partner nonetheless recognised and registered as nurses.
countries, the number of nurses per capita was generally Health care assistants (or nursing aids) who are not
low compared with the OECD average. In 2015, Colombia, recognised as nurses are excluded. Midwives are
Indonesia, South Africa, India and Brazil had fewer than excluded, except in some countries where they are
1.5 nurses per 1 000 population, although numbers have at least partly included because they are considered
been growing quite quickly in Brazil in recent years. as specialist nurses or for other reasons (Australia,
Ireland and Spain).
The number of nurses per capita increased in almost
all OECD countries since 2000. Korea and Portugal had a Austria and Greece report only nurses working in
relatively low density of nurses but have now converged hospital, resulting in an under-estimation.
towards the OECD average. France has also increased from
a relatively low density to a level above the OECD average. A
significant increase was registered in countries that already
had a high density of nurses in 2000, such as Switzerland, References
Finland and Denmark. In Ireland and Israel, the number of
nurses per capita declined between 2000 and 2015 as the Delamaire, M.-L. and G. Lafortune (2010), “Nurses in
size of the population grew more rapidly than the number Advanced Roles: A Description and Evaluation of
of nurses. In the Slovak Republic, the number of nurses Experiences in 12 Developed Countries”, OECD Health
declined both in absolute numbers and on a per capita Working Paper, No. 54, OECD Publishing, Paris, http://
basis. dx.doi.org/10.1787/5kmbrcfms5g7-en.
In 2015, there were about three nurses per doctor on OECD (2016), Health Workforce Policies in OECD Countries:
average across OECD countries, with about half of the Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
countries reporting between two to four nurses per doctor http://dx.doi.org/10.1787/9789264239517-en.
(Figure 8.13). The nurse-to-doctor ratio was highest in UN High-Level Commission on Health Employment and
Japan, Finland and Denmark (with 4.6 nurses per doctor). Economic Growth (2016), “Working for Health and Growth:
It was lowest in Chile, Turkey and Mexico with less than Investing in the Health Workforce”, WHO, Geneva,
1.2 nurse per doctor). retrieved from http://www.who.int/hrh/com-heeg/reports.
158 Health at a Glance 2017 © OECD 2017
Sw
0
1
2
3
4
5
Ja it z
0
4
8
12
16
20
F i pan er
Ratio
D nl 4.6 No land
S en an d 18.0
Un w i t ma 4.6 De r w
nm ay 17.3
licensed to practice.
ite zer r k 4.6 Ic ar
el k 16.7
L u d S l a nd
xe t a t 4.3
a
Fi nd
m es 15.5
bo ¹ 4.1 Ge nlan ¹
rm d
Per 1 000 population
Ic u r g
e 14.7
4.1 L u Ir e a n
No land xe la y 13.3
r 4.1 m nd
11.9
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
C a way
n 3.9
Un A bou ¹
Health at a Glance 2017 © OECD 2017
Ir a d ite ust rg 11.9
In e l a n a 3.9 d r al
St ia
do d a 11.5
n ¹ 3.8 Sw tes
Ne B e esi ed ¹ 11.3
w l gi a 3.7
Z e um Ja en 11.1
Au alan 3.6 B
Ne el pan
s d g 11.0
G e tr a 3.4 N e t h e r ium
r m li a w lan 10.8
3.3 Ze d
N e S l o a ny al s 10.5
th ven 3.2 Fr an d
er ia an 10.3
2000
la 3.1 Ca ce
Fr nd s ¹ 9.9
Un a 3.0
Ru O E n ad
ite A nc ss C a 9.9
2. Austria and Greece report only nurses employed in hospital.
3. Data in Chile refer to all nurses who are licensed to practice.
d us e¹ 3.0
ia S D
Ki tri n lo 3 5
ng a ² F e ve 9.0
2.9 de ni
OE dom Cz ra a 8.8
C 2.8 t
8.7
Sw D 3 Un e ch A u s i o n
Ru ed 5 2.8 i t e Re t r i
ss d pu a ² 8.1
ia Ko e n 2.7 Ki b
n li
n r Li gdo c 8.0
C z F e Pol e a 2.7 th m
e c de an u 7.9
h ra d
Re tio 2.2 Hu an
i
p n Po nga a 7.7
2.2
2. For Austria and Greece, the data refer to nurses and doctors employed in hospital.
Hu ubl r tu r y
6.5
ng ic 2.2 Sl Es gal
ar ov to ¹ 6.3
L i In d y 2.1 ak K ni a
Sl th ia
u 6.0
1.9
ov Re or
pu e a
ak Es ania 5.9
Re to n 1.8 bl
ic
So p ia 5.7
1.8
u t ub It a ¹
h lic
A ¹ Sp l y 5.4
1.7 Po ain
Gr f r i c la 5.3
2015
ee a 1.6 n
c Is d 5.2
La e² 1.5 r
tv
4.9
L a ael
Is i a 1.5
8.13. Ratio of nurses to doctors, 2015 (or nearest year)
ra Gr t v i
e ee a 4.7
It a l 1.4 M ce ²
l ex 3.2
Po Spa y 1.4 i
r tu in Ch c o 2.8
ga 1.4 i
Ch n a 2.4
concept for both nurses and doctors (except for Chile where numbers include all nurses and doctors licensed to practice).
Ch l ³ 1.4 Tu ile³
M in a rk 2.1
e 1.3 e
8.12. Practising nurses per 1 000 population, 2000 and 2015 (or nearest year)
Tu xico Br y¹ 2.0
rk 1.2 So az
e u t In i l 1.5
Ch y¹ 1.1 h d
B
il e In A f r i a 1.4
Co r a ¹ 1.0 do i c
lo z il Co ne a 1.2
m 0.8 lo si a
bi
a m 1.2
0.6 bi
a 1.1
3. The ratio for Portugal is underestimated because the numerator refers to professionally active nurses while the denominator includes all doctors
159
12 http://dx.doi.org/10.1787/888933604761
12 http://dx.doi.org/10.1787/888933604742
Nurses
8. HEALTH WORKFORCE
1. Data include not only nurses providing direct care to patients, but also those working in the health sector as managers, educators, researchers, etc.
1. For those countries which have not provided data for practising nurses and/or practising doctors, the numbers relate to the “professionally active”
8. HEALTH WORKFORCE
Nursing graduates
Many OECD countries have taken steps over the past number of nursing graduates in the earlier part of the
decade or so to increase the number of students admitted decade, but has shown some modest increase in recent
in nursing schools in response to concerns about current years.
or possible future shortages of nurses (OECD, 2016). In France, the number of graduates from nursing schools
Nonetheless, there are wide variations across countries in increased by 87% between 2000 and 2015. The numerus
training efforts of new nurses, which may be explained by: clausus set by the French Ministry of Health to control
differences in the current number and age structure of the entry in nursing education programmes was expanded
nursing workforce (and hence the replacement needs); in substantially since 1999. Most of the growth occurred in
the capacity of nursing schools to take on more students; the academic year of 2000/01 when the annual quota was
and the future employment prospects of nurses. increased by 43%, driven by a projected reduction in the
In 2015, there were on average around 46 new nurse supply of nurses resulting from the reduction of working
graduates per 100 000 population across OECD countries, up time to 35 hours per week, as well as a more general
from less than 40 in 2003. Korea, Switzerland and Denmark concern about the anticipated retirement of a large number
had the highest number of new nurse graduates relative of nurses.
to their population, with these three countries graduating
more than 90 new nurses per 100 000 population in 2015.
Mexico, Luxembourg and the Czech Republic had the Definition and comparability
lowest number, with less than 16 nurse graduates per
100 000 population (Figure 8.14). Nursing graduates refer to the number of students
Over the past decade, the number of nursing graduates who have obtained a recognised qualification required
has increased in all OECD countries, but at different rates to become a licensed or registered nurse. They include
(Figure 8.15). The number has increased strongly in many graduates from both higher level and lower level
of the countries which had relatively low number of nursing programmes. They exclude graduates from
graduates per capita. Mexico has among the lowest number Masters or PhD degrees in nursing to avoid double-
of nursing graduates, but between 2000 and 2015 there was counting nurses acquiring further qualifications.
an eight-fold increase in the number of nursing graduates The data for Denmark and the United Kingdom are
per capita. Over the same period, Italy has also shown a based on the number of new nurses receiving an
four-fold increase in the number of nursing graduates per authorisation to practice.
capita.
Among the countries already with above average number of
nursing graduates per capita, the increase has been more References
modest. Germany has shown an increase in the number
of nurse graduates since 2012 through the expansion Cassier-Woidasky, A.-K. (2013), Nursing Education in Germany
of registered nurse training programmes in several – Challenges and Obstacles in Professionalisation, DHBW,
universities, in addition to the programmes traditionally Stuttgart.
offered in vocational nursing schools (Cassier-Woidasky, OECD (2016), Health Workforce Policies in OECD Countries:
2013). Norway has also shown a modest increase in the Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
last few years. Japan and Finland showed a decline in the http://dx.doi.org/10.1787/9789264239517-en.
160 Health at a Glance 2017 © OECD 2017
8. HEALTH WORKFORCE
Nursing graduates
8.14. Nursing graduates, 2015 (or nearest year)
Per 100 000 population
120
109.0
100
93.4
92.4
77.5
76.9
80
73.5
72.9
66.8
63.1
58.8
60
54.5
54.2
52.9
50.9
49.7
46.0
46.0
43.9
43.6
38.9
35.8
40
33.2
32.1
31.1
28.9
27.6
27.0
26.2
23.8
23.2
21.5
20.6
19.0
20
15.8
12.8
12.0
0
Ge da
Au nia
ic
Be n
O um
Ze 5
Hu n d
er y
Fr s
Sw ce
Es en
Po ia
Tu d
Ir e y
d
ng a
m ic
it z rea
nm d
No li a
Ic ay
Un F in d
St d
Ca es
ov A u ny
R e r ia
Po m¹
Gr a l
ce
n
il e
e c Is y
Re r ael
M rg
o²
Sl rk¹
Ki vi
nd
l
th ar
De l a n
pa
n
an
ai
i te lan
w D3
bl
xe bl
n
It a
ic
u
rk
rw
an
at
ee
Ch
ra
ed
a
na
la
la
a
ak s t
do
a
d L at
e
to
Sp
r tu
S w Ko
Ne ng
bo
rm
pu
Lu pu
el
Ja
al
la
lg
ex
ov
Ne EC
er
st
h
ite
Cz
Sl
Un
1. In Denmark and the United Kingdom, the numbers refer to new nurses receiving an authorisation to practice, which may result in an over-
estimation if these include foreign-trained nurses.
2. In Mexico, the data include professional nursing graduates only.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604780
8.15. Evolution in the number of nursing graduates, selected OECD countries, 2000 to 2015 (or nearest year)
Countries above OECD average per capita Countries below OECD average per capita
Finland Germany France Israel
Japan Norway Italy Mexico
Index (2000 = 100) Index (2000 = 100)
200 800
175 700
150 600
125 500
100 400
75 300
50 200
25 100
0 0
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604799
Health at a Glance 2017 © OECD 2017 161
8. HEALTH WORKFORCE
Remuneration of nurses
The remuneration level of nurses is one of the factors Greece, the remuneration of nurses has been reduced
affecting their job satisfaction and the attractiveness of significantly, by as much as 25% in real terms between
the profession. It also has a sizeable impact on costs, since 2009 and 2015.
wages of nurses represent one of the largest spending
items in health systems.
The data presented in this section generally focus on the Definition and comparability
remuneration of nurses working in hospitals, although the
data coverage differs for some countries (see the box below The remuneration of nurses refers to average gross
on “Definition and comparability”). annual income, including social security contributions
and income taxes payable by the employee. It should
The data are presented in two ways. First, it is compared
normally include all extra formal payments, such as
with the average wage of all workers in each country,
bonuses and payments for night shifts and overtime.
providing some indication of the relative financial
In most countries, the data relate specifically to
attractiveness of nursing compared to other occupations.
nurses working in hospitals, although in Canada the
Second, the remuneration level in each country is converted
data also cover nurses working in other settings. In
into a common currency, the US dollar, and adjusted for
some federal states, such as Australia, Canada and
purchasing power parity, to provide an indication of the
the United States, the level and structure of nurse
relative economic well-being of nurses compared with their
remuneration is determined at the sub-national
counterparts in other countries.
level, which may contribute to variations across
In most OECD countries, the remuneration of hospital jurisdictions.
nurses was at or slightly above the average wage of all
Data refer only to registered (“professional”) nurses
workers in 2015 (Figure 8.16). In Mexico and Chile, the
in Chile, Ireland and the United States, resulting in
hospital nurses earned almost twice the average wage,
an over-estimation compared to other countries
while in Israel, Luxembourg and Spain, the wages of
where lower-level nurses (“associate professional”)
nurses were respectively 49%, 38% and 28% greater than
are also included. Data for Canada include registered
the average wage. In New Zealand, the United States,
(“professional”) nurses and unregistered nursing
Greece and Australia, it was about 20% greater than the
graduates. Data for New Zealand include all nurses
average wage. In most of the other countries, the wage of
employed by publically funded district health boards,
hospital nurses was roughly equal to the average wage in
registered and otherwise, and includes health
the economy, while in Hungary it was about 10% and in
assistants who have a different and significantly lower
Latvia about 20% lower.
salary structure than registered nurses.
When converted to a common currency (and adjusted for
The data relate to nurses working full time, with the
purchasing power parity), the remuneration of nurses was
exception of Belgium where part-time nurses are also
about five times higher in Luxembourg than in Hungary and
included (resulting in an under-estimation). The data
Latvia (Figure 8.17). Nurses in the United States also had
for some countries do not include additional income
relatively high earnings compared with their counterparts
such as overtime payments and bonuses (e.g. Italy
in other countries, which explains, at least partly, the
and Slovenia). Informal payments, which in some
ability of the United States to attract many nurses from
countries represent a significant part of total income,
other countries.
are not reported.
The economic crisis in 2008 has had a varying impact on
The income of nurses is compared to the average wage
the remuneration of nurses (Figure 8.18). The Netherlands,
of full-time employees in all sectors in the country.
for example, has seen a steady growth in remuneration for
The source for the average wage of workers in the
nurses. Some Central and Eastern European countries have
economy is the OECD Employment Database. For the
introduced a series of measures in recent years to increase
calculation of remuneration trends in real terms,
the retention of nurses and other health workers, including
economy-wide GDP deflators are used.
pay raises despite tight budget constraints. In Hungary, a
staged increase of 20% in the salaries of nurses and doctors
was introduced in 2012, phased over a three-year period.
In the Czech Republic, nurses also benefitted from a pay References
increase following protests of hospital workers in 2011
(although their pay raise was lower than that for doctors), OECD (2016), Health Workforce Policies in OECD Countries:
accompanied by some improvement in other aspects of Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
their working conditions (OECD, 2016). http://dx.doi.org/10.1787/9789264239517-en.
Following the recession, the remuneration of nurses OECD (2015), Fiscal Sustainability of Health Systems: Bridging
was cut down in some countries such as in Italy, which Health and Finance Perspectives, OECD Publishing, Paris,
has frozen wage increase over the past few years. In http://dx.doi.org/10.1787/9789264233386-en.
162 Health at a Glance 2017 © OECD 2017
8. HEALTH WORKFORCE
Remuneration of nurses
8.16. Remuneration of hospital nurses, ratio to 8.17. Remuneration of hospital nurses, USD PPP, 2015
average wage, 2015 (or nearest year) (or nearest year)
Mexico 1.83 Luxembourg 94.2
Chile¹ 1.81 United States¹ 73.0
Israel 1.49 Netherlands 66.1
Luxembourg 1.38 Australia 64.4
Spain 1.28 Ireland¹ 62.9
New Zealand 1.24 Belgium 59.6
United States¹ 1.24 Denmark 58.9
Greece 1.20 Israel 56.5
Australia 1.19 Iceland 56.0
Netherlands Canada² 55.2
1.15
OECD29 Chile¹ 55.0
1.14
Norway 54.2
Germany 1.13
Germany 53.7
Belgium 1.11
New Zealand 53.6
Canada² 1.09
Spain 53.2
Japan 1.09
United Kingdom 50.4
Czech Republic 1.08
OECD30 47.4
Denmark 1.07 Japan 44.7
Italy 1.06 France 42.5
Estonia 1.04 Finland 42.2
United Kingdom 1.04 Italy 42.1
Ireland¹ 1.01 Turkey 40.8
Poland 1.01 Slovenia 37.2
Norway 1.00 Greece 33.5
Iceland 0.97 Mexico 31.7
Slovak Republic 0.97 Czech Republic 26.9
Slovenia 0.97 Estonia 26.3
France 0.95 Poland 25.8
Finland 0.94 Slovak Republic 24.7
Hungary 0.91 Hungary 20.5
Latvia 0.78 Latvia 17.4
0 0.5 1 1.5 2 0 25 50 75 100
Ratio to average wage in each country USD PPP, thousands
1. Data refer to registered (“professional”) nurses in Chile, the United 1. Data refer to registered (“professional”) nurses in the United States,
States and Ireland (resulting in an over-estimation). Ireland and Chile (resulting in an over-estimation).
2. Data refer to registered (“professional”) nurses and unregistered nursing 2. Data refer to registered (“professional”) nurses and unregistered nursing
graduates. graduates.
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604818 12 http://dx.doi.org/10.1787/888933604837
8.18. Trend in the remuneration of hospital nurses in real terms, selected OECD countries, 2005-15
Australia Belgium Czech Republic France¹
Canada Netherlands¹ Greece² Hungary
United States Italy Mexico
Index (2005 = 100) Index (2005 = 100)
140 140
130 130
120 120
110 110
100 100
90 90
80 80
70 70
2005 2007 2009 2011 2013 2015 2005 2007 2009 2011 2013 2015
1. Index for France and the Netherlands, 2006 = 100.
2. Index for Greece, 2009 = 100.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604856
Health at a Glance 2017 © OECD 2017 163
8. HEALTH WORKFORCE
Foreign-trained doctors and nurses
International migration of doctors and nurses is not a new by the arrival of many nurses trained in Romania, who
phenomenon, but it has drawn considerable attention in now account for nearly half of all foreign-trained nurses.
recent years due to concerns that it might exacerbate In France the share of nurses trained abroad remains
shortages of skilled health workers in some countries. The low, but their numbers have been increasing, and many
Global Code of Practice on the International Recruitment of of these foreign-trained nurses are French citizens who
Health Personnel, adopted by the World Health Assembly received their diploma from Belgium. Israel has shown a
in May 2010, was designed to respond to these concerns. steady decline in the share of nurses trained abroad while
It provides an instrument for countries to promote a more increasing the number of domestic nursing graduates (see
ethical recruitment of health personnel, encouraging the indicator on “Nursing graduates”).
countries to achieve greater “self-sufficiency” in the training
of health workers, while recognising the basic human right
of every person to migrate. Definition and comparability
In 2015, the share of foreign-trained doctors ranged from
3% or less in Estonia, the Slovak Republic, the Netherlands, The data relate to foreign-trained doctors and nurses
Poland, Italy and Turkey, to more than 30% in Israel, New working in OECD countries measured in terms of total
Zealand, Ireland, Norway and Australia (Figure 8.19). The stocks. The OECD health database also includes data
very high proportion of foreign-trained doctors in Israel on the annual flows for most of the countries shown
reflects not only the importance of immigration in this here, as well as by country of origin. The data sources
country, but also that about one third of new licenses in most countries are professional registries or other
are issued to people born in Israel but trained abroad. In administrative sources.
Norway, roughly half of foreign-trained doctors are people The main comparability limitation relates to
who were born in the country but went to pursue their differences in the activity status of doctors and
medical studies in another country. In Luxembourg, all nurses. Some registries are regularly updated, making
doctors are foreign-trained, in the absence of a medical it possible to distinguish doctors and nurses who are
school in the country. still actively working in health systems, while other
Since 2000, the number and share of foreign-trained doctors sources include all doctors and nurses licensed to
has increased in many OECD countries (Figure 8.21). In the practice, regardless of whether they are still active
United States, the share has remained relatively stable over or not. The latter will tend to over-estimate not only
time, but the absolute number of doctors trained abroad has the number of foreign-trained doctors and nurses, but
continued to increase (OECD, 2016). Sweden has experienced also the total number of doctors and nurses (including
a strong rise in the number and share of foreign-trained the domestically-trained), making the impact on the
doctors, with most of these doctors coming from Germany, share unclear.
Poland and Iraq. The number and share of foreign-trained The data source in some countries includes interns
doctors has also increased in France and Germany, though and residents, while these physicians in training are
at a slower pace. In France, the rise is partly due to a fuller not included in other countries. Because foreign-
recognition of qualifications of foreign-trained doctors who trained doctors are often over-represented in the
were already working in the country, as well as the inflow categories of interns and residents, this may result in
of doctors from new EU member states. an under-estimation of the share of foreign-trained
In nearly all OECD countries, the proportion of foreign- doctors in countries where they are not included (e.g.,
trained nurses is much lower than that of foreign-trained Austria, France and Switzerland).
doctors. However, given that the overall number of nurses The data for Germany (on foreign-trained doctors) and
is usually much greater than the number of doctors, the some regions in Spain are based on nationality (or
absolute number of foreign-trained nurses tends to be place of birth in the case of Spain), not on the place
greater than for doctors (OECD, 2016). OECD countries vary of training.
widely in the number and share of foreign-trained nurses
working in their health system (Figure 8.20). While there
were almost no foreign-trained nurses working in Slovenia,
Turkey, the Netherlands and Estonia in 2015, they make up
References
over 25% of the nursing workforce in New Zealand, and
between 10% and 20% in Switzerland, Australia and the OECD (2016), Health Workforce Policies in OECD Countries:
United Kingdom. Right Jobs, Right Skills, Right Places, OECD Publishing, Paris,
The number and share of foreign-trained nurses has http://dx.doi.org/10.1787/9789264239517-en.
increased over the past ten years in several OECD countries, UN High-Level Commission on Health Employment and
including New Zealand, Australia and Canada (Figure 8.22). Economic Growth (2016), “Working for Health and Growth:
In Italy, an increase in the immigration of foreign-trained Investing in the Health Workforce”, WHO, Geneva,
nurses between 2000 and 2008 was primarily driven retrieved from http://www.who.int/hrh/com-heeg/reports.
164 Health at a Glance 2017 © OECD 2017
8. HEALTH WORKFORCE
Foreign-trained doctors and nurses
8.19. Share of foreign-trained doctors, 2015 8.20. Share of foreign-trained nurses, 2015
(or nearest year) (or nearest year)
Israel 57.9 New Zealand 25.4
New Zealand 42.1 Switzerland 18.7
Ireland 39.0 Australia 18.1
Norway 38.1
United Kingdom 14.4
Australia 32.4
Norway 9.1
Sweden 27.1
Switzerland 27.0 Israel 8.8
United Kingdom 26.9 Canada 7.7
United States 24.9 Germany 7.2
Canada 24.0 United States 6.0
Finland 19.9 OECD25 5.9
OECD28 16.9 Italy 5.7
Chile 16.1
Latvia 3.4
Slovenia 14.1
Belgium 3.2
Belgium 11.5
France 10.4
France 2.8
Germany¹ 10.3 Sweden 2.7
Spain¹ 9.4 Greece 2.5
Hungary 7.8 Chile 2.2
Latvia 6.5 Spain¹ 2.1
Czech Republic 6.1 Portugal 1.8
Denmark 5.3 Finland 1.8
Austria 5.1
Hungary 1.5
Estonia 3.0
Denmark 1.2
Slovak Republic 3.0
Netherlands 2.1 Slovenia 0.4
Poland 1.8 Turkey 0.3
Italy 0.8 Netherlands 0.1
Turkey 0.2 Estonia 0.1
0 10 20 30 40 50 60 0 5 10 15 20 25 30
% %
1. In Germany and some regions in Spain, the data are based on 1. The data for some regions in Spain are based on nationality or place
nationality (or place of birth in the case of Spain), not on the place of of birth, not on the place of training.
training. Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017. 12 http://dx.doi.org/10.1787/888933604894
12 http://dx.doi.org/10.1787/888933604875
8.21. Evolution in the share of foreign-trained 8.22. Evolution in the share of foreign-trained nurses,
doctors, selected OECD countries, 2000 to 2015 selected OECD countries, 2000 to 2015
France Germany Australia Canada
Norway Sweden France Israel
% United Kingdom United States % Italy New Zealand
40 30
35
25
30
20
25
20 15
15
10
10
5
5
0 0
14
15
00
01
02
03
04
05
06
09
10
11
12
13
14
15
07
08
00
01
02
03
04
05
06
07
08
09
10
11
12
13
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604913 12 http://dx.doi.org/10.1787/888933604932
Health at a Glance 2017 © OECD 2017 165
9. HEALTH CARE ACTIVITIES
Consultations with doctors
Medical technologies
Hospital beds
Hospital discharges
Average length of stay in hospitals
Hip and knee replacement
Caesarean sections
Ambulatory surgery
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.
Health at a Glance 2017 © OECD 2017 167
9. HEALTH CARE ACTIVITIES
Consultations with doctors
Consultations with doctors can take place in doctors’ offices where consultations with doctors in both primary care and
or clinics, in hospital outpatient departments or, in some hospital settings tend to be focused towards patients with
cases, in patients’ own homes. In many European countries more severe and complex cases.
(e.g., Denmark, Italy, Netherlands, Norway, Portugal, Slovak The number and type of doctor consultations can vary
Republic, Spain and the United Kingdom), patients are among different socio-economic groups in each country.
required or even incentivised to first consult a general An OECD study found that the probability of a visit to the
practitioner (GP) about any new episode of illness. The GP GP tends to be equally distributed in most countries, but in
may then refer them on to a specialist, if indicated. In other nearly all countries, higher income people are more likely
countries, patients may approach specialists directly. to see a specialist than those with low income, and also
In 2015, the number of doctor consultations per person more frequently (Devaux and de Looper, 2012).
ranged from less than 3 in Mexico and Sweden, to almost
13 and 16 in Japan and Korea respectively (Figure 9.1).
The OECD average was 6.9 consultations per person
per year, with most countries reporting between four Definition and comparability
and eight consultations. Cultural factors can play a role
in explaining some of the variations across countries, Consultations with doctors refer to the number of
although certain health system characteristics may also contacts with physicians, including both generalists
be important. Provider payment methods and the level of and specialists. There are variations across countries
co-payments are particularly relevant. For example, some in the coverage of these consultations, notably in
countries where doctors are paid on a fee-for-service basis outpatient departments of hospitals. The data come
tend to have above-average consultation rates (e.g. Japan mainly from administrative sources, although in some
and Korea), while countries with mostly salaried doctors countries (Ireland, Italy, Netherlands, New Zealand,
tend to have below-average rates (e.g. Mexico, Finland and Spain and Switzerland) the data come from health
Sweden). However, there are examples of countries such interview surveys. Estimates from administrative
as Switzerland and the United States where doctors are sources tend to be higher than those from surveys
paid mainly by fee-for-service and where consultation rates because of problems with recall and non-response
are below average. In these countries, patient co-payments rates.
can be high, which may result in patients not consulting In Hungary, the figures include consultations for
a doctor because of the cost of care (see the indicator on diagnostic exams such as CT and MRI scans (resulting
“Unmet needs for health care due to cost” in Chapter 5). in an over-estimation). The figures for the Netherlands
In Sweden and Finland, the low number of doctor exclude contacts for maternal and child care. The data
consultations may also be explained partly by the fact for Portugal exclude visits to private practitioners
that nurses and other health professionals play a more (resulting in an under estimation). In Germany, the
important role in providing primary care to patients in data include only the number of cases of physicians’
health centres, lessening the need for consultations with treatment according to reimbursement regulations
doctors (Delamaire and Lafortune, 2010). under the Social Health Insurance Scheme (a case only
counts the first contact over a three-month period,
The average number of doctor consultations per person even if the patient consults a doctor more often,
across the OECD has remained relatively stable since 2000 leading to an under-estimation). Telephone contacts
(from 6.5 to 6.9). But in some countries there have been are included in a few countries (e.g. Spain). In Turkey,
large increases over time (Korea, Turkey). In some other a majority of consultations with doctors occur in
countries, the number of doctor consultations per person outpatient departments in hospitals.
fell. This was the case in Japan, the Czech Republic and the
Slovak Republic, although the numbers remains well above
average in these three countries.
Information on the number of doctor consultations per References
person can be used to estimate the annual numbers of
consultations per doctor. This indicator should not be taken Delamaire, M.-L. and G. Lafortune (2010), “Nurses in
as a measure of doctors’ productivity, since consultations Advanced Roles: A Description and Evaluation of
can vary in length and effectiveness, and because it excludes Experiences in 12 Developed Countries”, OECD Health
the services doctors might deliver for hospital inpatients, Working Paper, No. 54, OECD Publishing, Paris, http://
as well as on administration and research. Keeping this in dx.doi.org/10.1787/5kmbrcfms5g7-en.
mind, the estimated number of consultations per doctor is Devaux, M. and M. de Looper (2012), “Income-related
highest in Korea and Japan, followed by Turkey and Hungary Inequalities in Health Service Utilisation in 19 OECD
(Figure 9.2). On the other hand, the estimated number of Countries”, OECD Health Working Papers, No. 58, OECD
consultations per doctor was lowest in Sweden and Norway, Publishing, Paris, http://dx.doi.org/10.1787/5k95xd6stnxt-en.
168 Health at a Glance 2017 © OECD 2017
0
2
4
6
8
10
12
14
16
18
0
1 000
2 000
3 000
4 000
5 000
6 000
7 000
8 000
Ko
Ko re
re Ja a 16.0
Ja a 7 140 Sl H pa
pa ov u n
a n 12.7
Tu n
r 5 385 C z k R gar y
Hu ke y e c ep 11.8
S n 4 651 h ub
Re lic
Sl o u t ga r Ru 11.4
ov h y 3 810 ss
pu
b
ak A f r G
i a er lic
Re i c a n 11.1
pu 3 414 F e ma
b d e ny
10.0
Annual consultations per doctor
ra
Po lic 3 303
Annual consultations per person
C l a Li tio
th n
n ua 10.0
Ru z e c C a d 3 179 ni
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
s s h R n ad
8.8
Health at a Glance 2017 © OECD 2017
ia Ne Tur a
n epu a
Fe b 3 024 t h ke
er y 8.4
de lic la
ra 3 010 n
Ge tion Ca ds 8.2
rm 2 526 na
d
S an S a 7.7
Ne lov y 2 416 A u pain
th eni
er a st 7.6
la 2 407 ra
OE nd s Po li a
l 7.4
CD 2 360 OE and
Be 32 CD 7.4
lg 2 295
2000
Be 32
A u ium lg 6.9
st 2 254 iu
m
L i r a li
th a 6.8
ua 2 101 Sl l y
It a
n ov 6.8
Fr i a
an 2 028 en
c A u ia 6.8
Lu S e 2 020 st
xe pa E s r ia
m in to 6.6
bo 2 000 n
u Fr i a 6.4
Ir e r g
la 1 995 an
Ic c e 6.3
Es nd el
to 1 983 an
ni Lu L a d 5.9
La a 1 870 xe t v
tv m ia
1. In Chile and Portugal, data for the denominator include all doctors licensed to practice.
ia bo 5.9
It a 1 845 u
l Ir e r g
l 5.8
Un C y
i t e hi 1 744 De nd a
d le¹ nm 5.7
1 682 a
2015
St
at Fi rk
Ic e s nl 4.4
el 1 624 an
an No d
4.3
Ne B d 1 619 rw
w raz
Z e il Un Por ay
t 4.3
al 1 487 ite ug
d
a a
Fi nd Sw Sta l 4.1
nl 1 363
a Ne er
it z tes
4.0
Au nd
s 1 310 w lan
Ze d
D e t r ia al 3.9
nm 1 295 an
d
9.2. Estimated number of consultations per doctor, 2015 (or nearest year)
a 3.7
M rk 1 230 Ch
e S w il e
Co x i c 3.4
9.1. Number of doctor consultations per person, 2000 and 2015 (or nearest year)
lo o 1 147 ed
m en
Po bi
a Br 2.9
Sw r tu 1 071 az
i t z g a l¹ S o M e il 2.8
er 1 000 ut xic
la h
N o nd A o 2.7
rw 996 Co fric
lo a 2.5
S w ay
976 m
ed bi
en a 1.9
692
169
12 http://dx.doi.org/10.1787/888933604970
12 http://dx.doi.org/10.1787/888933604951
Consultations with doctors
9. HEALTH CARE ACTIVITIES
9. HEALTH CARE ACTIVITIES
Medical technologies
New medical technologies are improving diagnosis and Clinical guidelines have been developed in several OECD
treatment, but they are also increasing health spending. countries to promote a more rational use of MRI and CT
This section presents data on the availability and use exams. In the United Kingdom, the National Institute for
of two diagnostic imaging technologies: computed Health and Clinical Excellence (NICE) has issued a number
tomography (CT) scanners and magnetic resonance of guidelines on the appropriate use of MRI and CT exams
imaging (MRI) units. CT and MRI exams help physicians (NICE, 2012). In the United States, a “Choosing Wisely”
diagnose a range of conditions. Unlike conventional campaign has developed clear guidelines for doctors and
radiography and CT scanning, MRI exams do not expose patients to reduce the use of unnecessary diagnostic tests
patients to ionising radiation. and procedures. The guidelines include, for instance,
The availability of CT scanners and MRI units has increased avoiding imaging studies such as MRI, CT or X-rays for
rapidly in most OECD countries over the past two decades. acute low back pain without specific indications (Choosing
Japan has, by far, the highest number of MRI and CT scanners Wisely, 2015). A similar “Choosing Wisely” campaign was
per capita, followed by the United States for MRI units and launched in Canada in 2014, and work has also started
by Australia for CT scanners (Figures 9.3 and 9.4). Germany, in several other OECD countries to produce similar clear
Greece, Iceland, Italy, Korea and Switzerland also have guidelines and recommendations to promote a more
significantly more MRI and CT scanners per capita than the efficient use of diagnostic tests and other procedures. It is
OECD average. The number of MRI units and CT scanners still too early to tell to what extent these campaigns will
per population is the lowest in Mexico, Hungary, Israel and succeed in reducing the overuse of MRI and CT exams.
the United Kingdom.
There is no general guideline or benchmark regarding the
ideal number of CT scanners or MRI units per population. Definition and comparability
However, if there are too few units, this may lead to access
The data in most countries cover MRI units and
problems in terms of geographic proximity or waiting
CT scanners installed both in hospitals and the
times. If there are too many, this may result in an overuse
ambulatory sector, but the coverage is more limited
of these costly diagnostic procedures, with little if any
in some countries. MRI units and CT scanners outside
benefits for patients.
hospitals are not included in Belgium, Portugal,
Data on the use of these diagnostic scanners are available Sweden and Switzerland (for MRI units). For the
for a smaller group of countries. Based on this more United Kingdom, the data only include equipment
limited country coverage, the number of MRI exams per in the public sector. For Australia and Hungary, the
capita is highest in Turkey, Germany, the United States, number of MRI units and CT scanners includes only
Japan and France, all of which have more than 100 MRI those eligible for public reimbursement.
exams per 1 000 population (Figure 9.5). In the United
Similarly, MRI and CT exams performed outside
States, the (absolute) number of MRI exams more than
hospitals are not included in Austria, Portugal,
doubled between 2000 and 2015. In Turkey, it has grown
Switzerland and the United Kingdom. In Australia, the
even faster, by three times between 2008 and 2015. In this
data only include exams for private patients (in or out
country, there is growing evidence that MRI exams are
of hospitals); while in Korea and the Netherlands they
being systematically prescribed for patients with various
only include publicly-financed exams.
health problems, resulting in overuse of these tests. The
number of CT exams per capita is highest in the United
States, followed by Japan and Luxembourg (Figure 9.6).
There are large variations in the use of CT and MRI References
scanners not only across countries, but also within
Choosing Wisely (2015), “Recommendations from the
countries. For example, in Belgium, there was almost
American Society of Anesthesiologists”, available at:
a two-fold variation in MRI and CT exams between
http://www.choosingwisely.org/clinician-lists/american-
provinces with the highest and lowest rates in 2010. In
society-anesthesiologists-imaging-studies-for-acute-low-
the United Kingdom (England), the utilisation of both
back-pain/.
types of diagnostic exams is generally much lower, but
the variation across regions is greater, with almost a four- NICE – National Institute for Health and Care Excellence
fold difference between the Primary Care Trusts that had (2012), Published Diagnostics Guidance, available at
the highest rates and lowest rates of MRI and CT exams in guidance.nice.org.uk/DT/Published.
2010/11. In Canada, there has been a strong rise in the use OECD (2014), Geographic Variations in Health Care: What Do
of both MRI and CT exams in all parts of the country over We Know and What Can Be Done to Improve Health System
the past decade, but there continues to be wide variations Performance?, OECD Publishing, Paris, http://dx.doi.
across provinces (OECD, 2014). org/10.1787/9789264216594-en.
170 Health at a Glance 2017 © OECD 2017
9. HEALTH CARE ACTIVITIES
Medical technologies
9.3. MRI units, 2015 (or nearest year) 9.4. CT scanners, 2015 (or nearest year)
Japan 51.7 Japan 107.2
United States 39.0 Australia² 59.6
Germany 33.6 United States 41.0
Italy 28.2 Iceland 39.3
Korea 26.3 Denmark 37.7
Finland 25.9 Korea 37.0
Greece 24.3 Latvia 36.9
Switzerland¹ 22.0 Switzerland 36.2
Iceland 21.2 Greece 35.1
Austria 20.7 Germany 35.1
OECD33 15.9 Italy 33.3
Spain 15.9 Austria 29.0
Sweden¹ 14.6 OECD34 25.7
Australia² 14.5 Belgium¹ 22.9
Ireland 14.1 Portugal¹ 22.5
New Zealand 13.3 Finland 21.5
Latvia 12.6 Lithuania 21.0
France 12.6 Sweden¹ 20.3
Spain 18.0
Netherlands 12.5
Slovak Republic 17.9
Luxembourg 12.3 New Zealand 17.8
Estonia 12.2 Ireland 17.8
Belgium¹ 11.7 Luxembourg 17.6
Lithuania 11.0 Poland 17.2
Turkey 10.2 Estonia 16.7
Canada 9.5 France 16.6
Chile 9.4 Czech Republic 16.1
Slovak Republic 8.9 Brazil 15.3
Slovenia 8.7 Canada 15.0
Czech Republic 8.3 Chile 14.8
Poland 7.6 Turkey 14.3
Portugal¹ 7.2 Netherlands 13.8
United Kingdom 7.2 Slovenia 13.1
Brazil 6.8 Russian Federation 12.8
Russian Federation 4.7 Israel 9.8
Israel 4.1 United Kingdom 9.5
Hungary² 3.6 Hungary² 8.4
Mexico 2.4 Mexico 5.9
0 10 20 30 40 50 60 0 20 40 60 80 100 120
Per million population Per million population
1. Equipment outside hospital not included. 1. Equipment outside hospital not included.
2. Only equipment eligible for public reimbursement. 2. Only equipment eligible for public reimbursement.
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933604989 12 http://dx.doi.org/10.1787/888933605008
9.5. MRI exams, 2015 (or nearest year) 9.6. CT exams, 2015 (or nearest year)
Turkey 144.3 United States 245.3
Germany 131.3 Japan 230.8
United States 117.8 Luxembourg 207.7
Japan 112.3 Belgium 198.8
France 104.8 France 197.4
Belgium 85.6 Iceland 190.9
Luxembourg 82.5 Greece 181.1
Denmark 82.1 Korea³ 174.9
Iceland 81.9 Turkey 174.8
Spain 78.3 Latvia 168.5
Switzerland¹ 69.9 Portugal¹ 165.3
Greece 67.9 Denmark 161.8
OECD29 64.8 Slovak Republic 156.2
Slovak Republic 56.8 Canada 152.8
Canada 55.5 Estonia 151.9
Austria¹ 55.0 Germany 143.8
United Kingdom¹ 52.6 OECD29 143.1
Netherlands³ 51.8 Austria¹ 142.2
Estonia 48.8 Israel 131.3
Czech Republic 48.0 Australia² 119.6
Latvia 42.6 Hungary 107.1
Slovenia 42.4 Spain 104.9
Australia² 41.0 Czech Republic 101.9
Finland 39.2 Switzerland¹ 100.3
Hungary 38.1 Poland 97.0
Portugal¹ 36.3 Chile 86.5
Israel 34.8 Netherlands³ 80.8
Korea³ 31.2 United Kingdom¹ 79.3
Poland 28.4 Slovenia 61.8
Chile 18.6 Finland 34.2
0 25 50 75 100 125 150 0 50 100 150 200 250
Per 1 000 population Per 1 000 population
1. Exams outside hospital not included. 1. Exams outside hospital not included.
2. Exams on public patients not included. 2. Exams on public patients not included.
3. Exams privately-funded not included. 3. Exams privately-funded not included.
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605027 12 http://dx.doi.org/10.1787/888933605046
Health at a Glance 2017 © OECD 2017 171
9. HEALTH CARE ACTIVITIES
Hospital beds
The number of hospital beds provides a measurement of beds stood at 76% on average across OECD countries in
the resources available for delivering services to inpatients 2015, only slightly above the 2000 level (Figure 9.9). This is
in hospitals. This section presents data on the number of because the general increase in occupancy rates (driven
overall hospital beds in 2000 and 2015 and for different by the reduction in number of beds) is offset by a few
types of care (curative care, rehabilitative care, long-term large decreases in occupancy rates observed in Norway,
care and other functions). It also presents an indicator of Japan and Latvia, along with some smaller decreases in
bed occupancy rates over time, focussing on curative care Switzerland, Germany, the Slovak Republic, Korea and
beds. more. Ireland and Israel had the highest rate of hospital
Among OECD countries, the number of hospital beds per bed occupancy at approximately 94%, followed by Canada
capita remains highest in Japan and Korea, with 13.2 and at 92% and the United Kingdom at 84%.
11.5 beds per 1 000 population in 2015 (Figure 9.7). In both
countries, hospitals have so-called “social admissions”,
that is, a significant part of hospital beds are devoted to
long-term care to tackle the increasing number of ageing Definition and comparability
population. The number of hospital beds is also well above
Hospital beds are defined as all beds that are regularly
the OECD average in the Russian Federation, Germany
maintained and staffed and are immediately available
and Austria. On the other hand, some of the key partner
for use. They include beds in general hospitals, mental
countries in Asia (India and Indonesia) have very few
health and substance abuse hospitals, and other
hospital beds compared to the OECD average. This is also
specialty hospitals. Beds in residential long-term care
the case for countries in Latin America (Mexico, Colombia,
facilities are excluded (OECD, 2017).
Chile and Brazil).
Curative care beds accommodate patients where the
The number of hospital beds per capita has decreased over
principal intent is to do one or more of the following:
the past decade in most OECD countries, falling on average
cure illness or provide definitive treatment of injury,
from 5.6 per 1 000 population in 2000 to 4.7 in 2015. This
perform surgery, relieve symptoms of illness or
reduction is part of a voluntary effort in most countries,
injury (excluding palliative care), reduce severity of
partly driven by progress in medical technology, which has
illness or injury, protect against exacerbation and/
enabled a move to day surgery for a number of procedures
or complication of illness and/or injury which could
and a reduced need for hospitalisation. In many European
threaten life or normal functions, perform diagnostic
countries, the financial and economic crisis, which started
or therapeutic procedures, manage labour (obstetric).
in 2008, provided an additional stimulus to reduce hospital
In some countries, these beds include all (curative and
capacity in line with policies to reduce public spending on
non-curative) psychiatric care beds.
health. Only in Korea, China and Turkey have the numbers
of hospital beds per capita grown since 2000. Generally, the Rehabilitative care beds accommodate patients with
largest decreases in the number of beds over time have the principal intent to stabilise, improve or restore
been observed in countries with an initially high number impaired body functions.
of beds in 2000. Long-term care beds are hospital beds accommodating
On average, about three-quarters of hospital beds (77%) are patients requiring long-term care due to chronic
allocated for curative care across OECD countries (Figure 9.8). impairments and a reduced degree of independence
The rest are distributed between long-term care (12%), in activities of daily living. They include beds in long-
rehabilitation (9%), and other types of care (2%). However, in term care departments of general hospitals, beds for
some countries, the share of beds allocated for rehabilitation long-term care in specialty hospitals, and beds for
and long-term care is much greater than the average. In palliative care.
Korea and Japan, for the reasons previously mentioned, The occupancy rate for curative (acute) care beds is
37% and 20% of hospital beds, respectively, are allocated calculated as the number of hospital bed-days related
for long-term care. In Finland, this share is also relatively to curative care divided by the number of available
high (28%), as local governments (municipalities) use beds curative care beds (multiplied by 365).
in health care centres (which are defined as hospitals) for
at least some of the institutional long-term care needs. In
France, Germany and Poland, around a quarter of all hospital
beds are devoted to rehabilitative care.
References
In several countries, the reduction in the number of
hospital beds has been accompanied by an increase in OECD (2017), OECD Health Statistics 2017, OECD Publishing,
their occupancy rates. The occupancy rate of curative care Paris, http://dx.doi.org/10.1787/health-data-en.
172 Health at a Glance 2017 © OECD 2017
0
3
6
9
12
15
Ru
50
60
70
80
90
100
%
0
20
40
60
80
100
%
Ir e Ja ss Ja
la pa ia pa
n
nd n¹ F e Ko n 13.2
94.7 Ko de re
Is re r a 11.5
ra
el
Ge a Ge atio
94.0 rm rm n 8.5
Un Ca an
ite na A u a ny
d da Au y 8.1
91.6
Per 1 000 population
Ki st Hu s t r i
ng H u r ia n a 7.6
Sw do L i ga
m ng th r y
7.0
it z 84.3 ar ua
er Cz y Cz
e c P nia
la e c Pol 7.0
nd h an h ola
83.3 Re d Re n
No pu pu d 6.6
rw b B e blic
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
ay
80.4
lg 6.5
Health at a Glance 2017 © OECD 2017
Ge B e lic Sl iu
rm lg ov
iu ak Fr a m 6.2
Curative care
an Sl m
y 79.8 ov Fr Re n c
ak a n pu e 6.1
It a Re c e b
ly L a lic 5.8
Be 78.9
pu
bl t
i Lu E s v i a
lg
iu xe t o 5.7
m La c m ni a
78.4 tv
ia¹ b 5.0
Ch Es O E ou r
il e to Sw C g 4.8
78.2 n it z D3
M OE ia
4.7
2000
er 5
2000
ex Lu C D
ic 3
Sl lan
o 77.9
xe
m 2
ov d 4.6
e
Sp S w bou F i nia
ai nl 4.5
n it z rg a
OE 75.8 er 4.4
N e Gr n d
CD la th ee
S l nd er ce
4.3
27
Rehabilitative care
75.7 ov la
Fr en nd
an F i ia Ch s 4.2
ce A u in
75.1
nl
Ja
an st a
r 3.9
pa G d No a li a
n Ne ree r 3.8
Au 74.5 th c Po w a
Cz
ec st er e
la
r tu y 3.8
h r ia nd ga
Re 74.3 No s It a l 3.4
pu rw Ic l y
bl
ic el 3.2
74.3 Po ay an
Note: Countries are ranked from highest to lowest total number of hospital beds per capita.
Gr r tu Is d 3.1
Lu ee ga r
xe ce
73.6
l Ir e a e l 3.0
m la
bo It a Un n
3.0
Long-term care
ur Ic l y i t e Sp d
g 71.6 el Ne S ain d
La an 3.0
d w tat
2015
2015
tv Ze es
ia
70.7
Is al 2.8
Hu ra
el a
ng Ir e Un Tu nd 2.7
ar la
y 69.3
i t e C r ke y
Sl
ov
nd d an 2.7
en Un S K i ad
ng a
ia i t e pai 2.6
68.8 d n
1. In Japan and Latvia, psychiatric care beds are reported in “other beds” rather than in the more specific categories.
Sl De dom
Tu nm 2.6
ov r ke
Ne Sta
w te
9.8. Hospital beds by function of health care, 2015 (or nearest year)
ak
Re y 68.8 Ze s S o S w ar k
2.5
al ut ed
9.7. Hospital beds per 1 000 population, 2000 and 2015 (or nearest year)
pu an h en
bl
ic d Af 2.4
Es 68.7 Tu ric
rk Br a 2.3
to
9.9. Occupancy rate of curative (acute) care beds, 2000 and 2015 (or nearest year)
ey az
ni Ca
a 67.0 n il 2.3
Other hospital beds
Co Ch
Ko D e ad a lo il e
re nm m 2.1
Po a 64.4 a M bia
r tu Sw rk In e x i 1.6
Un do c o
ga ed
ite l 64.0 ne 1.5
d en
si
St
at Ch In a 1.0
es il e di
62.8 a 0.5
173
12 http://dx.doi.org/10.1787/888933605103
12 http://dx.doi.org/10.1787/888933605084
12 http://dx.doi.org/10.1787/888933605065
Hospital beds
9. HEALTH CARE ACTIVITIES
9. HEALTH CARE ACTIVITIES
Hospital discharges
Hospital discharge rates measure the number of patients interventions often gradually extends to older population
who leave a hospital after staying at least one night. groups, as interventions become safer and more effective
Together with the average length of stay, they are important for people at older ages. But the diffusion of new medical
indicators of hospital activities. Hospital activities are technologies may also involve a reduction in hospitalisation
affected by a number of factors, including the capacity if it involves a shift from procedures requiring overnight
of hospitals to treat patients, the ability of the primary stays in hospitals to same-day procedures. In the group
care sector to prevent avoidable hospital admissions, of countries where discharge rates have decreased since
and the availability of post-acute care settings to provide 2000, there has been a strong rise in the number of day
rehabilitative and long-term care services. surgeries (see indicator on “Ambulatory surgery”). The
In 2015, hospital discharge rates were highest in Austria and number of beds available in a hospital might also affect
Germany, followed by Lithuania and the Russian Federation the timing of patient discharges, which in turn affects the
(Figure 9.10). They were the lowest in Colombia, Mexico, average length of stay (see indicator on “Average length of
Brazil and Canada. In general, those countries that have stay in hospitals”).
more hospital beds tend to have higher discharge rates. Hospital discharge rates vary not only across countries,
For example, the number of hospital beds per capita in but also within countries. In several OECD countries (e.g.,
Austria and Germany is more than two-times greater than Canada, Finland, Germany, Italy, Portugal, Spain and the
in Canada and Spain, and discharge rates are also more United Kingdom), hospital medical admissions (excluding
than two-times larger (see indicator on “Hospital beds”). admissions for surgical interventions) vary by more
Across OECD countries, the main conditions leading to than two-times across different regions in the country
hospitalisation in 2015 were circulatory diseases, pregnancy (OECD, 2014).
and childbirth, injuries and other external causes, diseases
of the digestive system, cancers, and respiratory diseases.
Germany, Austria, Hungary and Latvia have the highest
Definition and comparability
discharge rates for circulatory diseases; with Austria, Greece,
Germany and Hungary the highest for cancers (Figures 9.11 Discharge is defined as the release of a patient who
and 9.12). While the high rates of hospital discharges for has stayed at least one night in hospital. It includes
circulatory diseases in Hungary are associated with lots deaths in hospital following inpatient care. Same-day
of people having heart and other circulatory diseases discharges are usually excluded, with the exceptions
(see indicator on “Mortality from circulatory diseases” in of Chile, Japan, Norway, the Slovak Republic and
Chapter 3), this is not the case for Germany and Austria. the United States which include some same-day
Similarly, cancer incidence is not higher in Austria, Germany separations.
or Greece than in most other OECD countries (see indicator
Healthy babies born in hospitals are excluded
on “Cancer incidence” in Chapter 3). In Austria, the high
from hospital discharge rates in several countries
discharge rate is associated with a high rate of hospital
(Australia, Austria, Canada, Chile, Estonia, Finland,
readmissions for further investigation and treatment of
Greece, Ireland, Luxembourg, Mexico, Norway).
cancer patients (European Commission, 2008).
These comprise around 3 to 10% of all discharges.
Trends in hospital discharge rates vary widely across OECD Data for some countries do not cover all hospitals.
countries. Since 2000, discharge rates have increased in For instance, data for Mexico, New Zealand and the
some countries where discharge rates were low in 2000 United Kingdom are restricted to public or publicly-
and have increased rapidly since then (e.g. Korea, Turkey funded hospitals only. Data for Ireland cover public
and China) as well as in other countries such as Germany acute and psychiatric (public and private) hospitals.
where it was already above-average. In other countries Data for Canada and the United States include only
(e.g. France, Portugal and the United States), they have acute care/short-stay hospitals.
remained relatively stable, while in other countries (e.g.
Finland, Hungary, Iceland, Italy and Latvia), discharge rates
fell between 2000 and 2015.
Trends in hospital discharges reflect the interaction of References
several factors. Demand for hospitalisation may grow
as populations’ age, given that older population groups European Commission (2008), Hospital Data Project Phase 2,
account for a disproportionately high percentage of Final Report, European Commission, Luxembourg.
hospital discharges. However, population ageing alone OECD (2014), Geographic Variations in Health Care: What Do
may be a less important factor in explaining trends in We Know and What Can Be Done to Improve Health System
hospitalisation rates than changes in medical technologies Performance?, OECD Publishing, Paris, http://dx.doi.org/
and clinical practices. The diffusion of new medical 10.1787/9789264216594-en.
174 Health at a Glance 2017 © OECD 2017
9. HEALTH CARE ACTIVITIES
Hospital discharges
9.10. Hospital discharges, 2015 (or nearest year)
Per 1 000 population
300
255.8
255.3
238.6
228.5
250
206.4
200.5
200.1
196.5
187.0
184.6
183.6
200
174.8
171.5
171.5
171.1
169.6
166.2
165.8
164.4
164.1
158.9
156.0
153.0
153.1
147.8
145.7
143.8
140.6
131.9
150
125.5
124.1
118.6
116.5
114.4
113.9
109.2
93.2
83.5
100
55.2
49.9
34.4
50
0
Fr nia
n²
ak ep n
pu lic
S w s tr c e
rm a¹
n hu y
S l e ch er a i a
ng ²
Gr a r y
L a e¹
ov a
er ¹
E s l a nd
Tu ia¹
Po ey
F i land
rw m
Is a
C l
Sw D 3 5
Ch n
Lu Den in a
N e mb ar k
a ¹
i t e Ir l a n d
i t e ngd ¹
d om
Ja es²
er l y
Sp s
Ic a i n
r tu d
C gal
e l
a
lo o¹
¹
M r a zi
OE r ae
i t z a li a
Ko ¹ ²,
Ze rg
na ¹ ²,
B ¹ ³,
Un K i n d
B e and
Hu blic
ia Lit an
bi
Sl t v i
re
nd
ov R tio
Po lan
C z F ed an
Ge stri
th It a
c
N o l gi u
Co x i c
pa
rk
Au an
Re ub
ed
da
C a hil e
ay
w ou
e
ee
m
at
xe m
d ela
to
la
e
nl
St
Au
Ne
Un
ss
Ru
1. Data exclude discharges of healthy babies born in hospital (between 3-10% of all discharges).
2. Data include same-day discharges.
3. Data for Canada include discharges for curative (acute) care only.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605122
9.11. Hospital discharges for circulatory diseases, 2015 9.12. Hospital discharges for cancers, 2015
(or nearest year) (or nearest year)
Germany 37.4 Austria 28.6
Austria 33.5 Greece 24.4
Hungary 33.5 Germany 24.4
Latvia 31.5 Hungary 23.7
Slovak Republic 29.7 Japan 23.3
Estonia 28.6 Estonia 20.1
Czech Republic 27.2 Slovenia 18.2
Poland 27.1 Latvia 17.4
Greece 26.3 Korea 17.2
Finland 26.0 Slovak Republic 17.1
Slovenia 21.1 Finland 15.2
Norway 19.9 Norway 14.7
Sweden 19.5 Czech Republic 14.4
Belgium 19.2 Switzerland 14.2
France 19.1 Luxembourg 13.4
OECD35 19.0 Poland 13.3
Italy 18.8 OECD35 13.1
Switzerland 18.4 Australia 11.3
United States 18.2 Belgium 11.2
Luxembourg 16.5 Italy 11.1
Netherlands 16.4 Netherlands 11.0
Japan 16.2 France 11.0
Australia 15.9 Denmark 10.7
Denmark 15.7 Sweden 10.5
Turkey 14.3 Spain 9.8
New Zealand 13.5 Iceland 9.7
Spain 13.2 United Kingdom 7.9
United Kingdom 12.1 New Zealand 7.6
Israel 11.8 Portugal 7.4
Ireland 11.7 Ireland 6.9
Korea 11.2 Chile 6.7
Portugal 11.1 Turkey 6.5
Iceland 11.1 Israel 5.8
Canada 10.4 Canada 5.7
Chile 7.2 United States 5.1
Mexico 2.3 Mexico 3.6
0 10 20 30 40 0 10 20 30 40
Per 1 000 population Per 1 000 population
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605141 12 http://dx.doi.org/10.1787/888933605160
Health at a Glance 2017 © OECD 2017 175
9. HEALTH CARE ACTIVITIES
Average length of stay in hospitals
The average length of stay in hospitals is often regarded (OECD, 2013). Other options include promoting the uptake
as an indicator of efficiency. All else being equal, a shorter of less invasive surgical procedures, the expansion of early
stay will reduce the cost per discharge and shift care from discharge programmes which enable patients to return
inpatient to less expensive post-acute settings. Longer stays home to receive follow-up care, and support for hospitals
can be indicative of poor-value care: inefficient hospital to improve care co-ordination.
processes may cause delays in providing treatment; errors A few countries also collect data on delayed discharges –
and poor-quality care may mean patients need further the number of days that people stay in hospital after a
treatment or recovery time; poor care co-ordination may doctor declares them ready to be discharged or transferred.
leave people stuck in hospital waiting for ongoing care This provides a more precise measure of when a stay in
to be arranged. At the same time, some people may be a hospital is unnecessarily long. Denmark reported just
discharged too early, when staying in hospital longer could under 10 additional bed days per 1 000 population in 2014,
have improved their outcomes or reduced chances of re- a figure that has been relatively stable over time. Norway
admission. saw a sharp drop in delayed discharges, from 28 additional
In 2015, the average length of stay in hospitals for all causes bed days per 1 000 population in 2011 to about 12 in 2015.
across OECD countries was about eight days (Figure 9.13). Within the United Kingdom, England saw a significant
Turkey and Mexico had the shortest stays, with about four increase since 2013, reaching over 30 additional bed days
days, whereas Japan and Korea had the longest stays, with per 1 000 population in 2015. In England, this increase
over 16 days. In most countries, the average length of stay largely reflects ongoing health or social care services not
has fallen since 2000, with reductions particularly large being ready to receive patients (OECD 2017).
in Japan, Switzerland, the United Kingdom and Israel.
However, the average length of stay increased in Korea
and Hungary, with very slight increases in Italy, Canada
and South Africa. Definition and comparability
Focusing on specific diseases or conditions can remove
some of the effect of different case mix and severity. Average length of stay refers to the average number
Average length of stay following birth by normal delivery of days that patients spend in hospital. It is generally
was slightly less than three days on average in 2015 measured by dividing the total number of days
(Figure 9.14). This ranged from less than two days in stayed by all inpatients during a year by the number
Mexico, Turkey, the United Kingdom, Canada, Iceland and of admissions or discharges. Day cases are excluded.
the Netherlands, to around five days in the Slovak Republic The data cover all inpatient cases (including not only
and Hungary. In almost all OECD countries, the average curative/acute care cases) for most countries, with
length of stay following a delivery has fallen since 2000. the exceptions of Canada, Japan and the Netherlands
where the data refer to curative/acute care only
The average length of stay following acute myocardial
(resulting in an under-estimation).
infarction was 6.5 days on average in 2015. It was shortest
in Scandinavian countries (Norway, Denmark and Sweden), Healthy babies born in hospitals are excluded from
Turkey and the Slovak Republic, at fewer than five days, hospital discharge rates in several countries (Australia,
and highest in Chile and Germany, at more than ten days Austria, Canada, Chile, Estonia, Finland, Greece,
(Figure 9.15). Average length of stay following acute Ireland, Luxembourg, Norway, Mexico), resulting in a
myocardial infarction has fallen in all OECD countries slight over-estimation of the length of stay (e.g. the
since 2000, with reductions particularly marked in Austria, inclusion of healthy newborns would reduce the ALOS
Finland and the Slovak Republic. by 0.5 days in Canada). These comprise around 3 to
10% of all discharges.
Beyond differences in clinical need, several factors can
explain these cross-country variations. The combination of Data for normal delivery refer to ICD-10 code O80; for
an abundant supply of beds with the structure of hospital AMI they refer to ICD-10 codes I21-I22.
payments may provide hospitals with incentives to keep
patients longer. A growing number of countries (France,
Germany, Poland) have moved to prospective payment
methods, often based on diagnosis-related groups (DRGs), References
to set payments based on the estimated cost of hospital care
in advance of service provision. These payment methods OECD (2017). Tackling Wasteful Spending on Health,
encourage providers to reduce the cost of each episode of OECD Publishing, Paris, http://dx.doi.org/10.1787/
care. In Switzerland, cantons which moved from per diem 9789264266414-en.
payments to DRG-based payments have experienced a OECD (2013), OECD Reviews of Health Care Quality: Denmark
reduction in their lengths of stay (OECD and WHO, 2011). 2013: Raising Standards, OECD Publishing, Paris, http://
Strategic reductions in hospital bed numbers alongside dx.doi.org/10.1787/9789264191136-en.
development of community care services can also be OECD and WHO (2011), OECD Reviews of Health Systems:
expected to shorten average length of stay, as seen in Switzerland 2011, OECD Publishing, Paris, http://dx.doi.
Denmark’s quality-driven reforms of the hospital sector org/10.1787/9789264120914-en.
176 Health at a Glance 2017 © OECD 2017
9. HEALTH CARE ACTIVITIES
Average length of stay in hospitals
9.13. Average length of stay in hospital, 2000 and 2015 (or nearest year)
2000 2015
Days
25
20
16.1
16.5
15
9.5
11.4
10
7.8
10.1
7.4
9.6
9.4
9.3
6.5
9.1
6.3
9.0
8.8
8.5
8.4
8.3
7.9
7.8
7.6
7.6
7.3
7.3
7.2
7.0
7.0
6.8
6.8
6.8
4.2
6.7
6.4
5
6.2
6.1
5.9
5.7
5.5
5.5
5.5
3.9
0
Ir e i c a
Tu ico
Fr on
m lic
F e Ko ¹
ra a
C e
i t e G blic
S o No nia
a
ec Fi ar y
Lu Rep n d
rm g
r tu y
S w Au al
er a
L nd
th ia
a
CD y
Be 3 5
E s ium
Ca onia
Po da¹
ak S nd
pu n
ng e
m
a l
ov d
h ay
N e Ic a n d
Un h e r l a n d
St ¹
Sw a tes
en
s le
nm ia
lo r k
M bia
ey
Ze ae
d ds
n
Po an
it z stri
c
Hu hin
de re
ni
Ki ec
OE It al
Re pai
G e ou r
Sl lan
A u Chi
D e tr a l
g
Li at v
pa
do
Co a
rk
an
ut r w
xe ub
ed
la
ti
la
h nla
r
w Isr
m
ua
ex
na
ite lan
g
d re
l
lg
Af
t e
t
Ja
Ne
n
ov
ia
Cz
Un
Sl
ss
Ru
1. Data refer to average length of stay for curative (acute) care (resulting in an under-estimation). In Japan, the average length of stay for all inpatient
care was 29 days in 2015 (down from 39 days in 2000).
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605179
9.14. Average length of stay for normal delivery, 2015 9.15. Average length of stay for acute myocardial
(or nearest year) infarction (AMI), 2015 (or nearest year)
Hungary 5.0 Chile 10.2
Slovak Republic 4.9 Germany 10.2
Czech Republic 4.2 Estonia 9.1
France 4.0 Austria 8.2
Luxembourg 4.0 Italy 7.9
Poland 3.8 Hungary 7.8
Austria 3.7 Portugal 7.7
Slovenia 7.5
Belgium 3.7
Korea 7.4
Slovenia 3.6
Luxembourg 7.0
Italy 3.4 Spain 7.0
Latvia 3.4 Belgium 6.9
Switzerland 3.2 Ireland 6.9
Germany 3.1 Switzerland 6.8
Norway 3.1 United Kingdom 6.8
Greece 3.0 New Zealand 6.7
Israel 3.0 OECD34 6.5
Finland 2.9 Latvia 6.5
OECD32 2.9 Mexico 6.5
Chile 2.7 Finland 6.1
Australia 2.6 Poland 6.1
Denmark 2.4 Greece 6.0
Ireland 2.4 Czech Republic 5.9
Korea 2.4 France 5.9
Spain Israel 5.9
2.4
Netherlands 5.6
Sweden 2.3
Iceland 5.5
New Zealand 2.1
United States 5.4
United States 2.0 Australia 5.2
Netherlands 1.9 Canada 5.2
Iceland 1.7 Slovak Republic 4.9
Canada 1.5 Turkey 4.8
United Kingdom 1.5 Sweden 4.3
Turkey 1.4 Denmark 4.0
Mexico 1.1 Norway 3.7
0 2 4 6 0 4 8 12
Days Days
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605198 12 http://dx.doi.org/10.1787/888933605217
Health at a Glance 2017 © OECD 2017 177
9. HEALTH CARE ACTIVITIES
Hip and knee replacement
Significant advances in surgical treatment have provided 2000. Similarly, knee surgeries have seen a large increase in
effective options to reduce the pain and disability associated the past decades in all OECD countries, with the exception
with certain musculoskeletal conditions. Joint replacement of Chile and Estonia, which showed small decreases in the
surgery (hip and knee replacement) is considered the most past few years.
effective intervention for severe osteoarthritis and hip
fractures, reducing pain and disability and restoring some
patients to near normal function.
Definition and comparability
Osteoarthritis is one of the ten most disabling diseases
in developed countries. Worldwide, estimates show that Hip replacement is a surgical procedure in which
10% of men and 18% of women aged over 60 years have the hip joint is replaced by a prosthetic implant. It is
symptomatic osteoarthritis, including moderate and generally conducted to relieve arthritis pain or treat
severe forms (WHO, 2014). Age is the strongest predictor severe physical joint damage following hip fracture.
of the development and progression of osteoarthritis. It
Knee replacement is a surgical procedure to replace
is more common in women, increasing after the age of 50
the weight-bearing surfaces of the knee joint in order
especially in the hand and knee. Other risk factors include
to relieve the pain and disability of osteoarthritis. It
obesity, physical inactivity, smoking, excessive alcohol
may also be performed for other knee diseases such
consumption and injuries. While joint replacement surgery
as rheumatoid arthritis.
is mainly carried out among people aged 60 and over, it can
also be performed on people at younger ages. Classification systems and registration practices vary
across countries, which may affect the comparability of
In 2015, Switzerland, Germany, Austria and Belgium had
the data. While most countries include both total and
the highest rates for both of hip and knee replacement
partial replacement, some countries only include total
(Figures 9.16 and 9.17). In Mexico and Chile, the rates of hip
hip replacement. In Ireland, Mexico, New Zealand and
and knee replacement are particularly low, with less than
the United Kingdom, the data only include activities in
40 hip replacements and less than 10 knee replacements
publicly-funded hospitals, therefore underestimating
per 100 000 population. Differences in population structure
the number of total procedures presented here (for
may explain part of this variation across countries, and
example, approximately 15% of all hospital activity
age standardisation reduces it to some extent. Still, large
in Ireland is undertaken in private hospitals). Data
differences persist and the country ranking does not
for Portugal relate only to public hospitals on the
change significantly after age standardisation (McPherson
mainland. Data for Spain only partially include
et al., 2013; OECD 2014).
activities in private hospitals.
National averages can mask important variation in
hip and knee replacement rates within countries. In
Australia, Canada, Germany, France and Italy, the rate
of knee replacement is more than two times higher in References
certain regions compared with others, even after age-
standardisation (OECD, 2014). McPherson, K., G. Gon and M. Scott (2013), “International
Variations in a Selected Number of Surgical Procedures”,
The number of hip and knee replacements has increased
OECD Health Working Papers, No. 61, OECD Publishing,
rapidly since 2000 in most OECD countries (Figures 9.18 and
Paris, http://dx.doi.org/10.1787/5k49h4p5g9mw-en.
9.19). On average, the rate of hip replacement increased
by 30% between 2000 and 2015 and the rate of knee OECD (2014), Geographic Variations in Health Care: What Do
replacement nearly doubled. For hip replacement, most We Know and What Can Be Done to Improve Health System
OECD countries show increasing trends of varying degrees, Performance?, OECD Publishing, Paris, http://dx.doi.
but countries like Ireland and Portugal show much slower org/10.1787/9789264216594-en.
growth than the average, with Ireland being the only OECD WHO (2014), “Chronic Rheumatic Conditions”, Fact Sheet,
country to show a decrease in hip replacement rates from Geneva, available at: www.who.int/chp/topics/rheumatic/en/.
178 Health at a Glance 2017 © OECD 2017
9. HEALTH CARE ACTIVITIES
Hip and knee replacement
9.16. Hip replacement surgery, 2015 (or nearest year) 9.17. Knee replacement surgery, 2015 (or nearest year)
Switzerland 308 Switzerland 240
Germany 299 United States 226
Austria 271 Austria 215
Belgium 255 Germany 206
Finland 247 Belgium 202
Norway 244 Australia 197
France 241 Luxembourg 189
Denmark 237 Finland 187
Sweden 234 Canada 178
Netherlands 228 Denmark 168
Luxembourg 207 France 160
United States 204 United Kingdom 149
Australia 184 Netherlands 138
United Kingdom 182 Iceland 132
Czech Republic 180 Czech Republic 126
Slovenia 173
OECD33 126
Italy 172
Sweden 124
OECD34 166
Korea 121
New Zealand 161
Spain 120
Estonia 156
Canada 148 New Zealand 112
Iceland 147 Italy 112
Latvia 142 Slovenia 107
Hungary 138 Slovak Republic 103
Greece 132 Norway 101
Ireland 131 Estonia 94
Slovak Republic 123 Hungary 87
Spain 113 Latvia 82
Poland 112 Turkey 67
Portugal 91 Portugal 62
Israel 62 Israel 55
Korea 53 Ireland 52
Turkey 44 Poland 40
Chile 33 Chile 10
Mexico 8 Mexico 3
0 50 100 150 200 250 300 350 0 50 100 150 200 250 300
Per 100 000 population Per 100 000 population
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605236 12 http://dx.doi.org/10.1787/888933605255
9.18. Hip replacement surgery trends, 9.19. Knee replacement surgery trends,
2000 to 2015 (or nearest year) 2000 to 2015 (or nearest year)
Australia Finland Austria France
France Portugal Ireland United Kingdom
OECD29 OECD26
Per 100 000 population Per 100 000 population
250 250
200 200
150 150
100 100
50 50
0 0
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605274 12 http://dx.doi.org/10.1787/888933605293
Health at a Glance 2017 © OECD 2017 179
9. HEALTH CARE ACTIVITIES
Caesarean sections
Rates of caesarean delivery have increased over time in A number of countries have taken different measures to
nearly all OECD countries, although in a few countries this reduce unnecessary caesarean sections. Public reporting,
trend has reversed, at least slightly, in the past few years. provider feedback, the development of clearer clinical
Reasons for the increase include the rise in first births among guidelines, and adjustments to financial incentives have
older women and in multiple births resulting from assisted been used to try to reduce the inappropriate use of
reproduction, malpractice liability concerns, scheduling caesareans. In Australia, where caesarean section rates are
convenience for both physicians and patients, and the high relative to most OECD countries, a number of States
increasing preference of some women to have a caesarean have developed clinical guidelines and required reporting
delivery. Nonetheless, caesarean delivery continues to of hospital caesarean section rates, including investigation
result in increased maternal mortality, maternal and infant of performance against the guidelines. These measures
morbidity, and increased complications for subsequent have discouraged variations in practice and contributed
deliveries, raising questions about the appropriateness of to slowing down the rise in caesarean sections. Other
caesarean deliveries that may not be medically required. countries have reduced the gap in hospital payment rates
In 2015, much as in previous years, caesarean section rates between a caesarean section and a normal delivery, with
were lowest in Nordic countries (Iceland, Finland, Sweden the aim to discourage the inappropriate use of caesareans
and Norway), Israel and the Netherlands, with rates ranging (OECD, 2014).
from 15% to 17% of all live births (Figure 9.20). They were
highest in Turkey, Mexico and Chile, with around one out
of two live births delivered by caesarean section.
Caesarean rates have increased since 2000 in most OECD Definition and comparability
countries, with the average rising from 20% in 2000 to 28%
The caesarean section rate is the number of total
in 2015, although the rate of growth seems to have slowed
caesarean deliveries performed per 100 live births.
over the past 5 years (Figure 9.21). Growth rates have been
particularly rapid in Poland, the Slovak Republic and the In Ireland, Mexico, New Zealand and the United
Czech Republic which have historically had relatively low Kingdom, the data only include activities in publicly-
rates, as well as some of the countries with the highest funded hospitals (though for Ireland all of maternity
rates today (Turkey, Korea). In other countries, the growth units are located in publicly-funded hospitals). This
rate has shown a notable slowing since the mid-2000s, such may lead to an underestimate of caesarean section
as in Israel, Finland and Sweden. In Italy, caesarean rates rates in these countries, since there is some evidence
have come down significantly in recent years, although that private hospitals tend to perform more caesarean
they remain among the highest in Europe. sections than public hospitals.
There can be substantial variations in caesarean rates across
regions and hospitals within the same country. In Italy, there
continue to be huge variations in caesarean rates, mainly
driven by the southern regions of the country. Spain shows References
similar large variations across its regions (OECD, 2014).
FHF – Fédération hospitalière de France (2008), Étude sur les
In several countries, there is evidence that private hospitals
césariennes [Study on caesareans], Paris.
tend to perform more caesarean sections than public
hospitals. In France, private for-profit hospitals authorised OECD (2014), Geographic Variations in Health Care: What Do
to provide maternity care for pregnancies without We Know and What Can Be Done to Improve Health System
complications have caesarean rates as high as public Performance?, OECD Publishing, Paris, http://dx.doi.org/
hospitals which have to deal with more complicated cases 10.1787/9789264216594-en.
(FHF, 2008). In Switzerland, caesarean sections have been OFSP – Office fédéral de la santé publique (2013),
found to be substantially higher in private clinics (41%) Accouchements par césarienne en Suisse [Births by caesareans
than in public hospitals (30.5%) (OFSP, 2013). in Switzerland], Bern.
180 Health at a Glance 2017 © OECD 2017
0
10
20
30
40
50
60
0
10
20
30
40
Tu
20 r ke
00 M y 53.1
ex
20 ic
01 o 46.8
20 Ch
il e
Per 100 live births
02
Ko 46.0
Per 100 live births
20
Hu e a r
03
ng 38.0
20 ar
04 Po y 37.2
20 la
05 nd
36.2
20 I
Korea
06 Au t al y
35.3
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
OECD31
20 S w s tr a
Germany
Health at a Glance 2017 © OECD 2017
07 i t z li a
er 34.0
20 la
08 P o nd
Un r t 32.9
20 ite ug
09 d a
St l 32.3
20 at
10 Sl Ge e s
20
ov 32.2
ak rma
11 Re ny
20 pu 30.2
bl
Italy
12
30.2
Poland
20 Ir e i c
la
13 n
20 Au d 30.1
st
14
O r ia
20 Lu E C D 28.7
15 xe
m 3 3
N e bo 27.9
Un w Z u r g
ite ea 27.8
d l
K i and
ng 26.3
do
0
10
20
30
40
Cz m
20 ec Can 26.2
00 h ad
Re a
20 pu 26.0
01 bl
ic
20 Sp 25.4
02 a
Per 100 live births
De in
20 nm 24.5
03 ar
20 k 21.1
04 La
tv
ia
20 Fr 21.0
05 an
France
Estonia
Be ce 20.8
OECD31
20
lg
06 i
9.20. Caesarean section rates, 2015 (or nearest year)
S l um
20
07 ov 20.8
en
20 Es ia 20.8
08 to
20 n
09
Sw ia 18.7
ed
20 en
10 Is 17.3
ra
20
11
No el 16.2
rw
Finland
20 ay
12 I 16.1
9.21. Caesarian section trends in selected OECD countries, 2000 to 2015 (or nearest year)
Ne c ela
20 th nd
13 er
la 16.0
United Kingdom
20 nd
14 Fi s
nl 15.9
an
20
15 d 15.5
181
12 http://dx.doi.org/10.1787/888933605331
12 http://dx.doi.org/10.1787/888933605312
Caesarean sections
9. HEALTH CARE ACTIVITIES
9. HEALTH CARE ACTIVITIES
Ambulatory surgery
In the past few decades, the number of surgical procedures cataract, with a 34% OECD average and a maximum of 86%
carried out on a same-day basis has markedly increased in Finland. Many countries still lag behind, but show signs
in OECD countries. Advances in medical technologies of catching up. These large differences in the share of same-
– in particular the diffusion of less invasive surgical day surgery may reflect variations in the perceived risks of
interventions – and better anaesthetics have made this postoperative complications, or simply clinical traditions
development possible. These innovations have improved of keeping children for at least one night in hospital after
patient safety and health outcomes, and have also, in the operation.
many cases, reduced the unit cost per intervention by Financial incentives can affect the extent to which minor
shortening the length of stay in hospitals. However, the surgeries are conducted on a same-day basis. In Hungary,
impact of the rise in same-day surgeries on overall health budget caps for same-day surgery financially discouraged
spending may not be straightforward since the reduction in the practice. A recent policy change to abolish this budget
unit cost (compared to inpatient surgery), may be offset by cap is expected to increase the rates of same-day surgeries
the overall growth in the volume of procedures performed. for cataracts and other minor surgeries. In Denmark and
There is also a need to take into account any additional cost France, diagnostic-related group (DRG) systems have been
related to post-acute care and community health services adjusted to incentivise same-day surgery. In the United
following the interventions. Kingdom, a financial incentive of approximately GBP 300
Cataract surgery and tonsillectomy provide good examples per case was awarded for selected surgical procedures
of high-volume surgeries which are now carried out mainly if the patient was managed on a day-case basis (OECD,
on a same-day basis in many OECD countries. 2017).
Day surgery now accounts for 90% or more of all cataract
surgeries in a majority of OECD countries (Figure 9.22).
In several countries, nearly all cataract surgeries are
performed as day cases. However, the use of day surgery Definition and comparability
is still relatively low in Poland, Turkey, Hungary, the Slovak
Cataract surgery consists of removing the lens of the
Republic and Mexico, where they still account for less than
eye because of the presence of cataracts which are
two thirds of all cataract surgeries. While this may be partly
partially or completely clouding the lens, and replacing
explained by limitations in the data coverage of outpatient
it with an artificial lens. It is mainly performed on
activities in hospital or outside hospital, this may also
elderly people. Tonsillectomy consists of removing the
reflect more advantageous reimbursement for inpatient
tonsils, glands at the back of the throat. It is mainly
stays or constraints on the development of day surgery.
performed on children.
The number of cataract surgeries performed on a same-day
The data for several countries do not include outpatient
basis has grown very rapidly since 2000 in many countries,
cases in hospital or outside hospital (i.e., patients who
such as Portugal and Austria (Figure 9.22). Whereas fewer
are not formally admitted and discharged), leading
than 10% of cataract surgeries in Portugal were performed
to some under-estimation. In Ireland, Mexico, New
on a same-day basis in 2000, this proportion has increased
Zealand and the United Kingdom, the data only include
to 97% by 2015. In Austria, the share of cataract surgeries
cataract surgeries carried out in public or publicly-
performed as day cases increased from 1% only in 2000 to
funded hospitals, excluding any procedures performed
75% in 2015. The number of cataract surgeries carried out
in private hospitals (in Ireland, it is estimated that
as day cases has also risen rapidly in many other countries,
approximately 15% of all hospital activity is undertaken
with many of them carrying out 90% or more cases as
in private hospitals). Data for Portugal relate only to
ambulatory in 2015.
public hospitals on the mainland. Data for Spain only
Tonsillectomy is one of the most frequent surgical partially include activities in private hospitals.
procedures on children, usually performed on children
suffering from repeated or chronic infections of the tonsils
or suffering from breathing problems or obstructive sleep
apnea due to large tonsils. Although the operation is
References
performed under general anaesthesia, it is now carried
out mainly as a same-day surgery in several countries, OECD (2017), Tackling Wasteful Spending on Health,
with children returning home the same day (Figure 9.23). OECD Publishing, Paris, http://dx.doi.org/10.1787/
However, the percentage of cases is not yet as high as for 9789264266414-en.
182 Health at a Glance 2017 © OECD 2017
9. HEALTH CARE ACTIVITIES
Ambulatory surgery
9.22. Share of cataract surgeries carried out as ambulatory cases, 2000 and 2015 (or nearest year)
% 2000 2015
99.8 99.1 98.7 98.7 98.6 98.6
98.2 98.2 97.8 96.7 96.7
100 96.6 96.6 96.2 95.6 95.4
94.2 93.1 92.7
90.4
86.8
81.8
79.2
80 74.9
60 63.8 57.2
54.3 53.3
40 34.6
20
Re c o
ia
Hu lic
No ic
da
De ni a
Fi k
Un t h e n d
Ki ds
Sw in
Sl en
Po l y
us al
Re a
y
ey
Ze y
Be d
OE e
nd
m
nd
el
xe a ny
Au g
Ge 28
r ia
a
h r a li
c
ar
re
ar
an
ur
en
bl
g
ra
It a
a
do
iu
rk
rw
an
b
ed
na
xi
n
la
Ne nla
la
CD
st
Sp
to
r tu
nm
Ko
ng
bo
Lu m
pu
pu
al
ite rla
lg
Is
ov
ov M e
Tu
Po
Ir e
t
ng
Fr
Ca
Es
r
m
A
ak
d
Ne
ec
Cz
Sl
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605350
9.23. Share of tonsillectomy carried out as ambulatory cases, 2000 and 2015 (or nearest year)
% 2000 2015
100
86.3
80
73.6 73.5
70.8
67.9
63.3
60 55.6
53.0
47.8
42.8
49.8
37.9
40 34.1
30.6 30.4
24.2
18.3
20
21.0 10.7
8.8
6.6 5.9
3.7 3.7
0.1 0.0 0.0
0
o
ia
ce
r ia
y
d
a
26
ly
el
Au y
li a
m a
g
Ge d
y
nd
da
en
s
ay
l
m
ga
an
ni
e
ar
nd
ar
ic
ai
n
an
an
ur
en
ra
It a
iu
do
rk
rw
an
L u Kor
ra
ed
na
la
la
st
CD
Sp
ex
to
r tu
nm
ng
bo
rm
nl
al
la
lg
Is
ov
Tu
Po
Ir e
st
ng
Au
Fr
Sw
Ca
No
Es
M
Ze
er
Fi
OE
Be
Hu
Po
De
Sl
Ki
th
xe
w
Ne
Ne
ite
Un
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605369
Health at a Glance 2017 © OECD 2017 183
10. PHARMACEUTICAL SECTOR
Pharmaceutical expenditure
Pharmacists and pharmacies
Pharmaceutical consumption
Share of generic market
Research and development in the pharmaceutical sector
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.
Health at a Glance 2017 © OECD 2017 185
10. PHARMACEUTICAL SECTOR
Pharmaceutical expenditure
Pharmaceuticals play a vital role in the health system. some “blockbuster” pharmaceuticals – while it increased
Policymakers need to balance access for new medicines by 2.3% each year in the 2003-09 period. The reduction
while providing the right incentives to industry and was particularly steep in European countries that were
acknowledging that health care budgets are limited. After affected by the economic and financial crisis, such as
inpatient and outpatient care, pharmaceuticals represent Greece (-6.5%), Portugal (-5.9%) and Ireland (-4.4%). As a
the third largest expenditure item of health care spending; response to mounting pressures on public budgets, many
accounting for more than a sixth (16%) of health expenditure governments made reducing pharmaceutical expenditure
on average across OECD countries in 2015 (not taking into a priority to rein in public spending. The policy measures
account spending on pharmaceuticals in hospitals). included the de-listing of products (i.e. excluding them
Similar to other health care functions, the cost of from reimbursement) and the introduction or increase of
pharmaceuticals is predominantly covered by government user charges for retail prescription drugs (Belloni et al.,
financing or compulsory insurance schemes (Figure 10.1). 2016).
Across OECD countries, these schemes cover on average In more recent years a number of countries, including
around 57% of all retail pharmaceutical spending, with Germany, Switzerland, Belgium and the United States have
out-of-pocket payments (39%) and voluntary private seen the return of higher pharmaceutical spending growth
insurance (4%) financing the remaining part. Coverage again, partly due to steep increases in spending for certain
is most generous in Germany and Luxembourg where high cost drugs such as Hepatitis C drugs or oncology drugs.
government and compulsory insurance schemes pay for
80% or more of all pharmaceutical costs. In eight OECD
countries, public or mandatory schemes cover less than Definition and comparability
half the amount spent on medicines. This is the case in
Poland (34%), Latvia (35%), Canada and the United States Pharmaceutical expenditure covers spending on
(both 36%). In these countries, voluntary private insurance prescription medicines and self-medication, often
or out-of-pocket payments play a much bigger role in referred to as over-the-counter products. In some
financing pharmaceuticals. countries, other medical non-durable goods are also
The total retail pharmaceutical bill across OECD countries included. It also includes pharmacists’ remuneration
was more than USD 800 billion in 2015. However, there are when the latter is separate from the price of medicines.
wide variations in pharmaceutical spending per capita Final expenditure on pharmaceuticals includes
across countries, reflecting differences in volume, patterns wholesale and retail margins and value-added
of consumption and pharmaceutical prices, as well as in tax. Total pharmaceutical spending refers in most
the use of generics (Figure 10.2). The United States spent countries to “net” spending, i.e. adjusted for possible
far more on pharmaceuticals than any other OECD country rebates payable by manufacturers, wholesalers or
on a per capita basis (USD 1 162), and more than double the pharmacies.
OECD average. Switzerland (USD 982) and Japan (USD 798) Pharmaceuticals consumed in hospitals and other
also spent significantly more on medicines per capita health care settings as part of an inpatient or day
than other OECD countries. At the other end of the scale, case treatment are excluded (data available suggests
Denmark (USD 282), Israel (USD 313) and Estonia (USD 326) that their inclusion would add another 10-20% to
had relatively low spending levels. pharmaceutical spending). Comparability issues exist
Around 80% of total retail pharmaceutical spending with regards to the administration and dispensing of
is for prescribed medicines, with the rest spent on pharmaceuticals for outpatients in hospitals. In some
over-the-counter medicines (OTC). OTC medicines are countries the costs are included under curative care
pharmaceuticals that can generally be bought without whereas in others under pharmaceuticals.
prescription and their costs are in most cases fully borne Pharmaceutical expenditure per capita is adjusted to
by patients. The share of OTC medicines is particularly high take account of differences in purchasing power.
in Poland, accounting for half of pharmaceutical spending,
but also in Spain (34%) and Australia (31%).
Average annual pharmaceutical spending growth in the
References
2009-15 period has been much lower compared with pre-
crisis years (Figure 10.3). Between 2009 and 2015, expenditure Belloni, A., D. Morgan and V. Paris (2016), “Pharmaceutical
on pharmaceuticals dropped by 0.5% per year on average Expenditure and Policies: Past Trends And Future
across the OECD – mainly driven by cuts in spending by Challenges”, OECD Health Working Papers, No. 87, OECD
government or compulsory schemes and patent expiry of Publishing, Paris, http://dx.doi.org/10.1787/5jm0q1f4cdq7-en.
186 Health at a Glance 2017 © OECD 2017
10. PHARMACEUTICAL SECTOR
Pharmaceutical expenditure
10.1. Expenditure on retail pharmaceuticals¹ by type of financing, 2015 (or nearest year)
Government and compulsory schemes Voluntary health insurance
% Out-of-pocket Other
100
16 13 17
25 27 29 24 29
7 31 31 33 34 33
80 38 41 41 42 39 40
12 44 44 48 47 48 49 45
1 51 51
1 58 65 66
1 26
60 4 5 1 4 34
1 31
6
40 84 80 75 72 71 71 69 68 67 65 62 59 59 58 57 55 55 55 52 52 51 51 51 50 48
20 44 38 36 36 35 34
da
Au nia
Re c e
B e lic
ic
y
g
nd
ov Fr n
n ia
th om
s
Re al y
No n
OE ay
F 0
it z nd
Po nd
Gr a l
ce
Es a
Sw ia
Hu en
Sl ar y
De a li a
Un Ic e k
d
es
Po ia
nd
an
re
nd
ar
a
ai
ur
i te lan
3
bl
d str
tv
iu
ak a n
rw
b
ee
at
ed
p
na
la
la
S w inla
la
CD
t
e
Sp
r tu
to
Ko
nm
Ne gd
ng
bo
r
rm
pu
pu
I
Ja
la
La
lg
ov
St
Ir e
er
st
ite Au
Ca
er
m
Ge
d
Ki
xe
h
Lu
ec
Cz
Un
Sl
Note: “Other” includes financing from non-profit-schemes, enterprises and the rest of the world.
1. Includes medical non-durables.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605388
10.2. Expenditure on retail pharmaceuticals per capita, 2015 (or nearest year)
Prescribed medicines Over-the-counter medicines Total (no breakdown)
USD PPP
1 400
1 162
1 200
982
1 000
798
766
756
800
684
663
637
621
617
601
572
553
553
550
535
525
600
509
497
484
480
479
417
404
413
401
387
369
352
326
400
313
282
200
0
Ir e d a
ce
O E lic
Ne we a
it z tes
nd
Ge an
C a ny
B e d¹
m
A u t r ia
li a
ov Gr y¹
Re ce¹
31
H ain
m y
g
ng a
er n
No lic
Po ia
Es d
Is a
D e a e l¹
k
Fi ¹
ov d
Cz Po ds¹
R e g a l¹
Ic ay
m
S i
e
ni
ar
xe ar
th de
n
ur
an
an
en
tv
l
iu
n
an
rw
b
d Ko r
b
ra
a
p
na
It a
la
la
CD
do
ak e e
Sp
to
Sw Sta
nm
Lu ung
la
bo
rm
pu
pu
r
nl
el
Ja
La
lg
ec tu
er
st
la
Au
Fr
r
Sl
d
Ki
i te
h
Un
ite
Sl
Un
1. Includes medical non-durables (resulting in an overestimation of around 5-10%).
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605407
10.3. Average annual growth in retail pharmaceutical expenditure¹ per capita, in real terms,
2003-09 and 2009-15 (or nearest period)
% 2003-09 2009-15
12
10
11.0
8
7.0
6.4
6.3
6
4.2
3.9
3.8
3.8
4
3.1
3.1
2.7
2.5
2.5
2.5
2.5
2.3
2.2
2.1
2.1
2.0
2.0
2.0
1.9
1.8
1.5
2
1.0
1.1
0.9
0.8
0.7
0.7
0.7
0.7
0.7
0.0
0.6
0.7
0.5
0.4
0.4
0.0
0.1
0
-0.1
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-0.6
-2
-0.7
-0.8
-0.9
-1.0
-1.5
-1.3
-1.7
-2.2
-2.2
-4
-2.8
-5.9
-6.5
-3.9
-4.4
-6
-8
Re c o
da
ia
y
Po e c e
Hu blic
Sl ce
B e lic
Ir e a l
nd
th and
ec Me k
Fr y
Sl xem an d
Re r g
OE m
Ca 1
Po n
Sw nd
en
Au r ael
G e a li a
Un A u l y
St a
E s
i t z nia
N o nd
ay
a
n
ia
De nds
an
i
e
re
ar
ar
ai
pa
3
en
ite str
g
It a
tv
iu
ak bou
an
rw
b
at
ed
xi
na
la
la
la
CD
Sp
r tu
S w sto
nm
Ko
ng
e
r
rm
pu
pu
Ne c el
Lu F inl
Ja
la
La
lg
Is
ov
er
st
Gr
er
I
d
h
ov
Cz
1. Includes medical non-durables.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605426
Health at a Glance 2017 © OECD 2017 187
10. PHARMACEUTICAL SECTOR
Pharmacists and pharmacies
Pharmacists are educated and trained health care (Vogler et al., 2012). Denmark has fewer community
professionals who manage the distribution of medicines pharmacies, but these are often large, including branch
to consumers/patients and help ensure their safe and pharmacies and supplementary pharmacy units attached
efficacious use. The role of the pharmacist has changed to the main pharmacy (Vogler et al., 2012).
over recent years. Although their main role is to The range of products and services provided by the
dispense medications in retail pharmacies, pharmacists pharmacies varies across countries. In most European
are increasingly providing direct care to patients (e.g. countries, for example, pharmacies can also sell cosmetics,
flu vaccinations in Ireland and New Zealand, medicine food supplements, medical devices and homeopathic
adherence support in Australia, Japan, England and New products. In a few countries pharmacies can also sell
Zealand), both in community pharmacies and as part of reading glasses and didactic toys (Martins et al., 2015).
integrated health care provider teams.
Between 2000 and 2015, the number of pharmacists has
increased by 30% in OECD countries. Japan has by far the Definition and comparability
highest density of pharmacists, at twice the OECD average,
while the density of pharmacists is low in Turkey, Chile and Practising pharmacists are defined as the number of
the Netherlands (Figure 10.4). Between 2000 and 2015, the pharmacists who are licensed to practice and provide
number of pharmacists per capita has increased in nearly direct services to clients/patients. They can be either
all OECD countries, with the exception of Switzerland. It salaried or self-employed, and work in community
increased most rapidly in Portugal, Spain, Slovenia and pharmacies, hospitals and other settings. Assistant
the Slovak Republic. pharmacists and the other employees of pharmacies
In Japan, the increase in the number of pharmacists are normally excluded.
can be largely attributed to the government’s efforts to In Ireland, the figures include all pharmacists
separate more clearly drug prescribing by doctors from registered with the Pharmaceutical Society of Ireland,
drug dispensing by pharmacists (the Bungyo system). possibly including some pharmacists who are not in
Traditionally, the vast majority of prescription drugs in activity. Assistant pharmacists are included in Iceland.
Japan were dispensed directly by doctors. However, in recent
Community pharmacies are premises which in
decades, the Japanese government has taken a number of
accordance to the local legal provisions and definitions
steps to encourage the separation of drug prescribing from
may operate as a facility in the provision of pharmacy
dispensing.
services in the community settings. The number of
Most pharmacists work in community retail pharmacies, community pharmacies reported are the number of
but some also work in hospital, industry, research and premises where dispensing of medicines happened
academia (FIP, 2015). For instance, in Canada more than under the supervision of a pharmacist.
three-quarters of practising pharmacists worked in
a community pharmacy, while about 20% worked in
hospitals and other health care facilities in 2012 (CIHI,
2015). In Japan, around 55% of pharmacists worked in References
community pharmacies in 2014, while around 20% worked
CIHI – Canadian Institute for Health Information (2015),
in hospitals or clinics and the other 25% worked in other
“Pharmacist Workforce, 2012 – Provincial/Territorial
settings (Survey of Physicians, Dentists and Pharmacists
Highlights”, Ottawa, Canada.
2014).
FIP – International Pharmaceutical Fededation (2015),
Variation in the number of community pharmacies across
“Global Trends Shaping Pharmacy – Regulatory
OECD countries (Figure 10.5) can be explained by the
Frameworks, Distribution of Medicines and Professional
different dispensing channels for medicines. In addition
Services. 2013-2015”.
to community pharmacies, medicines can be dispensed
through hospital pharmacies (both for inpatient and Martins, S.F. et al. (2015), “The Organizational Framework of
outpatient use) or can be provided directly by doctors in Community Pharmacies in Europe”, International Journal
some countries. For example, the relatively low number of of Clinical Pharmacy, May 28.
community pharmacies in the Netherlands may be partly Vogler, S. et al. (2012), “Impact of Pharmacy Deregulation
explained by the fact that patients can also purchase and Regulation in European Countries”, Gesundheit
their prescription drugs directly from some doctors Österreich GmbH, Vienna.
188 Health at a Glance 2017 © OECD 2017
10. PHARMACEUTICAL SECTOR
Pharmacists and pharmacies
10.4. Practising pharmacists, 2000 and 2015 (or nearest year)
2000 2015
Per 100 000 population
180
170
160
140
121
119
116
120
112
112
111
110
104
104
100
92
85
84
83
82
80
76
76
80
74
74
74
72
72
71
71
70
66
65
64
63
60
51
44
54
40
35
21
20
K c
i t z nia
Fi d²
Be pan
m
n
Fr ly¹
Ic ce¹
Ir e n d ¹
Ca d
Un G da¹
St ¹
ec b ¹
Re r g
rm a
S w love y
st ¹
i t e or i a
S w n d¹
ng al
Ne L 3 4
Ze via
en
No el
Po ay
Es nd
Hu ni a
ov A ar y
L u e p u ia
D e l a nd
Ch k
Ne Tur ²
er ¹
s
O E om
Au tes
C z em blic
d ce
th ey
il e
i
S an
G e or e
ar
nd
ai
an
bl
R tr
Un P r a l
ra
d tug
iu
h ou
rw
ed
It a
la
CD
an
k
w at
ite ee
Sp
to
na
nm
d
ng
a
a
la
ak u s
pu
nl
Ja
la
lg
Is
er
el
al
r
Ki
x
Sl
1. Data include not only pharmacists providing direct services to patients, but also those working in the health sector as researchers, for pharmaceutical
companies, etc.
2. Data refer to all pharmacists licensed to practice.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605445
10.5. Community pharmacies, 2015 (or nearest year)
Per 100 000 population
50
47.2
45.0
43.9
45
41.8
40
37.5
35.7
34.0
35
31.5
29.9
30
28.0
26.7
24.7
24.8
23.6
25
23.1
22.1
21.3
19.2
20
16.9
15.4
15.0
14.9
15
13.3
11.7
10.1
10
6.0
5
3.9
0
ce
n
a
nd
nd
g
r ia
y
ly
l
da
y
26
y
li a
nd
ay
s¹
el
k
ga
es
en
an
ke
re
ar
ar
ai
pa
an
ur
an
ra
It a
nd
iu
do
rw
an
ra
na
la
la
la
st
CD
at
Sp
r tu
ed
nm
Ko
ng
bo
r
rm
el
nl
Ja
lg
Is
Tu
Po
er
Ir e
st
la
ng
Au
Fr
Ca
St
No
Ic
Sw
Fi
OE
Be
Hu
m
Po
er
Au
it z
Ge
De
Ki
d
xe
th
Sw
i te
d
Ne
Lu
ite
Un
Un
1. Estimates.
Source: FIP (2015), “Global Trends Shaping Pharmacy – Regulatory Frameworks, Distribution of Medicines and Professional Services. 2013-2015”.
12 http://dx.doi.org/10.1787/888933605464
Health at a Glance 2017 © OECD 2017 189
10. PHARMACEUTICAL SECTOR
Pharmaceutical consumption
In general, pharmaceutical consumption continues to
increase, partly driven by a growing need for drugs to active ingredient(s) in a given therapeutic class by
treat ageing-related and chronic diseases, and by changes international expert consensus. For instance, the
in clinical practice. This section examines consumption DDD for oral aspirin equals 3 grams, which is the
of four categories of pharmaceuticals: antihypertensive, assumed maintenance daily dose to treat pain in
cholesterol-lowering, antidiabetic and antidepressant adults. DDDs do not necessarily reflect the average
drugs. daily dose actually used in a given country. DDDs can
Consumption of antihypertensive drugs has nearly doubled be aggregated within and across therapeutic classes
in OECD countries between 2000 and 2015. It has nearly of the Anatomic-Therapeutic Classification (ATC). For
quadrupled in Luxembourg and Estonia (Figure 10.6). It is more detail, see www.whocc.no/atcddd.
highest in Germany and Hungary, which report almost five The volume of antihypertensive drugs consumption
times the consumption levels in Korea and Turkey. These presented in Figure 10.6 refers to the sum of five ATC2
variations reflect both differences in the prevalence of high categories, which can all be prescribed for hypertension
blood pressure and in clinical practice. (Antihypertensives, Diuretics, Beta-blocking agents,
The use of cholesterol-lowering drugs has nearly quadrupled Calcium channel blockers and Agents acting on the
in OECD countries between 2000 and 2015 (Figure 10.7). The Renin-Angiotensin system).
Slovak Republic, Denmark and the United Kingdom report Data generally refer to outpatient consumption
the highest consumption per capita in 2015. Across OECD only, except for Chile, the Czech Republic, Estonia,
countries, there is an eight fold variation in consumption Finland, France, Iceland, Italy, Korea, Norway, the
levels of cholesterol-lowering drugs. Slovak Republic and Sweden, where data also include
The use of antidiabetic drugs has almost doubled in OECD hospital consumption. The data for Canada relate to
countries between 2000 and 2015 (Figure 10.8). This growth three provinces only (British Columbia, Manitoba
can be explained by the rising prevalence of diabetes, and Saskatchewan). The data for Luxembourg and
largely linked to increases in the prevalence of obesity Spain refer to outpatient consumption for prescribed
(see indicators on overweight and obesity in Chapter 4), a drugs covered by the National Health System (public
major risk factor for the development of type 2 diabetes. In insurance). Data for Luxembourg are underestimated
2015, the consumption of antidiabetic drugs was highest in due to incomplete consideration of products with
Finland, the Czech Republic and Greece. multiple active ingredients.
Consumption of antidepressant drugs has doubled in OECD
countries between 2000 and 2015 (Figure 10.9). This might
reflect improved recognition of depression, availability of References
therapies, guidelines and changes in patient and provider
Belloni, A., D. Morgan and V. Paris (2016), “Pharmaceutical
attitudes (Mars et al., 2017). However, there is significant
Expenditure and Policies: Past Trends and Future
variation in consumption of antidepressants between
Challenges”, OECD Health Working Papers, No. 87, OECD
countries. Iceland reports the highest level of consumption
Publishing, Paris, http://dx.doi.org/ 10.1787/5jm0q1f4cdq7-en.
of antidepressants in 2015, twice the OECD average,
followed by Australia, Portugal and the United Kingdom. Grandfils, N. and C. Sermet (2009), “Evolution 1998-2002 of
Latvia, Korea and Estonia report the lowest consumption the Antidepressant Consumption in France, Germany
levels of antidepressants. and the United Kingdom”, Document de travail IRDES,
No. 21, Paris.
Mars, B. et al. (2017), “Influences on Antidepressant
Definition and comparability Prescribing Trends in the UK: 1995–2011”, Social Psychiatry
Defined daily dose (DDD) is the assumed average and Psychiatric Epidemiology, Vol. 52, No. 2, pp. 193-200
maintenance dose per day for a drug used for its OECD (2014), Making Mental Health Count: The Social and Economic
main indication in adults. DDDs are assigned to each Costs of Neglecting Mental Health Care, OECD Publishing,
Paris, http://dx.doi.org/10.1787/9789264208445-en.
190 Health at a Glance 2017 © OECD 2017
10. PHARMACEUTICAL SECTOR
Pharmaceutical consumption
10.6. Antihypertensive drugs consumption, 2000 and 10.7. Cholesterol-lowering drugs consumption, 2000
2015 (or nearest year) and 2015 (or nearest year)
2000 2015 2000 2015
Turkey 126.6 Turkey 17.8
Korea 127.6 Chile 48.8
Austria 170.0 Estonia 51.6
Latvia 189.7 Korea 58.8
Israel 210.2 Latvia 62.5
Luxembourg 214.8 Austria 75.9
Australia 229.2 Germany 79.3
Portugal 247.9 Iceland 89.7
Chile 260.4 France 91.7
France 265.7 Italy 92.0
Iceland 271.6 Sweden 99.3
Norway 273.1 OECD28 100.7
Spain 289.2 Canada 103.4
Greece 290.5 Greece 104.6
Netherlands 299.3 Finland 105.3
Canada 304.6 Portugal 107.0
OECD28 317.0 Hungary 107.6
Belgium 325.5 Spain 110.2
Estonia 331.1 Israel 112.2
Sweden 374.3 Luxembourg 117.3
United Kingdom 380.4 Czech Republic 117.7
Denmark 396.1 Slovenia 118.3
Slovenia 411.5 Netherlands 120.4
Slovak Republic 420.9 Norway 127.9
Italy 432.7 Australia 134.1
Finland 433.7 Belgium 134.3
Czech Republic 451.1 United Kingdom 139.0
Hungary 557.2 Denmark 141.3
Germany 591.3 Slovak Republic 152.0
0 100 200 300 400 500 600 0 30 60 90 120 150 180
Defined daily dose, per 1 000 people per day Defined daily dose, per 1 000 people per day
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605483 12 http://dx.doi.org/10.1787/888933605502
10.8. Antidiabetic drugs consumption, 2000 and 2015 10.9. Antidepressant drugs consumption, 2000 and
(or nearest year) 2015 (or nearest year)
2000 2015 2000 2015
Austria 38.0 Latvia 12.3
Latvia 43.8 Korea 20.3
Iceland 46.1 Estonia 24.8
Norway 51.4 Hungary 28.3
Denmark 53.8 Chile 37.3
Australia 57.5 Turkey 38.0
Canada 58.3
Slovak Republic 40.0
Sweden 58.7
Israel 44.4
Estonia 59.6
Netherlands 45.1
Israel Italy 46.5
59.6
Turkey Greece 48.1
60.7
Korea France 49.8
61.7
Luxembourg Luxembourg 53.7
65.2
Czech Republic 55.0
France 65.7
Germany 56.4
OECD28 66.5 Norway 56.5
Portugal 67.3 Slovenia 56.8
Belgium 68.1 OECD29 60.3
Hungary 69.9 Austria 60.3
Italy 70.9 Finland 68.2
Spain 72.9 New Zealand 72.8
Chile 74.0 Spain 73.1
Slovak Republic 75.2 Denmark 77.0
Netherlands 75.6 Belgium 78.3
Slovenia 76.1 Canada 90.1
Germany 82.6 Sweden 92.5
United Kingdom 84.7 United Kingdom 94.2
Greece 86.2 Portugal 95.1
Czech Republic 88.1 Australia 104.2
Finland 90.1 Iceland 129.6
0 20 40 60 80 100 0 20 40 60 80 100
Defined daily dose, per 1 000 people per day Defined daily dose, per 1 000 people per day
Source: OECD Health Statistics 2017. Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605521 12 http://dx.doi.org/10.1787/888933605540
Health at a Glance 2017 © OECD 2017 191
10. PHARMACEUTICAL SECTOR
Generics and biosimilars
All OECD countries view the development of generic
markets as a good opportunity to increase efficiency in Definition and comparability
pharmaceutical spending, but many do not fully exploit
A generic is defined as a pharmaceutical product
the potential of generics (Figure 10.10). In 2015, generics
which has the same qualitative and quantitative
accounted for more than three-quarters of the volume
composition in active substances and the same
of pharmaceuticals sold in the United States, Chile,
pharmaceutical form as the reference product, and
Germany, New Zealand and the United Kingdom, while
whose bioequivalence with the reference product
they represented less than one-quarter of the market in
has been demonstrated. Generics can be classified in
Luxembourg, Italy, Switzerland and Greece.
branded generics (generics with a specific trade name)
Some of the differences in generic uptake can be explained and unbranded generics (which use the international
by market structures, notably the number of off-patent non-proprietary name and the name of the company).
medicines, and by prescribing practices, but generic uptake
Countries were requested to provide data for the
also depends on policies implemented by countries (EGA,
whole market. However many countries provided data
2011; Vogler, 2012). Several countries have expanded their
covering only the community pharmaceutical market
efforts to encourage generic uptake since the onset of the
or the reimbursed pharmaceutical market (see Figure
economic crisis in 2008.
notes). The share of generic market expressed in value
Financial incentives for physicians, pharmacists and can be the turnover of pharmaceutical companies,
patients have been implemented to boost the development the amount paid for pharmaceuticals by third-party
of generic markets. For instance, France (in 2009 and payers, or the amount paid by all payers (third-party
2012) introduced incentives for GPs to prescribe generics and consumers). The share of generic market in
through a pay-for-performance scheme while in Japan (in volume can be expressed in DDDs or as a number of
2012) payment bonuses also contributed to an increased packages/boxes or standard units.
share of generics in total prescribing. Pharmacies are often
A Biosimilar Medicinal Product is the product granted
paid through mark-ups based on the price of medicines.
regulatory approval, demonstrating similarity to the
This disincentive to substitute a generic for a more
Reference Medicinal Product in terms of quality
expensive drug has been addressed in some countries.
characteristics, biological activity, safety and efficacy.
France guarantees pharmacists an equivalent mark-up,
Referenced Medicinal Product is the original product,
while in Switzerland pharmacists receive a fee for generic
which was granted market exclusivity at the start of
substitution. Patients have a financial interest to choose
its life, but once exclusivity has expired the product
cheaper drugs when their co-payment is lower for generic
has been categorised as referenced. The biosimilar
drugs than its equivalent. This is generally the case in all
market share is the number of biosimilar treatment
systems using reference prices (or fixed reimbursement
days as a share of biosimilar and referenced product(s)
amount) for clusters of products. In Greece, patients
volume. Volume is measured in Defined Daily Dose
choosing originator over generic drugs are now required
which is a measure of the average dose prescribed as
to pay for the difference.
defined by the WHO.
A biosimilar is a biological medicine highly similar
to another already approved biological medicine (the
“reference medicine”). Biological medicines contain
active substances from a biological source, such as living References
cells or organisms. The rationale behind the introduction
EGA – European generic medicines (2011), “Market Review –
of biosimilars is to increase price competition, thereby
The European Generic Medicines Markets”, European
reducing prices. There is large variation in the uptake
generic medicines.
for two biosimilars – Epoetin and Anti-Tumour Necrosis
Factor (Anti-TNF) – across OECD countries (Figure 10.11). OECD (2017). Tackling Wasteful Spending on Health, OECD
Biosimilars have 100% of the Epoetin market share in Publishing, Paris, http://dx.doi.org/10.1787/9789264266414-en.
Finland, Hungary, Poland, the Slovak Republic and the Quintiles IMS (2017), “The Impact of Biosimilar Competition
Czech Republic, whereas it is 2% in Belgium and 6% in in Europe”, London.
the United Kingdom. For Anti-TNF, biosimilars have 90% Vogler, S. (2012), “The Impact of Pharmaceutical Pricing
and 82% of the market share in Denmark and Norway and Reimbursement Policies on Generic Uptake:
respectively, while it is 2% in Switzerland and 5% in Implementation of Policy Options on Generics in
Belgium and Ireland. 29 European Countries – An Overview”, Generics and
Biosimilars Initiative Journal, Vol. 1, No. 2, pp. 44-51.
192 Health at a Glance 2017 © OECD 2017
10. PHARMACEUTICAL SECTOR
Generics and biosimilars
10.10. Share of generics in the total pharmaceutical market, 2015 (or nearest year)
% Value Volume
90 86
84
81
79 78
75
73 72
71
70
61
60
55
53 52
50
48 49 48
47
45
42 42
39 40 39
36 36
34
31 32
29 30
30 28 28
25 24
23 23 23 22
20 19
19 18 18
17 16 17 16 17 16
15
12 11
8
6
0
da
nm c
Fi c
²
k¹
ey
OE ia¹
Sl 2 7
No ²
ay
C z Por ¹
Re al
Ir e d
Es ¹
a
Be n
m
Gr ¹
it z e¹
nd
y¹
d¹
¹
ia
s¹
y
g¹
m
ce
nd
ia
es
il e
ni
xe It al
pa
an
bl
bl
g
tv
an
nd
Sw eec
ar
iu
ai
ur
an
rk
rw
r
na
en
la
do
CD
an
tu
to
at
Ch
st
la
pu
pu
Sp
nl
Ja
La
lg
bo
rm
Tu
er
la
al
Ca
ov
St
ng
Au
Fr
Re
er
Ze
m
Ge
De
Ki
d
th
ite
h
ak
Ne
d
ec
Lu
Ne
Un
ite
ov
Sl
Un
1. Reimbursed pharmaceutical market.
2. Community pharmacy market.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605559
10.11. Biosimilar market share (volume) for Epoetin and Anti-Tumour Necrosis Factor (Anti-TNF)
vs reference product , 2015 (or nearest year)
Epoetin Anti-TNF
% 100 100 100 100 99
100
94
91 90
90 87 87
82 81
80 76
70
70 68
65
61 60
60
50 46 45
40
32 33
29 30
30 26 27
24 25
23 22
20 19
20 18 17
14 14
10 6 6
5 5
2 2
0
ic
ic
ia
ce
d
nd
en
nd
ay
r ia
20
ly
m
s
nd
m
ga
an
ar
nd
ar
ai
an
bl
bl
en
It a
do
iu
rw
an
ed
la
la
la
CD
st
Sp
r tu
nm
ng
rm
pu
pu
nl
la
lg
ov
Po
Ir e
er
ng
Au
Fr
Sw
No
er
Fi
OE
Be
Hu
Po
Re
Re
it z
Ge
De
Sl
Ki
th
Sw
h
ak
Ne
d
ec
ite
ov
Cz
Un
Sl
Source: Quintiles IMS (2017), “The Impact of Biosimilar Competition in Europe”, London.
12 http://dx.doi.org/10.1787/888933605578
Health at a Glance 2017 © OECD 2017 193
10. PHARMACEUTICAL SECTOR
Research and development in the pharmaceutical sector
Funding for pharmaceutical research and development complex combination of factors. These include growing
(R&D) is the result of a complex mix of private and public requirements to obtain market approval that have increased
sources. Governments mainly support basic and early- clinical trial costs and an ever-increasing “back catalogue”
stage research. Such funding is made through direct budget of effective drugs that has shifted research efforts to more
allocations, research grants, publicly-owned research complex conditions (Scannell et al., 2012). Rising R&D costs
institutions and funding of higher education institutions. can be both a cause and a result of higher drug prices,
The pharmaceutical industry translates and applies as the acceptance of higher prices by payers can make
knowledge generated by basic research to develop products, increasingly expensive R&D financially viable. Increasing
and invests in large clinical trials required to gain market R&D costs can then in turn drive up prices.
approval. The industry also receives direct R&D subsidies
or tax credits in many countries.
In 2014, governments of OECD countries budgeted about
Definitions and comparability
USD 51 billion on health-related R&D (a broader category
than pharmaceuticals). This figure understates total Business enterprise expenditure on R&D (BERD) covers
government support, since it excludes most tax incentive R&D carried out by corporations, regardless of the origin
schemes or funding for higher education or publicly-owned of funding, which can include government subsidies.
corporations. Meanwhile, the pharmaceutical industry BERD is recorded in the country where the R&D activity
spent approximately USD 100 billion on R&D across OECD took place, not the country providing funding. National
countries. In high-income countries, the business sector statistical agencies collect data primarily through
has been estimated to contribute 60% of all health-related surveys and according to the Frascati Manual (OECD,
research, while 30% comes from governments and 10% from 2015) but there is some variation in national practices.
other sources, including private not-for-profit organisations “Pharmaceutical R&D” refers to BERD by businesses
and universities’ own funds (Røttingen et al., 2013). classified in the pharmaceutical industry.
Most pharmaceutical R&D takes place in OECD countries. Government budgets for R&D (GBARD) capture both
However, the share of non-OECD countries in global R&D performed directly by government and amounts
industry R&D expenditure is increasing (Chakma et al., paid to other institutions for R&D. “Health-related
2014), especially in China, where the industry spent R&D” refers to GBARD aimed at protecting, promoting
approximately USD 11 billion on R&D in 2014 (0.05% of and restoring human health, including all aspects of
GDP). More than half of the spending in OECD countries medical and social care. It does not cover spending by
(Figure 10.12) occurs in the United States, where the public corporations or general university funding that
pharmaceutical industry spent about USD 56 billion is subsequently allocated to health.
(0.3% of GDP), and direct government budgets on health-
The gross value added (GVA) of a sector equals gross
related R&D were USD 33 billion (0.2% of GDP). Industry
output less intermediate consumption. It includes the
spent USD 26 billion (0.1% of GDP) and governments
cost of wages, consumption of fixed capital and taxes on
budgeted USD 11 billion (0.05% of GDP) in Europe; and
production. Because GVA does not include intermediate
USD 15 billion (0.3% of GDP) and USD 1.6 billion (0.03%
consumption, it is less sensitive than gross output to
of GDP) respectively in Japan. As a share of GDP, industry
sector-specific reliance on raw materials. OECD averages
spending is highest in Switzerland (0.6%), Belgium (0.6%)
in Figure 10.13 are based on 15 countries for air and
and Slovenia (0.4%), smaller countries with relatively
spacecraft, and 25-29 countries for all other industries.
large pharmaceutical sectors.
The pharmaceutical industry is highly R&D intensive. On
average across OECD countries, the industry spent some
14% of its gross value added on R&D. This is almost as References
high as in the air and spacecraft (18%) and electronics and
optical products industries (17%), and considerably higher Chakma, J. et al. (2014), “Asia’s Ascent – Global Trends in
than the average across manufacturing as a whole (6%) Biomedical R&D Expenditures”, New England Journal of
(Figure 10.13). Medicine, Vol. 370, No. 1, pp. 3-6.
Expenditure on R&D in the pharmaceutical industry in OECD OECD (2015), Frascati Manual 2015: Guidelines for Collecting
countries grew by more than 50% in real terms between 2004 and Reporting Data on Research and Experimental
and 2014. However, this increase is not associated with higher Development, OECD Publishing, Paris, http://dx.doi.
output in terms of new drug approvals (NDAs). In the United org/10.1787/9789264239012-en.
States, the annual number of NDAs has remained relatively Røttingen, J.A. et al. (2013), “Mapping of Available Health
stable since the 1980s (Figure 10.14) while the number of Research and Development Data: What’s There, What’s
approvals per inflation-adjusted R&D spending has declined Missing, What Role Is There for a Global Observatory?”,
steadily. Exceptions are the late 1990s, when a backlog of The Lancet, Vol. 382, No. 9900, pp. 1286-1307.
pending applications was cleared, and the years since 2010. Scannell, J.W. et al. (2012), “Diagnosing the Decline in
This pattern of constant output at increasing costs despite Pharmaceutigal R&D Efficiency”, Nature Reviews Drug
advances in technology (“Eroom’s Law”) is driven by a Discovery, Vol. 11, No. 3, pp. 191-200.
194 Health at a Glance 2017 © OECD 2017
10. PHARMACEUTICAL SECTOR
Research and development in the pharmaceutical sector
10.12. Business enterprise expenditure for pharmaceutical R&D (BERD) and government budgets
for health-related R&D (GBARD), 2014 or nearest year
Business R&D expenditure, pharma Direct government R&D budgets, health
Billion USD % of GDP
60 56.6 0.6
50 0.5
40 0.4
33.5 0.33
0.29
30 26.4 0.3
0.19
20 14.6 0.2
0.14
11.4
10 0.1 0.06
1.6 2.8 4.2 0.03 0.03 0.05
0 0
pe
pe
D
es
es
pa
pa
EC
EC
at
at
ro
ro
Ja
Ja
St
rO
St
rO
Eu
Eu
d
d
he
he
ite
ite
Ot
Ot
Un
Un
Note: 2012 BERD data for Switzerland and 2011 GBARD data for Mexico; all other countries 2014 or 2013. Europe includes 21 EU member countries that
are also members of the OECD, Iceland, Norway and Switzerland; no BERD data available for Luxembourg and no GBARD data for Latvia.
Source: OECD Main Science and Technology Indicators and Research and Development Statistics Databases.
12 http://dx.doi.org/10.1787/888933605597
10.13. R&D intensity by industry: business enterprise R&D expenditure (BERD) as a proportion
of gross value added (GVA), 2014 or nearest year
BERD/GVA, percentage
United States, 43.8
40 Japan, 39.0
Belgium, 32.1
30
20
18.2 17.2
OECD average, 14.2
10
5.6
0.7 0.7 0.4 0.4 0.2
0
Air and Electronic and Pharmaceuticals Total Mining and Total services Utilities Agriculture, Construction
spacecraft optical products manufacturing quarrying forestry and fishing
Note: The air & spacecraft, electronic & optical products and pharmaceutical industries are sub-categories of total manufacturing. All other industries
are totals at the same level as total manufacturing.
Source: OECD Analytical Business Enterprise R&D (ANBERD), Structural Analysis (STAN) and System of National Accounts (SNA) Databases. National
statistics offices for GVA in the pharmaceutical industry in Australia and the air & spacecraft industry in Canada.
12 http://dx.doi.org/10.1787/888933605616
10.14. Annual new drug approvals (NDAs) per billion USD pharmaceutical business expenditure
on R&D in the United States, inflation-adjusted
Number of NDAs NDAs per USD bn pharma BERD
Number of NDAs, 3yr average Number of NDAs per bn USD BERD
140 21
120 18
100 15
80 12
60 9
40 6
20 3
0 0
0
1
2
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
8
8
8
20
20
20
20
20
19
19
19
19
19
20
20
19
19
19
19
19
19
19
19
19
20
20
19
19
20
19
19
19
20
20
20
20
19
20
Source: United States Food and Drug Administration (FDA); Pharmaceutical Research and Manufacturers of America (PhRMA).
12 http://dx.doi.org/10.1787/888933605635
Health at a Glance 2017 © OECD 2017 195
11. AGEING AND LONG-TERM CARE
Demographic trends
Life expectancy and healthy life expectancy at age 65
Self-reported health and disability at age 65
Dementia prevalence
Recipients of long-term care
Informal carers
Long-term care workers
Long-term care beds in institutions and hospitals
Long-term care expenditure
The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.
Health at a Glance 2017 © OECD 2017 197
11. AGEING AND LONG-TERM CARE
Demographic trends
Longer life expectancies (see indicators on life expectancy Although the pressure that this growing proportion of
in Chapter 3) and declining fertility rates mean that older people aged 65 and 80 over will put on long-term care
people make up an ever-increasing proportion of the systems will depend on the health status of people as
populations of OECD countries. they reach these ages, population ageing will likely lead to
On average across OECD countries, the share of the greater demand for elderly care and contribute to increases
population aged over 65 years increased from less than in health spending. Nevertheless, most studies have found
9% in 1960 to 17% in 2015, and is expected to continue new technologies and rising incomes to be more significant
to increase, reaching 28% in 2050 (Figure 11.1, left panel). drivers of health spending growth than population ageing
In more than two-thirds of OECD countries, at least one- (OECD, 2015).
quarter of the population will be over 65 years of age by As populations age, the potential supply of labour in the
2050. This proportion is expected to be especially large in economy is expected to decline. On average across OECD
Japan, Spain, Portugal, Greece and Korea, where nearly countries, there were slightly more than four people of
40% of the population will be aged over 65 years by 2050. working age (15-64 years) for every person 65 years and
Population ageing will also occur rapidly in China, where older in 2012. This rate is projected to halve from 4.2 in 2012
the share of the population over 65 is expected to nearly to 2.1 on average across OECD countries over the next 40
triple between 2015 and 2050, to reach a level just below years (OECD, 2013). Moreover, ageing may lead to shortfalls
the OECD average. Conversely, Israel, the United States and in certain revenue-raising mechanisms, particularly payroll
Mexico will see a more gradual increase in the share of the taxes, making it more difficult for countries to maintain or
elderly population due to significant inflows of migrants or increase government spending on health.
higher fertility rates.
The growth in the share of the population aged 80 years
and over will be even more dramatic (Figure 11.1, right Definition and comparability
panel). On average across OECD countries, nearly 5% of the
population was 80 years old and over in 2015. By 2050, the Data on the population structure have been extracted
percentage will increase to more than 10%. In Italy, Spain, from the OECD historical population data and
Portugal, and Germany, the proportion of the population projections (1950-2050). The projections are based
aged over 80 is expected to more than double between 2015 on the most recent “medium-variant” population
and 2050. The rise will be even faster in Korea, where the projections from the United Nations, World Population
share of the population aged over 80 years will grow from Prospects – 2017 Revision.
3% to 14% over the next four decades.
Population ageing is a phenomenon affecting most
countries around the world, but the speed of the process References
varies (Figure 11.2). The speed of population ageing has
been particularly fast in Japan, where the share of the Muir, T. (2017), “Measuring Social Protection for Long-
population aged 80 years and over increased from 2% in term Care”, OECD Health Working Papers, No. 93, OECD
1990 to nearly 8% in 2015, and is expected to rise to 15% by Publishing, Paris, http://dx.doi.org/10.1787/a411500a-en.
2050. The population in Korea remains relatively young, but
OECD (2015), Fiscal Sustainability of Health Systems: Bridging
is expected to age rapidly in the coming decades, so that by
Health and Finance Perspectives, OECD Publishing, Paris,
2050 the share of the population over 80 will be nearly the
http://dx.doi.org/10.1787/9789264233386-en.
same as in Japan. The pace of population ageing has been
slower in non-OECD countries, although it is expected to OECD (2013), Pensions at a Glance 2013: OECD and G20 Indicators,
accelerate. In large partner countries including Brazil and OECD Publishing, Paris, http://dx.doi.org/10.1787/pension_
China, less than 2% of the population was 80 years and over glance-2013-en.
in 2015, though this share is expected to reach close to 7% United Nations (2017), 2017 Revision of World Population
in Brazil and more than 8% in China by 2050. Prospects, United Nations, https://esa.un.org/unpd/wpp/.
198 Health at a Glance 2017 © OECD 2017
11. AGEING AND LONG-TERM CARE
Demographic trends
11.1. Share of the population aged over 65 and 80 years, 2015 and 2050
2015 2050
Population aged 65 years and over Population aged 80 years and over
Japan 27 Japan 8
Spain 19 Italy 7
Portugal 20 Korea 3
Greece 21 Spain 6
Korea 13 Portugal 6
Italy 22 Germany 6
Slovenia 18 Greece 6
Poland 15 Austria 5
Germany 21 Slovenia 5
Austria 18 Switzerland 5
Czech Republic 18 Netherlands 4
Estonia 19 France 6
Switzerland 18 Canada 4
Slovak Republic 14 Belgium 5
Hungary 18 Finland 5
Netherlands 18 OECD35 4
Latvia 20 United Kingdom 5
OECD35 17 Estonia 5
France 18 Poland 4
Belgium 18 Denmark 4
China 10 Sweden 5
Finland 20 Lithuania 5
Canada 16 New Zealand 4
Lithuania 19 Iceland 4
Ireland 13 Latvia 5
United Kingdom 18 Czech Republic 4
Iceland 14 Norway 4
Sweden 20 United States 4
Denmark 19 Luxembourg 4
Chile 10 Chile 2
Norway 16 Australia 4
New Zealand 15 Ireland 3
Costa Rica 7 Costa Rica 2
Luxembourg 14 China 2
Brazil 8 Slovak Republic 3
Australia 15 Hungary 4
United States 15 Brazil 2
Russian Federation 14 Russian Federation 3
Colombia 8 Israel 3
Turkey 8 Turkey 2
Mexico 7 Mexico 2
Israel 11 Colombia 1
Indonesia 5 India 1
India 6 Indonesia 1
South Africa 5 South Africa 1
0 10 20 30 40 0 5 10 15 20
% %
Source: OECD Health Statistics 2017, OECD Historical Population Data and Projections Database, 2017.
12 http://dx.doi.org/10.1787/888933605654
11.2. Trends in the share of the population aged over 80 years, 1990-2050
Korea Japan Partner countries¹ World OECD
%
16
14
12
10
0
1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
1. Partner countries include Brazil, China, Colombia, Costa Rica, India, Indonesia, Lithuania, the Russian Federation and South Africa.
Source: OECD Historical Population Data and Projections Database, 2017.
12 http://dx.doi.org/10.1787/888933605673
Health at a Glance 2017 © OECD 2017 199
11. AGEING AND LONG-TERM CARE
Life expectancy and healthy life expectancy at age 65
Life expectancy at age 65 has increased significantly 65-year-old men with a high level of education could expect
for both men and women over the past few decades in to live five years longer than those with a low education
OECD countries, rising by 5.4 years on average since 1970 level in 2015. By contrast, differences in life expectancy by
(Figure 11.3). Some of the factors explaining these gains education level are much smaller (less than two years) in
in life expectancy at age 65 include advances in medical Nordic countries (Denmark, Finland, Norway and Sweden)
care combined with greater access to health care, healthier and Portugal (see Eurostat Database 2017).
lifestyles and improved living conditions before and after
people reach age 65.
Japan and Korea have achieved the highest gains in life Definition and comparability
expectancy at age 65 since 1970, with an increase of about
eight years. The gains have been much more modest in Life expectancy measures how long on average a person
Hungary, the Slovak Republic and Mexico, with an increase of a given age can expect to live, if current death rates
of only about three years. do not change. However, the actual age-specific death
In 2015, people at age 65 in OECD countries could expect to rate of any particular birth cohort cannot be known
live another 19.5 years: 21 years for women and 18 years in advance. If rates are falling, as has been the case
for men (Figure 11.4). This gender gap of three years on over the past decades in OECD countries, actual life
average across OECD countries has been fairly stable over spans will be higher than life expectancy calculated
time. In 2015, life expectancy at age 65 was highest in Japan using current death rates. The methodology used to
for women (24 years) and in Japan, Australia, and Iceland calculate life expectancy can vary slightly between
for men (nearly 20 years). Among OECD countries, it was countries. This can change a country’s estimates by
lowest in Hungary for women (18 years) and in Latvia for a fraction of a year. Life expectancy at age 65 is the
men (14 years). unweighted average of the life expectancy at age 65
of women and men.
Countries’ relative positions with respect to life expectancy
at age 65 mirror closely their relative positions with regard Disability-free life expectancy (or “healthy life
to life expectancy at age 80. Life expectancy at age 80 in years”) is defined as the number of years spent free
2015 was highest in Japan for women (who can expect to of activity limitation. In Europe, this indicator is
live an additional 12 years) and highest in France for men calculated annually by Eurostat for EU countries and
(who can expect to live an additional 9 years). some EFTA countries. The disability measure is based
on the Global Activity Limitation Indicator (GALI)
Increased life expectancy at age 65 does not necessarily
question, which comes from the European Union
mean that the extra years lived are in good health. In
Statistics on Income and Living Conditions (EU-SILC)
Europe, an indicator of disability-free life expectancy
survey. The question asks: “For at least the past six
known as “healthy life years” is calculated regularly, based
months, have you been hampered because of a health
on a general question about disability in the European
problem in activities people usually do? Yes, strongly
Union Statistics on Income and Living Conditions (EU-SILC)
limited / Yes, limited / No, not limited”. While healthy
survey. Among European countries participating in the
life years is the most comparable indicator to date,
survey, the average number of healthy life years at age 65
there are still problems with translation of the GALI
was almost the same for women and men, at 9.3 years for
question, although it does appear to satisfactorily
women and 9.4 years for men in 2015 (Figure 11.5). The
reflect other health and disability measures (Jagger
absence of any significant gender gap in healthy life years
et al., 2010).
means that many of the additional years of life that women
experience relative to men are lived with some type of
activity limitation. Nordic countries (with the exception of
Finland) had the highest number of healthy life years at References
age 65 in 2015. In Sweden, women could expect to live an
average of an additional 17 years, and men 16 years, free Jagger, C. et al. (2010), “The Global Activity Limitation
of disability. Indicator (GALI) Measured Function and Disability
Similarly across European Countries”, Journal of Clinical
Life expectancy and healthy life expectancy vary by
Epidemiology, Vol. 63, pp. 892-899.
educational status. For both men and women, highly
educated people are likely to live longer and in better Mäki, N. et al. (2013), “Educational Differences in
health. Differences in life expectancy by education level Disability-free Life Expectancy: A Comparative study
are particularly large in Central and Eastern European of Long-standing Activity Limitation in Eight European
countries, especially for men. In the Slovak Republic, Countries”, Social Science & Medicine, Vol. 94, pp. 1-8.
200 Health at a Glance 2017 © OECD 2017
0
2
4
6
8
10
12
14
16
18
20
0
5
10
15
20
25
30
0
5
10
15
20
25
30
Sw Ja Ja
p p
Years
Years
ed
en
16.3
Fr an
an 21.9 Fr an
an 21.9
Ic c c
el
an S w Sp e 21.5 S w Sp e 21.5
Healthy years
d 15.3
i t z ain i t z ain
No er 21.0 er 21.0
rw Au lan Au lan
ay st d 20.9 st d 20.9
Ge 15.3 ra
li a
ra
li a
rm Ne 20.9 Ne 20.9
an w It a w It a
y 11.9 Ze l y Ze l y
al 20.6 al 20.6
Ir e a a
la
nd Lu Ic e n d 20.4 Lu Ic e n d 20.4
Source: Eurostat Database 2017.
De 11.7 xe lan
m d
xe lan
m d
nm 20.4 20.4
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
bo bo
Health at a Glance 2017 © OECD 2017
ar ur ur
k 11.5 Ko g 20.4 Ko g 20.4
Un Be
lg No r e a
ite iu No r e a
20.3 20.3
d m rw rw
Ki 11.1 C a ay C a ay
ng
do na 20.3 na 20.3
m d d
10.3 Is a 20.2 Is a 20.2
Total
r r
Total
Fr S w ael S w ael
an 20.2 20.2
Sw ce ed ed
it z 10.3 F i en F i en
er nl 20.2 nl 20.2
Ne la a a
nd
th 10.1
Gr n d
ee 20.1 Gr n d
ee 20.1
er Po c e Po c e
la
Lu nd r tu 19.9 r tu 19.9
xe s B g B g
1970
m 10.0
bo Ne elg al 19.9 Ne elg al 19.9
ur t h ium t h ium
g er er
9.7 la 19.9 la 19.9
OE Un n Un n
CD ite Au ds ite Au ds
25 d st 19.8 d st 19.8
Ki ri Ki ri
9.4 ng a 19.7 ng a 19.7
Sp do do
ai 19.7 19.7
n Ir e m Ir e m
Men
9.2 l l
Male
Cz Fi OE and OE and
ec
nl
an CD 19.7 CD 19.7
h d Sl 3 5 Sl 3 5
Re 9.2 ov 19.5 ov 19.5
pu Ge eni Ge eni
bl
ic rm a 19.5 rm a 19.5
8.3 a a
Po Un D e n ny Un D e n ny
la i t e ma 19.5 i t e ma 19.5
2015
nd d rk d rk
Sl 8.0 St 19.4 St 19.4
ov at at
Note: Countries are ranked in descending order of healthy life expectancy for the whole population.
Note: Countries are ranked in descending order of healthy life expectancy for the whole population.
en es es
ia Ch 19.3 Ch 19.3
Au 7.9 Es le i Es le i
st to 18.5 to 18.5
r ia ni ni
7.8 Br a 18.1 Br a 18.1
Gr az az
ee
ce Po il
la 18.0 Po il
la 18.0
7.7 n n
Women
Female
Tu d 17.9 Tu d 17.9
11.4. Life expectancy at age 65 by sex, 2015 (or nearest year)
It a Cz r ke Cz r ke
11.3. Life expectancy at age 65, 1970 and 2015 (or nearest year)
Po
ly
7.7 e M y 17.8 e M y 17.8
S l ch R e x i S l ch R e x i
r tu ov e c o ov e c o
ga ak pu 17.7 ak pu 17.7
Hu l 6.2 Re bli Re blic
17.7 17.7
11.5. Healthy life years at age 65, European countries, 2015 (or nearest year)
ng pu c pu
ar bl bl
y Ru L ic Ru L ic
5.9 ss a 16.9 ss a 16.9
Es i a Hu t v ia i a Hu t v i a
to
ni n 16.6 n 16.6
a F e nga F e nga
5.3 d d
Sl La S o er a r y 16.4 S o er a r y 16.4
ov ut tio ut tio
ak tv
ia h n h n
15.5 15.5
Re 4.1 Co Afr Co Afr
pu st ica st ica
bl aR 13.9 aR 13.9
ic ic ic
4.0 a 7.3 a 7.3
201
12 http://dx.doi.org/10.1787/888933605730
12 http://dx.doi.org/10.1787/888933605711
12 http://dx.doi.org/10.1787/888933605692
Life expectancy and healthy life expectancy at age 65
11. AGEING AND LONG-TERM CARE
11. AGEING AND LONG-TERM CARE
Self-reported health and disability at age 65
Most OECD countries conduct regular health surveys
which allow respondents to report on different aspects Definition and comparability
of their health. These surveys often include a question
on self-perceived health status, along the lines of: “How Self-reported health reflects people’s overall
is your health in general?”. Although these questions are perception of their own health, including both
subjective, indicators of perceived general health have physical and psychological dimensions. Typically,
been found to be a good predictor of future health care use survey respondents are asked a question such as:
and mortality (Hirosaki et al., 2017; Schnittker and Bacak, “How is your health in general? Very good, good, fair,
2014). However, cross-country differences may be difficult poor, very poor”. OECD Health Statistics provides figures
to interpret, as survey questions may differ slightly and related to the proportion of people rating their health
cultural factors can affect responses. to be “good/very good” combined.
More than half of the population aged 65 years and over Caution is required in making cross-country comparisons
report being in good health in 14 of 34 OECD countries of perceived health status, for at least two reasons. First,
(Figure 11.6). In New Zealand, Canada and the United people’s assessment of their health is subjective and
States, more than three-quarters of older people report can be affected by cultural factors. Second, there are
good health, though the response categories offered to variations in the question and answer categories used to
survey respondents in these three countries are different measure perceived health across surveys/countries. In
from those used in most other OECD countries (see particular, the response scale used in Australia, Canada,
“Definition and comparability” box). Among European New Zealand and the United States is asymmetric
countries, older people in Norway, Sweden, Ireland, (skewed on the positive side), including the following
Switzerland, and the Netherlands report the best health response categories: “excellent, very good, good, fair,
status, with more than 60% assessing their health to be poor”. The data reported in OECD Health Statistics refer
good. At the other end of the scale, less than 15% of the to respondents answering one of the three positive
population aged 65 years and over in Latvia and Portugal responses (“excellent, very good or good”). By contrast,
report being in good health. In nearly all countries, men in most other OECD countries, the response scale is
over 65 were more likely than women to rate their health symmetric, with response categories being: “very good,
to be good. On average across OECD countries, 47% of men good, fair, poor, very poor”. The data reported from these
aged over 65 rated their health to be good or better, while countries refer only to the first two categories (“very
41% of women did so. good, good”). Such differences in response categories
may introduce an upward bias in the results from those
Self-reported health status varies substantially by income.
countries that are using an asymmetric scale.
Across OECD countries on average, less than one-third
of people aged 65 years and older in the lowest income Perceived general disability is measured by the
quintile considered their health to be good, compared Global Activity Limitation Indicator (GALI) question,
with close to 60% of those in the highest income quintile which comes from the European Union Statistics on
(Figure 11.7). In Sweden, more than four-fifths of people Income and Living Conditions (EU-SILC) survey. The
aged 65 and older in the highest income quintile consider question is: “For at least the past six months, have
their health to be good, while fewer than half of people in you been hampered because of a health problem in
the lowest income quintile say the same. activities people usually do? Yes, strongly limited /
Yes, limited / No, not limited”. Persons in institutions
Although measures of disability are not fully standardised
are not surveyed, resulting in an under-estimation of
across countries, the European Union Statistics on Income
disability prevalence. Again, the measure is subjective,
and Living Conditions (EU-SILC) survey collects data on
and cultural factors may affect survey responses.
the limitations that people face in their daily activities
(Figure 11.8). Such limitations often correspond to a
need for long-term care. On average across 26 European
countries, 51% of all over-65s reported that they were References
limited either to some extent or severely in their usual
Hirosaki, M. et al. (2017), “Self-rated Health Is Associated
daily activities because of a health problem in 2015
with Subsequent Functional Decline Among Older
(Figure 11.8). The lowest rates of disability were reported
Adults in Japan”, International Psychogeriatrics,
in Nordic countries, with around one in five over-65s
Vol. 29, No. 9, pp. 1475-1483, http://dx.doi.org/10.1017/
reporting at least some limitation in daily activities in
S1041610217000692.
Sweden and Norway. The highest rates were found in
Eastern European countries such as Latvia and the Slovak Schnittker, J. and V. Bacak (2014), “The Increasing Predictive
Republic, where three-quarters of over-65s reported at Validity of Self-Rated Health, PLoS ONE, Vol. 9, No. 1,
least some limitations. e84933, https://doi.org/10.1371/journal.pone.0084933.
202 Health at a Glance 2017 © OECD 2017
11. AGEING AND LONG-TERM CARE
Self-reported health and disability at age 65
11.6. Perceived health status in adults aged 65 years and over, 2015 (or nearest year)
Total Male Female
% reporting to be in good or very good health
100
80
60
85.5
40
78.6
78.1
72.7
19.8
66.7
65.4
65.3
64.3
60.3
59.7
57.2
56.3
16.6
16.0
53.3
52.9
15.7
12.7
48.3
47.8
44.0
43.6
42.8
41.4
40.3
20
38.6
9.6
32.5
31.9
29.3
25.4
24.0
20.7
21.9
0
U n C a d¹
xe om
Fi rg
OE nd
F r ia
Ge nce
ia
Tu ic
ic
St a¹
st ¹
No ia¹
Gr n
S w ay
I n
it z nd
t h a nd
De nd s
Sl ec e
il e
ec Ja y
Re p an
Po y
Es r y
Po ni a
La l
ia
Ic r k
d
ite el l
m
Re a
Hu d
Au 4
e
ga
Au tes
an
e
l
e
e
ai
n
an
bl
en
bl
r
ra
tv
It a
i t e n ad
d gi u
an
a
u
rk
rw
Ch
r
l
ed
la
S w r ela
st
CD
ra
Sp
r tu
to
nm
a k Ko
Lu ingd
ng
bo
e
rm
pu
a
pu
Ne erl
nl
el
a
la
Is
ov
al
er
m
Ze
B
d
K
w
h
Ne
ov
Cz
Un
Sl
Note: Numbers are close together for males and females for New Zealand, the United States, Canada, the United Kingdom and Denmark.
1. Data for New Zealand, Canada, the United States and Australia are biased upwards relative to other countries and so are not directly comparable.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605749
11.7. Perceived health status in adults aged 65 years and over by income quintile, European countries, 2015
(or nearest year)
Lowest quintile Highest quintile
% reporting to be in good or very good health
100
80
60
40
20
0
ia
ic
ic
ce
nd
en
ay
nd
r ia
ce
ly
nd
ia
l
25
ga
an
ni
ar
nd
ar
an
ai
an
ur
en
bl
bl
It a
tv
iu
do
rw
an
ee
ed
la
la
la
CD
st
Sp
to
r tu
nm
ng
bo
rm
pu
pu
nl
el
la
La
lg
ov
er
Po
Ir e
ng
Au
Gr
Fr
Sw
No
Es
Ic
er
Fi
OE
Be
Hu
m
Po
Re
Re
it z
De
Ge
Sl
Ki
th
xe
Sw
ak
Ne
Lu
ec
ite
ov
Cz
Un
Sl
Source: Eurostat Database.
12 http://dx.doi.org/10.1787/888933605768
11.8. Limitations in daily activities in adults aged 65 years and over, European countries, 2015 (or nearest year)
% Some limitations Severe limitations
100
49.5
45.4
80
44.8
45.5
42.4
38.1
30.3
35.4
36.9
36.4
34.9
40.1
60
36.0
32.7
36.1
40.6
29.2
29.5
34.9
23.6
25.2
32.2
25.3
23.5
40
8.2
14.9
11.9
20
26.5
22.5
22.5
23.8
20.8
28.9
28.9
17.8
14.2
15.2
19.3
21.8
18.0
13.6
19.8
19.4
26.1
12.7
11.9
12.1
21.1
17.2
17.0
8.5
5.8
7.0
7.7
0
ic
en
ay
nd
ia
ia
m
ce
ic
g
nd
nd
r ia
ce
ly
a
26
ga
an
ke
ni
ar
nd
ar
ai
an
ur
an
bl
bl
en
tv
It a
iu
do
rw
an
ee
ed
la
la
la
st
CD
Sp
r tu
to
nm
ng
bo
r
rm
pu
pu
nl
el
La
la
lg
ov
er
Po
Tu
Ir e
ng
Au
Gr
Fr
Sw
No
Es
Ic
er
Fi
OE
Be
Hu
m
Po
Re
Re
it z
Ge
De
Sl
Ki
th
xe
Sw
ak
Ne
d
Lu
ec
ite
ov
Cz
Un
Sl
Source: Eurostat Database.
12 http://dx.doi.org/10.1787/888933605787
Health at a Glance 2017 © OECD 2017 203
11. AGEING AND LONG-TERM CARE
Dementia prevalence
Dementia describes a variety of brain disorders which to be driven in part by higher rates of dementia at older
progressively lead to brain damage and cause a gradual ages. However, the use of antipsychotics rises less steeply
deterioration of the individual’s functional capacity and than dementia prevalence (Figure 11.10).
social relations. Alzheimer’s disease is the most common
form of dementia, representing about 60% to 80% of cases.
There is currently no cure or disease modifying treatment, Definition and comparability
but better policies can improve the lives of people with
dementia by helping them and their families adjust to The prevalence estimates in Figure 11.9 are taken
living with the condition and ensuring that they have from the World Alzheimer Report 2015, which includes
access to high quality health and social care. a systematic review of studies of dementia prevalence
In 2017, there were an estimated 18.7 million people living around the world. Prevalence by country has been
with dementia in OECD countries. This is equivalent estimated by applying these age-specific prevalence
to around one in every 69 people in the population as a rates for the relevant region of the world to country-
whole, but dementia prevalence increases rapidly with specific population estimates from the United Nations
age. Across all OECD countries, around 2% of people aged (World Population Prospects: The 2017 Revision).
65-69 have dementia, compared with more than 40% of Differences between countries are therefore driven by
those aged over 90 (Figure 11.10). As a result, countries the age structure of populations – i.e. older countries
with older populations have more people with dementia: have more people with dementia. The World Alzheimer
Japan, Italy, and Germany are estimated to have more than Report 2015 analysis includes studies carried out since
20 people with dementia per 1 000 population, while the 1980, with the assumption that age-specific prevalence
Slovak Republic, Turkey and Mexico have fewer than nine is constant over time. This assumption is retained in
(Figure 11.9). the construction of this indicator, so that fixed age-
specific prevalence rates are applied for both 2017 and
Ageing populations mean that dementia will become
2037. Although gender-specific prevalence rates were
more common in the future, and the most rapidly ageing
available for some regions, overall rates were used in
countries will see prevalence more than double in the next
this analysis.
20 years. This includes fast-ageing OECD countries (Korea
and Chile) and partner countries such as Brazil, China, Antipsychotics are defined consistently across
Colombia and Costa Rica. However, there is some evidence countries using Anatomical Therapeutic Classification
that the age-specific prevalence of dementia may be falling (ATC) codes. The numerator includes all patients on
in some countries (Matthews et al., 2013) and it may be the medications register with a prescription for a drug
possible to reduce the risk of dementia through healthier within the ATC subgroup N05A. The denominator
lifestyles and preventive interventions. If such efforts are is the total number of people on the register. Most
successful, the rise in prevalence may be less dramatic than countries are unable to identify which prescriptions
these numbers suggest. relate to people with dementia, so the antipsychotics
indicator covers all people aged over 65. Some caution
Behavioural and psychological symptoms affect many
is needed when making inferences about the dementia
people with dementia and can make caring for them
population, since it is not certain that a higher rate of
difficult. Antipsychotic drugs can reduce these symptoms,
prescribing among all over-65s translates into more
but the associated risks and ethical issues – and the
prescriptions for people with dementia. Nonetheless,
availability of a range of effective non-pharmacological
measuring this indicator, exploring the reasons for
interventions – mean they are only recommended as a
variation and reducing inappropriate use can help to
last resort. However, the inappropriate use of these drugs
improve the quality of dementia care.
remains widespread and reducing their overuse is a policy
priority for many OECD countries.
New data collected by the OECD show that rates of
prescribing of antipsychotics to older people vary by more References
than a factor of two across OECD countries (Figure 11.12).
In 2015, Sweden, Norway the Netherlands, France, Australia Matthews, F.E. et al. (2013), “A Two-decade Comparison of
and Denmark prescribed antipsychotics to fewer than 35 in Prevalence of Dementia in Individuals Aged 65 Years and
every thousand people aged over 65, with rates either falling Older from Three Geographical Areas of England: Results
or constant. At the other extreme, more than 70 in every of the Cognitive Function and Ageing Study I and II”, The
thousand people aged over 65 in Slovenia had a prescription Lancet, Vol. 382, No. 9902.
of antipsychotics, an increase of 14% since 2011. OECD (2015), Addressing Dementia: the OECD Response, OECD
Rates of antipsychotic prescribing rise with age Health Policy Studies, OECD Publishing, Paris, http://
(Figure 11.11). On average across 13 OECD countries, 3% of dx.doi.org/10.1787/9789264231726-en.
people aged 65-69 had a prescription for antipsychotics in Prince, M. et al. (2015), World Alzheimer Report 2015: The Global
2015, compared to 12% of people aged over 90. This is likely Impact of Dementia, Alzheimer’s Disease International.
204 Health at a Glance 2017 © OECD 2017
11. AGEING AND LONG-TERM CARE
Dementia prevalence
11.9 Dementia prevalence
2017 2037
People with dementia per 1 000 population
45
40
35
30
25
20
15
10
5
0
Ch o
do i c a
lo ia
M bia
ur a
ak B y
C o epu il
n l
ov e
Ic n i a
th ia
Ja l y
a
aR c
n P rea
I ic
ite rel y
St d
Lu C a and
N e e m b ad a
CD a
Gr a i n
r e
n
Ko c a
e c de nd
pu n
L li a
Ne No 3 5
er ay
Es nia
S w ng r k
A u um
S w s t r ia
F i de n
Sp d
Fr c e
rm al
y
Ze rg
st d
d nm s
es
i t z dom
lg d
Hu sr a e
s t bli
an
in
ke
Sl Chil
Po a n c
si
OE oni
R raz
i t e De nd
Un I ga r
ic
pa
d an
Re io
B e la n
an
Au alan
C o In d
bl
Ge tug
Li at v
It a
Ki a
w ou
th r w
ee
i
at
ra
C z F e ola
ne
In A f r
e
m
ex
ua
h r at
i
x n
e
el
nl
la
t
er
T
h
ut
So
ov
ia
Un
Sl
ss
Ru
Source: OECD analysis of data from the World Alzheimer Report 2015 and the United Nations.
12 http://dx.doi.org/10.1787/888933605806
11.10 Prevalence of dementia across all OECD 11.11 Proportion of population with a prescription of
countries by age group, 2017 antipsychotics, by age group, 2015 or nearest year
% Highest rate in any country
45
41 Average of 13 OECD countries
40
% Lowest rate in any country
35
45
30
40
25
20 35
20
30
15 12
25
10 7
4 20
5 2
15
0
10
9
+
-7
-6
-7
-8
-8
90
5
70
75
85
65
80
0
Source: OECD analysis of data from the World Alzheimer Report 2015 and
9
+
-7
-6
-7
-8
-8
90
the United Nations.
70
75
85
65
80
12 http://dx.doi.org/10.1787/888933605825
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605844
11.12 People with a prescription of antipsychotics, 2015 or nearest year
2015 2011
Per 1 000 people aged 65 and over
80 75.0
70 65.0
56.5 70.4
60 55.7 58.2
51.7
50 45.8 47.6 47.8
40
30 33.7 34.6
29.7 30.9 31.3
30.5
20
10
0
ia
ce
en
li a
ly
15
da
d
m
n²
m
¹
ni
nd
ar
an
an
ay
en
It a
iu
do
ai
an
ra
ed
na
CD
to
nm
rw
nl
al
la
lg
Sp
ov
st
ng
Fr
Sw
Ca
Es
er
Ze
OE
Fi
Be
No
Au
De
Sl
Ki
th
w
Ne
d
Ne
ite
Un
1. Data for Norway do not include people in institutional care, so underestimate the use of antispychotics.
2. Data for Spain refer to 2014.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605863
Health at a Glance 2017 © OECD 2017 205
11. AGEING AND LONG-TERM CARE
Recipients of long-term care
As people age, they are more likely to develop disabilities in the use of institutional care, but there has been an even
and need support from family, friends and long-term care larger decrease in the number of “curators” appointed by
(LTC) services. As a result, while LTC services are delivered local government to care for people at home.
to younger disabled groups, the majority of LTC recipients
are older people. On average across the OECD, 13% of
people over 65 receive long-term care (Figure 11.13). The Definition and comparability
proportion of over-65s receiving long-term care varies from
2% in Portugal and 6% in Estonia to more than 20% in Israel LTC recipients are defined as persons receiving long-
and Switzerland. term care by paid providers, including nonprofessionals
People aged over 80 make up on average more than half of receiving cash payments under a social programme.
all LTC recipients, and almost two-thirds of LTC recipients in They also include recipients of cash benefits such
Japan, Denmark and Australia (Figure 11.14). As populations as consumer-choice programmes, care allowances
age, demand for LTC services is likely to increase – although or other social benefits which are granted with the
this effect may be partially offset by improving health in primary goal of supporting people with long-term care
old age. Nonetheless, a significant number of younger needs. LTC institutions refer to nursing and residential
disabled people require long-term care, making up as many care facilities which provide accommodation and
as a third of all LTC recipients in Norway, Slovenia and the long-term care as a package. LTC at home is defined as
Netherlands. people with functional restrictions who receive most
of their care at home. Home care also applies to the
While population ageing is a significant driver of the
use of institutions on a temporary basis, community
growth in LTC users over time, it explains relatively
care and day-care centres and specially designed
little of the cross-country variation. For example, Israel
living arrangements. Data for Poland, the United
has one of the youngest populations in the OECD but a
States, Ireland, Canada, the Slovak Republic, Iceland
greater than average proportion receiving LTC. A more
and Belgium are only available for people receiving
important driver is the availability of publicly funded LTC
long-term care in institutions, so the total number
services. However, data for people receiving care outside
of recipients will be underestimated. In Estonia, data
of public systems are more difficult to collect and may
on recipients of home care refer only to those who
be underreported, meaning that figures for countries
have a “curator” appointed by local government. Other
that rely more heavily on privately-funded care may be
social services, without a personal care component,
artificially low. Cultural norms around the degree to which
are not included in the data. It is possible that some
families look after older people may also be an important
of the decrease in recipients reflects the replacement
driver of the utilisation of formal services (see indicator
of curators with these other services.
on “Informal carers”).
Data on LTC services is difficult to collect in many
In response to most people’s preference to receive LTC
countries and there are some known limitations of the
services at home, many OECD countries have implemented
figures. Data for some countries refer only to people
programmes and benefits to support home-based care, in
receiving publicly-funded care, while other countries
particular for older people. In most countries for which
include people who are paying for their own care.
trend data are available, the proportion of LTC recipients
Data from France and the Czech Republic refer to
aged 65 and over receiving long-term care at home has
the number of people claiming care benefits, which
increased over the past ten years (Figure 11.15), with
may not correspond directly to the number receiving
particularly large increases in Portugal and Sweden. In
services. Some countries use different age categories:
Portugal this reflects an expansion of home care services
instead of reporting people aged 65 and over, Belgium
from a very low level in 2005. In Sweden it results from a
reports those aged 60 and over and Iceland those aged
deliberate policy to reduce institutional care capacity and
66 and over.
encourage community care (Colombo et al., 2011).
While the proportion of LTC recipients living at home has
increased over the past decade in most OECD countries, it
has declined significantly in Finland and Estonia. In Finland, References
this has not been driven by an increase in traditional
institutional care. Instead, there has been an increase in Colombo, F. et al. (2011), Help Wanted? Providing and Paying
the use of “service housing” – where older people move for Long-Term Care, OECD Publishing, Paris, http://dx.doi.
into specially adapted houses where 24/7 care is available. org/10.1787/9789264097759-en.
Although this is classified as institutional care, it allows Muir, T. (2017), “Measuring Social Protection for Long-
more independence and autonomy than a traditional care term Care”, OECD Health Working Papers, No. 93, OECD
institution. In Estonia, there has been a significant increase Publishing, Paris, http://dx.doi.org/10.1787/a411500a-en.
206 Health at a Glance 2017 © OECD 2017
0
5
10
25
15
20
%
0
20
40
60
80
100
%
0
10
20
30
40
50
60
70
80
90
100
%
Po
Po r tu
r tu Ja ga
ga pa
n Es l 2.1
l 38.1 66 30 3 to
De ni
Es
nm a 5.7
to ar Ko
ni k 64 26 9 re
Lu a 56.1 Au a 7.4
xe st Sp
m ra ai
li a n
bo
ur 62 27 11 Fr 8.5
g 57.0 Fi an
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
nl ce
an 10.3
Health at a Glance 2017 © OECD 2017
Sl
ov d 58 25 18
Fi
nl
en an
ia Ko Sl d 11.4
57.8 re ov
Ne a 57 38 5
en
w ia
Fi
nl Ze Hu 11.6
an al Lu ng
d 58.8
an ar
d xe y
0-64
57 25 19 m 12.8
bo
Fr ur
an
ce
Sp
ai OE g 12.8
59.2 n 56 17 27
Cz
ec CD
h
2005
Ge
2005
18
rm Re 13.0
Ko an pu
re
a
Lu y 54 27 20
bl
ic
64.9 xe
m
Ge 13.2
bo rm
OE ur an
CD g Au y 13.4
16
Sw 53 22 24 st
66.8
it z ra
er li a
la 14.4
1. These values include only recipients of long-term care in institutions.
Ge nd No
rm 53 26 21 Ne rw
an w ay
Ne y 69.4 Is Ze 16.1
th ra al
el an
er 52 29 19 d 16.4
65-79
la Sw Sw
nd ed
s 71.2 ed Ne
(or nearest year)
en th en
23 27 er 17
No 50 la
OE nd
rw CD s
ay 18.4
71.4 19
52
Ne
w 27 21 S w Isr a
Po
Ze r tu i t z el 20.7
al er
an ga la
d 72.0 l 48 37 15 nd
Sw Sl 21.6
2015
2015
it z ov
er en
la ia
nd 47 22 31 Un Po
72.7 ite la
No nd
Sw rw d ¹ 0.9
ay St
45 21 34 at
ed Ne
80+
en th es
73.5 er Ir e ¹ 3.3
Cz la la
nd
Hu ec
h s 45 24 31
nd
¹
ng Re
Sl
ov C a n 3.5
ar
11.14 Share of long-term care recipients, by age, 2015 (or nearest year)
y 76.6
pu ak ad
bl a¹ 3.8
ic Re
44 26 30 pu
Sp Es bl
ic¹
ai
n to Ic 3.9
78.8 ni el
a 40 33 26 an
Is
Hu B e d¹ 6
ra ng lg
ar iu
el 90.8 y m
29 44 28 ¹ 8.8
11.13 Proportion of people aged 65 and over receiving long-term care, 2005 and 2015 (or nearest year)
11.15 Share of long-term care recipients aged 65 years and over receiving care at home, 2005 and 2015
207
12 http://dx.doi.org/10.1787/888933605920
12 http://dx.doi.org/10.1787/888933605901
12 http://dx.doi.org/10.1787/888933605882
Recipients of long-term care
11. AGEING AND LONG-TERM CARE
11. AGEING AND LONG-TERM CARE
Informal carers
Family and friends are the most important source of care for participation rates of women in the labour market mean
people with long-term care (LTC) needs in OECD countries. that there is a risk that fewer people will be willing and able
Because of the informal nature of care that they provide, it to provide informal care in the future. Coupled with the
is not easy to get comparable data on the number of people effects of an ageing population, this could lead to higher
caring for family and friends across countries, nor on the demand for professional LTC services. Public LTC systems
frequency of their caregiving. The data presented in this will need adequate resources to meet increased demand
section come from national or international health surveys while maintaining access and quality.
and refer to people aged 50 years and over who report
providing care and assistance to family members and friends.
On average across OECD countries for which data is Definition and comparability
available, around 13% of people aged 50 and over report
providing informal care at least weekly – but this figure Informal carers are defined as people providing any
is more than 20% in the Czech Republic and Belgium and help to older family members, friends and people in
less than 10% in Poland and Portugal (Figure 11.16). There their social network, living inside or outside of their
is also variation in the intensity of the care provided. The household, who require help with everyday tasks. The
lowest rates of daily care provision are found in Sweden, data relate only to the population aged 50 and over,
Switzerland, Denmark and the Netherlands – countries and are based on national surveys for Australia (Survey
where the formal LTC sector is well-developed and public of Disability, Ageing and Carers, SDAC), the United
coverage is comprehensive. Kingdom (English Longitudinal Study of Ageing, ELSA),
Intensive caregiving is associated with a reduction in the United States (Health and Retirement Survey,
labour force attachment for caregivers of working age, HRS) and an international survey for other European
higher poverty rates, and a higher prevalence of mental countries (Survey of Health, Ageing and Retirement
health problems. Many OECD countries have implemented in Europe, SHARE).
policies to support family carers with a view to mitigating Questions about the intensity of care vary between
these negative impacts. These include paid care leave (e.g., surveys. In SHARE, carers are asked about how often
Belgium), flexible work schedules (e.g., Australia and the they provided care in the last year and this indicator
United States), respite care (e.g., Austria, Denmark and includes people who provided care at least weekly.
Germany) and counselling/training services (e.g., Sweden). In ELSA, people are asked if they have provided care
Moreover, a number of OECD countries provide cash in the last week, which may be broadly comparable
benefits to family caregivers or cash-for-care allowances with “at least weekly”. Questions in HRS and SDAC
for recipients which can be used to pay informal caregivers are less comparable with SHARE. Carers in HRS are
(Colombo et al., 2011). included if they provided more than 200 hours of care
On average across OECD countries, 60% of those providing in the last year. In SDAC, a carer is defined as someone
daily informal care are women (Figure 11.17). Poland and who has provided ongoing informal assistance for at
Portugal have the greatest gender imbalance, with 70% of least six months. People caring for disabled children
informal carers being women. Sweden is the only country have been excluded for European countries but are
where more men than women report that they provide at included for the United States and Australia. However,
least weekly informal care. the United States data only include those caring for
someone outside of their household. As a result,
Around two thirds of carers are looking after a parent or a
data for Australia and the United States may not be
spouse, but patterns of caring vary for different age groups.
comparable with other countries.
Younger carers (aged between 50 and 65) are much more
likely to be caring for a parent (Figure 11.18). They are more
likely to be women – daughters provide much more care to
their parents than sons – and may not be providing care
References
every day. Carers aged over 65 are more likely to be caring
for a spouse. Caring for a spouse tends to more intensive, Bauer, J.M. and A. Sousa-Poza (2015), “Impacts of Informal
requiring daily care, and men and women are equally likely Caregiving on Caregiver: Employment, Health, and
to take on this role. Family”, Journal of Population Ageing, Vol. 8, No. 3,
The fact that fewer people provide daily care in countries pp. 113-145.
with stronger formal LTC systems suggests that there is Colombo, F. et al. (2011), Help Wanted? Providing and Paying
a trade-off between informal and formal care. Declining for Long-Term Care, OECD Publishing, Paris, http://dx.doi.
family size, increased geographical mobility and rising org/10.1787/9789264097759-en.
208 Health at a Glance 2017 © OECD 2017
11. AGEING AND LONG-TERM CARE
Informal carers
11.16. Share of informal carers among population aged 50 and over, 2015 (or nearest year)
% Daily Weekly Total
25
20
10
15
11
5
8
8
12
7
10
10
6
19
4
6
17
1
5
5
7
2
7
5
11
10
9
9
8
8
8
7
7
7
6
6
6
6
5
4
0
ia
ic
ce
m
li a
r ia
18
nd
ce
ly
el
en
es
nd
ga
m
an
ni
nd
ar
ai
ur
en
bl
ra
It a
iu
an
ee
at
ra
ed
CD
la
la
st
do
Sp
to
r tu
nm
bo
rm
pu
la
lg
Is
ov
St
er
Po
st
Au
Gr
Fr
Sw
Es
ng
er
OE
Be
Po
Re
Au
it z
Ge
De
Sl
d
th
Ki
xe
ite
Sw
h
Ne
Lu
d
ec
Un
ite
Cz
Un
Note: The definition of informal carers differs between surveys (see Definition and comparability).
1. United Kingdom refers to England.
Source: Wave 6 of the Survey of Health, Ageing and Retirement in Europe (2015), Survey of Disability, Ageing and Carers for Australia (2015), wave 7 of
the English Longitudinal Study of Ageing (2015), wave 12 of the Health and Retirement Survey for the United States (2014).
12 http://dx.doi.org/10.1787/888933605939
11.17. Share of women among informal daily carers aged 50 and over, 2015 (or nearest year)
%
80
70.1
69.5
66.0
70
64.3
63.3
62.9
62.4
61.8
61.8
61.2
60.8
59.8
59.1
58.4
57.8
56.8
56.0
55.8
55.7
60
54.5
52.9
45.6
50
40
30
20
10
0
ia
l
nd
ce
es
ly
ic
nd
21
el
r ia
li a
ce
en
ga
an
ni
nd
ar
ai
ur
en
bl
ra
It a
iu
an
ee
at
ra
ed
la
la
do
CD
st
Sp
to
r tu
nm
bo
rm
pu
la
Is
lg
ov
St
Po
er
st
Au
Gr
Fr
Sw
Es
ng
er
OE
Be
m
Po
Re
Au
it z
Ge
De
Sl
d
th
Ki
xe
ite
Sw
h
Ne
Lu
d
ec
Un
ite
Cz
Un
Note: The definition of informal carers differs between surveys (see Definition and comparability).
1. United Kingdom refers to England.
Source: Wave 6 of the Survey of Health, Ageing and Retirement in Europe (2015), Survey of Disability, Ageing and Carers for Australia (2015), wave 7 of
the English Longitudinal Study of Ageing (2015), wave 12 of the Health and Retirement Survey for the United States (2014).
12 http://dx.doi.org/10.1787/888933605958
11.18. Share of informal carers in European1 population aged 50 and over, by recipients of care and age,
daily and weekly, 2015
Weekly care Daily care Weekly care Daily care
Care to parents, % Care to spouse, %
14 14
12 12
10 10
8 8
6 6
4 4
2 2
0 0
50 55 60 65 70 75 80 85+ 50 55 60 65 70 75 80 85+
Age of carer Age of carer
1. Data refer to population aged 50 and over for countries included in SHARE wave 6.
Source: Wave 6 of the Survey of Health, Ageing and Retirement in Europe (2015).
12 http://dx.doi.org/10.1787/888933605977
Health at a Glance 2017 © OECD 2017 209
11. AGEING AND LONG-TERM CARE
Long-term care workers
Long-term care (LTC) is a labour-intensive service. Formal work conditions); and increase productivity (e.g. through
LTC workers are defined as paid staff, typically nurses and reorganisation of work processes and more effective use
personal carers, providing care and/or assistance to people of new technologies) (Colombo et al., 2011; European
limited in their daily activities at home or in institutions, Commission, 2013).
excluding hospitals. Formal care is complemented by
informal, usually unpaid, support from family and friends,
which accounts for a large part of care for older people
in all OECD countries (see indicator on “Informal carers”). Definition and comparability
Relative to the population aged 65 and over, Norway, Sweden Long-term care workers are defined as paid workers
and the United States have the most LTC workers and the who provide care at home or in institutions (outside
Slovak Republic and Portugal the fewest (Figure 11.19). In hospitals). They include qualified nurses and personal
all countries except for Israel, Japan, Estonia and Korea, care workers providing assistance with ADL and other
the majority of LTC staff work in institutions, even though personal support. Personal care workers include
the majority of recipients usually receive care at home different categories of workers who may be called
(see indicator on care recipients). This reflects the fact that under different names in different countries. They may
those in institutions often have more severe needs and have some recognised qualification or not. Because
require more intensive care. personal care workers may not be part of recognised
Most LTC workers are women and work part-time. At least occupations, it is more difficult to collect comparable
90% of LTC workers are women in Korea, Denmark, the data for this category of LTC workers across countries.
Slovak Republic, the Netherlands and Norway (Figure 11.20). LTC workers also include family members or friends
Foreign-born workers also play an important role in LTC who are employed under a formal contract either by
provision, though their presence is uneven across OECD the care recipient, an agency, or public and private
countries. While Germany has very few foreign-born care service companies. They exclude nurses working
LTC workers, nearly one in four care workers in the United in administration. The numbers are expressed as head
States is foreign-born (Colombo et al., 2011). counts, not full-time equivalent.
The LTC sector represents a small but growing share of There are some differences in the methodologies that
total employment in OECD countries, averaging just over countries use to calculate the data, which could bias
2%. The number of LTC workers increased by more than the results. Data for some countries refers only to
50% in Japan, Korea, and Israel between 2005 and 2015 workers employed in the public sector, while other
(Figure 11.21). In Japan and Korea, this is related to the countries include the private and non-profit sectors.
introduction of universal LTC insurance and the increasing Data from the Czech Republic and Japan are based
professionalisation of LTC work. However, the Japanese and on surveys of establishments, meaning that people
Korean populations are ageing rapidly and even with these who work in more than one establishment are double-
changes, the growth in the LTC workforce has only just kept counted.
pace with the growth in the population aged over 80 – the
people most likely to need LTC. In contrast, the number of
long-term care workers decreased in Estonia, the Slovak
Republic, and the Netherlands, despite large increases in References
the population aged 80+ over the same period (Figure 11.21).
Colombo, F. et al. (2011), Help Wanted? Providing and Paying
On average, around one third LTC workers are nurses and
for Long-Term Care, OECD Publishing, Paris, http://dx.doi.
the other two thirds are personal care workers (also referred
org/10.1787/9789264097759-en.
to as nursing aides, health assistants in institutions or
home-based care assistants) with less formal training. OECD and European Commission (2013), A Good Life in Old
Many OECD countries have set educational and training Age? Monitoring and Improving Quality in Long-Term Care,
requirements for personal care workers, although these OECD Health Policy Studies, OECD Publishing, Paris,
vary substantially, especially where home-based care is http://dx.doi.org/10.1787/9789264194564-en.
concerned (OECD/European Commission, 2013). Scheil-Adlung, X. (2015), “Long-term Care (LTC) Protection
As populations continue to age, demand for LTC workers is for Older Persons: A Review of Coverage Deficits in 46
likely to rise. Responding to increasing demand will require Countries”, ESS Working Paper, No. 50, International
policies to improve recruitment (e.g. encouraging more Labour Office, Geneva.
unemployed people to consider training and working in United Nations (2016), “The Growing Need for Long-term
the LTC sector); improve retention (e.g. enhancing pay and Care, Assumptions and Realities”, Briefing Paper.
210 Health at a Glance 2017 © OECD 2017
11. AGEING AND LONG-TERM CARE
Long-term care workers
11.19. Long-term care workers per 100 people aged 65 and over, 2015 (or nearest year)
% Institution Home Institution + Home
14
12
1
10
8
9
13
6
12
1
10
1
4
3
4
1
3
2
1
3
3
3
3
3
3
1 1
2
2
1
1
1
1
0
1
ic
ic
¹
es
el
nd
li a
18
da
n¹
r ia
nd
l
ga
en
ay
an
ni
re
ar
nd
ar
pa
an
bl
bl
ra
ai
at
ra
na
CD
la
la
st
rw
to
r tu
ed
nm
Ko
ng
rm
pu
pu
Sp
Ja
al
la
Is
St
er
Ir e
st
Au
Ca
Es
er
Ze
OE
Sw
No
Hu
Po
Re
Re
Au
it z
Ge
De
d
th
ite
w
Sw
ak
Ne
Ne
ec
Un
ov
Cz
Sl
1. In Norway, Sweden and Spain it is not possible to distinguish LTC workers in institutions and at home. They are not included in the OECD average.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933605996
11.20. Proportion of long-term care workers who are women, 2015 (or nearest year)
%
100
95.0
94.9
93.9
93.8
95
92.5
89.7
89.1
90
87.9
87.9
87.6
86.8
87.0
85
80.3
80
75
70
ic
a
ay
12
nd
r ia
es
el
en
n
an
re
ar
nd
ai
bl
ra
rw
at
ed
CD
la
st
Sp
Ko
nm
rm
pu
la
Is
St
Ir e
Au
Sw
No
er
OE
Re
De
Ge
d
th
ite
ak
Ne
Un
ov
Sl
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933606015
11.21. Long term care workers and population aged 80 and over, 2005 and 2015 (or nearest year)
% change in population aged 80+ 2005-15 % change in number of LTC workers 2005-15
% change
80
61
60
57
55
52
51
47
39
40
37
32
32
30
28
28
26
24
24
21
21
19
18
17
17
20
16
16
14
13
12
8
8
7
6
4
0
0
-20
-20
-29
-40
-37
-60
ic
a
en
r ia
nd
ay
li a
17
es
nd
el
an
ni
nd
ar
re
ar
ai
pa
bl
ra
rw
at
ra
ed
CD
la
la
st
Sp
to
nm
Ko
ng
rm
pu
Ja
la
Is
St
Ir e
er
st
Au
Sw
No
Es
er
OE
Hu
Re
Au
it z
De
Ge
d
th
ite
Sw
ak
Ne
Un
ov
Sl
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933606034
Health at a Glance 2017 © OECD 2017 211
11. AGEING AND LONG-TERM CARE
Long-term care beds in institutions and hospitals
The number of beds in long-term care (LTC) institutions less of a contribution and LTC systems often pick up board,
and in LTC departments in hospitals provides a measure lodging and care costs. Moreover, LTC users generally prefer
of the resources available for delivering LTC services to to remain at home. Most countries have taken steps in recent
individuals outside of their home. years to support this preference and promote community
On average across OECD countries, there were 50 LTC beds care. However, depending on individual circumstances,
per 1 000 people aged 65 and over in 2015 – 46 in LTC a move to LTC institutions may be the most appropriate
institutions and four beds in LTC departments in hospitals option, for example for people living alone and requiring
(Figure 11.22). The Netherlands had the highest number round-the-clock care and supervision (Wiener et al., 2009)
of LTC beds in 2015, with around 87 beds per 1 000 people or people living in remote areas with limited home-care
aged 65 and over, the vast majority of which were in LTC support. It is therefore important that countries retain an
institutions. On the other hand, there were fewer than appropriate level of residential LTC capacity, and that care
20 beds per 1 000 people aged 65 and over in Italy, Latvia, institutions develop and apply models of care that promote
Poland, and Turkey. dignity and autonomy.
On average, there has been almost no change in the
number of LTC beds per 1 000 population over 65 since
2005, though this masks substantial variation between Definition and comparability
countries (Figure 11.23). At one extreme, some countries
Long-term care institutions refer to nursing and
with well-established, comprehensive LTC systems have
residential care facilities which provide accommodation
reduced residential LTC capacity. Between 2005 and 2015,
and long-term care as a package. They include specially
Sweden reduced the number of LTC beds in institutions by
designed institutions or hospital-like settings where
23.5 beds per 1 000 population over 65. These reductions
the predominant service component is long-term
are attributable to a drive to move LTC out of residential
care for people with moderate to severe functional
facilities and into the community (Colombo et al., 2011).
restrictions. Beds in adapted living arrangements for
Iceland, Canada and Norway have also made significant
persons who require help while guaranteeing a high
reductions in the number of beds available. In contrast,
degree of autonomy and self-control are not included.
Korea has seen a massive increase in capacity, increasing
For international comparisons, beds in rehabilitation
the number of beds from 13 to 58 per 1 000 population
centers should not be included.
over 65 between 2005 and 2015, with the increase
particularly marked since the introduction of a public LTC However, there are variations in data coverage across
insurance scheme in 2008. countries. Several countries only include beds in
publicly-funded LTC institutions, while others also
While most countries allocate very few beds for LTC in
include private institutions (both profit and non-for-
hospitals, some still use hospital beds quite extensively for
profit). Some countries also include beds in treatment
LTC purposes. Despite recent increases in the number of
centers for addicted people, psychiatric units of
beds in LTC institutions in Korea, the majority of LTC beds
general or specialised hospitals, and rehabilitation
are still in hospitals – although this may be driven in part
centers.
by the reimbursement rules of the Korean LTC insurance,
which require some facilities to be classified as hospitals.
In Japan, many hospital beds are used for long-term care,
though the number has decreased in recent years. Some References
European countries, such as Finland, Hungary and the
Czech Republic, still have a significant number of LTC beds Colombo, F. et al. (2011), Help Wanted? Providing and Paying
in hospitals, but in general there has been a move towards for Long-Term Care, OECD Publishing, Paris, http://dx.doi.
replacing hospital beds with institutional facilities, which org/10.1787/9789264097759-en.
are often cheaper and provide a better living environment Muir, T. (2017), “Measuring Social Protection for Long-
for people with LTC needs. term care”, OECD Health Working Papers, No. 93, OECD
Providing LTC in institutions can be more efficient than Publishing, Paris, http://dx.doi.org/10.1787/a411500a-en.
community care for people with intensive needs, due to Wiener, J. et al. (2009), “Why Are Nursing Home Utilization
economies of scale and the fact that care workers do not Rates Declining”, Real Choice Systems Change Grant
need to travel to each person separately. However, it often Program, US Department of Health and Human Services,
costs more to public budgets, since informal carers make Centers for Medicare and Medicaid Services.
212 Health at a Glance 2017 © OECD 2017
Ne
-30
-20
-10
0
10
20
30
40
50
th
0
20
40
60
80
100
Ko Lu erla
re xe nd
a +45.4 m s 87.4
Es bo
to
ni u
a 86.3
Be rg
+15.2 lg
Ir e iu
la
nd S m 72.1
Sw we
Sl Ge +6.5
ov rm i t z de n
ak an er 66.4
y la
Re +5.1 n
pu F i d¹ 65.9
Source: OECD Health Statistics 2017.
Source: OECD Health Statistics 2017.
nl
bl a n
Health at a Glance 2017 © OECD 2017
ic Sl
+3.5 ov Ic d 65.3
It a ak ela
ly Re n d
60.6
+3.3
Per 1 000 population aged 65 and over
Fr pu
an bl
ic
ce 58.3
Hu +1.3 Ko
ng Hu r e a
ar Ne ng 58.2
y +1.1 w a
Au
st
Ze r y
al 57.5
r ia an
+1.0
56.1
Be Fr d
lg a
iu
m Ge nce
+0.2 rm 55.7
La
tv Au an
ia st y 54.4
-0.4 ra
Institutions
Po
la Sl lia¹
ov 52.2
nd
OE -0.6 en
CD C a ia 51.7
25 na
De -2.0 d
51.3
Absolute difference in LTC beds per 1 000 population aged 65 and over, 2005-15
nm Ir e a
ar la
k -5.6 Es
nd
51.2
Fi to
nl
an
d -6.4
O E ni a
CD 50.8
Is De 3 1
ra nm 49.7
el
-6.5 ar
Ja Cz
pa e No k 48.9
Au n -7.0 Un ch R r w a
st i t e ep y 48.8
Sw ra
li a d u
K i bli
it z -7.6 ng c 48.0
Un er do
la m
Hospitals
ite nd ¹
Un d -8.0 Sp 47.6
ite St ai
d at
es Un A u n 47.3
Cz
Ki
ng -8.2 ite str
d i
ec
h
do 45.6
m St a
Re -8.8 at
es
pu Ja 36.8
bl
ic pa
-11.4 n
No
Is 34.3
rw ra
ay el
Ca -11.7 25.6
na It a
11.22. Long-term care beds in institutions and hospitals, 2015 (or nearest year)
l
19.2
1. The numbers of long-term care beds in hospitals are not available for Australia, Switzerland, Turkey and the United Kingdom.
da
Ic -12.2 La y
el tv
an Po ia 17.3
d -15.4 la
Sw n
11.23. Trends in long-term care beds in institutions and hospitals, 2005-15 (or nearest year)
ed Tu d 12.3
en rk
-23.5 ey
¹ 8.0
213
12 http://dx.doi.org/10.1787/888933606072
12 http://dx.doi.org/10.1787/888933606053
Long-term care beds in institutions and hospitals
11. AGEING AND LONG-TERM CARE
11. AGEING AND LONG-TERM CARE
Long-term care expenditure
Long-term care (LTC) spending has seen the highest growth rate was 4.6% between 2005 and 2015 across OECD countries
across the various functions (see Indicator on “Health (Figure 11.26). Spending growth stands out for Korea, which
expenditure by type of service”) and is expected to rise has implemented a number of measures to expand the
further in the coming years. Population ageing leads to coverage of their LTC systems in recent years, although
more people needing ongoing health and social care; rising total LTC spending still remains below the OECD average
incomes increase expectations on the quality of life in old as a share of GDP.
age; the supply of informal care is potentially shrinking; Projection scenarios suggest that public resources allocated
and productivity gains are difficult to achieve in such a to LTC as a share of GDP could double or more by 2060 (De
labour-intensive sector. All these factors create upward La Maisonneuve and Oliveira Martins, 2013). One of the
pressures on spending. main challenges in the future will be to strike the right
A significant share of LTC services is paid for out of balance between providing appropriate social protection to
government or compulsory insurance schemes. Total people with LTC needs and ensuring that this protection is
government/compulsory spending on LTC (including both fiscally sustainable.
the health and social care components) accounted for
1.7% of GDP on average across OECD countries in 2015
(Figure 11.24). At 3.7% of GDP, the highest spender was the
Netherlands, where public expenditure on long-term care Definition and comparability
was around double the OECD average. At the other end of the
LTC spending comprises both health and social services
scale, Hungary, Estonia, Poland, Israel and Latvia allocated
to LTC dependent people who need care on an on-
less than 0.5% of their GDP, to the public provision of long-
going basis. Based on the System of Health Accounts
term care. This variation can partly reflect differences in
(SHA), the health component of LTC spending relates
the population structure, but mostly the development
to nursing and personal care services (i.e. help with
of formal LTC systems, as opposed to more informal
activities of daily living (ADL)). It covers palliative care
arrangements based mainly on care provided by unpaid
and care provided in LTC institutions or at home. LTC
family members. Despite problems of underreporting
social expenditure primarily covers assistance with
which limit comparability, available data on privately-
instrumental activities of daily living (IADL). Despite
funded LTC expenditure suggests in some cases it can be
progress made in improving the general comparability
substantial, playing a relatively large role in Switzerland
of LTC spending in recent years there is still some
(0.7% of GDP), Germany and the United Kingdom (both 0.6%).
variation in reporting practices between the health
Consequently, the share of private spending – mainly out-
and social components for some LTC activities across
of-pocket expenditure – in total spending on LTC accounts
countries. In addition, LTC expenditure funded by
for more than 30% in those countries.
governments and compulsory insurance schemes
The boundaries between health and social LTC spending is more suitable for international comparisons as
are still not fully consistent across countries, with some there is more variation in the comprehensiveness in
reporting particular components of LTC as health care, reporting of privately-funded LTC expenditure across
while others view it as social spending. Sweden and Norway OECD countries.
spend 2.5% or more of their GDP on the health part of LTC
Finally, some countries (e.g. Estonia, Israel, and the
financed from government/compulsory schemes, which is
United States) can only report spending data for
around double the OECD average (1.3%). With 1.3% of GDP,
institutional care, and hence underestimate the total
the Netherlands report the highest level of public spending
amount of spending on long-term care services by
on social LTC, much higher than the OECD average of 0.4%.
government and compulsory insurance schemes.
The way LTC is organised in countries affects the
composition of LTC spending and may also have an impact
on overall LTC spending. Across the OECD, two-thirds of
government and compulsory spending on LTC (health) References
was for inpatient LTC in 2015. This is mainly provided
in residential LTC facilities (Figure 11.25). Yet in Poland, De La Maisonneuve, C. and J.O. Martins (2013), “Public
Finland, Denmark, Austria and Germany, spending on Spending on Health and Long-term Care: A New Set of
home-based LTC accounts for more than 50% of all LTC Projections”, OECD Economic Policy Papers, No. 6, OECD
(health) spending. Spending for home-based LTC can be Publishing, http://dx.doi.org/10.1787/5k44t7jwwr9x-en.
either due to services provided by professional LTC workers Muir, T. (2017), “Measuring Social Protection for Long-
or informal workers, when a care allowance exists which term Care”, OECD Health Working Papers, No. 93, OECD
remunerates the caregiver for the LTC services provided. Publishing, Paris, http://dx.doi.org/10.1787/a411500a-en.
Spending by government and compulsory insurance OECD, Eurostat and WHO (2011), A System of Health Accounts,
schemes on LTC has increased more rapidly than health 2011 Edition, OECD Publishing, Paris, http://dx.doi.
care expenditure over the last decade. The annual growth org/10.1787/9789264270985-en.
214 Health at a Glance 2017 © OECD 2017
11. AGEING AND LONG-TERM CARE
Long-term care expenditure
11.24. Long-term care expenditure (health and social components) by government and compulsory
insurance schemes, as a share of GDP, 2015 (or nearest year)
Long-term care (health) Long-term care (social)
% GDP
4 3.7
3.2
3
2.5 2.5
2.3 2.2
2.0
2 1.8 1.7 1.7 1.7
1.5 1.4 1.3 1.3 1.3 1.2 1.2
0.9
1 0.8 0.8 0.7
0.5 0.5 0.4 0.4 0.4
0.2 0.2
Ge lic
ia
Sw s
De en
No k
Be ay
m
Re n d
xe a ny
Au g
Ic n
d
C a ia
Sl da
Po al y
St l
es
ia
Po l
Es d
Hu a
y
Un O nd
Ki 15
ite uga
e
c
ni
re
nd
ar
ar
pa
ai
n
an
an
ur
en
ra
tv
iu
do
rw
Sw an
at
ed
na
D
la
la
st
It
Sp
to
nm
Ko
ng
bo
Lu r m
pu
nl
e c Ir e l
Ja
el
la
La
lg
Is
ite EC
ov
er
Un r t
ng
Fr
er
Fi
m
it z
d
th
h
Ne
Cz
Note: The OECD average only includes the 15 countries that report health and social LTC.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933606091
11.25. Government and compulsory insurance spending on LTC (health) by mode of provision, 2015
(or nearest year)
% Inpatient long-term care Home-based long-term care Other
3 1 3 2 1 1 1
100 1 4 6 5 6 4 5
7 12 13 11
11 13 17 16 17
14 13 23 26 27 31
80 23 30 31 44 48
15 41 41
23 54 57 64 70
60 18
85
96 92 88 87 86 85 83
40 82 82 77 74 73 68 66 65 64 60 55 55 55 52 51
45
20 42
36 30
14
0
ic
ia
y
ce
ly
r ia
d
nd
y
a
da
a
nd
ia
n
26
en
nd
ay
ga
an
ni
re
nd
ar
ar
pa
ai
an
an
ur
bl
en
tv
It a
iu
do
an
rw
ed
na
la
la
la
st
CD
Sp
to
r tu
Ko
nm
ng
bo
rm
pu
nl
el
Ja
La
la
lg
ov
er
Po
Ir e
ng
Au
Fr
Sw
Ca
No
Es
Ic
er
Fi
OE
Be
Hu
m
Po
Re
it z
Ge
De
Sl
Ki
th
xe
Sw
h
Ne
d
Lu
ec
ite
Cz
Un
Note: “Other” includes LTC day cases and outpatient LTC.
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933606110
11.26. Annual growth rate in expenditure on LTC (health and social) by government and compulsory
insurance schemes, in real terms, 2005-15 (or nearest year)
% 32.1
10
8.2
8
6
5.1 5.0 4.9
4.6 4.6 4.4 4.3 4.2 4.2 4.0
4 3.4 3.3 3.3
2.9 2.8 2.8 2.7 2.5
2.1 2.0 2.0 1.8
2 1.4
0.3
0
ic
nd
ia
a
ce
n
l
ay
nd
ia
y
25
s
r ia
el
da
en
es
y
ga
an
ni
re
nd
ar
ar
pa
ai
ur
an
an
bl
en
ra
tv
iu
rw
an
at
ed
na
la
la
CD
st
Sp
to
r tu
Ko
nm
ng
bo
rm
pu
Ja
nl
el
La
la
lg
Is
ov
St
Po
er
Au
Fr
Sw
Ca
No
Es
Ic
er
Fi
OE
Be
Hu
m
Po
Re
it z
Ge
De
Sl
d
th
xe
ite
Sw
h
Ne
Lu
ec
Un
Cz
Source: OECD Health Statistics 2017.
12 http://dx.doi.org/10.1787/888933606129
Health at a Glance 2017 © OECD 2017 215
ORGANISATION FOR ECONOMIC CO-OPERATION
AND DEVELOPMENT
The OECD is a unique forum where governments work together to address the economic, social and
environmental challenges of globalisation. The OECD is also at the forefront of efforts to understand and
to help governments respond to new developments and concerns, such as corporate governance, the
information economy and the challenges of an ageing population. The Organisation provides a setting
where governments can compare policy experiences, seek answers to common problems, identify good
practice and work to co-ordinate domestic and international policies.
The OECD member countries are: Australia, Austria, Belgium, Canada, Chile, the Czech Republic,
Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Israel, Italy, Japan, Korea,
Latvia, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the
Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States.
The European Union takes part in the work of the OECD.
OECD Publishing disseminates widely the results of the Organisation’s statistics gathering and
research on economic, social and environmental issues, as well as the conventions, guidelines and
standards agreed by its members.
OECD PUBLISHING, 2, rue André-Pascal, 75775 PARIS CEDEX16
(81 2017 30 1P) ISBN 978-92-64-28039-7 – 2017
Health at a Glance 2017
OECD Indicators
Health at a Glance provides the latest comparable data and trends on different aspects of the performance of
health systems in OECD countries. It provides striking evidence of large variations across countries in indicators
of health status and health risks, as well as in the inputs and outputs of health systems.
This new edition of Health at a Glance presents the most recent comparable data on the health status of
populations and health system performance in OECD countries. Where possible, it also reports data for partner
countries (Brazil, China, Colombia, Costa Rica, India, Indonesia, Lithuania, Russian Federation and South
Africa). The data presented in this publication come from official national statistics, unless otherwise stated.
This edition contains a range of new indicators, particularly on risk factors for health. It also places greater
emphasis on time trend analysis. Alongside indicator-by-indicator analysis, this edition offers snapshots and
dashboard indicators that summarise the comparative performance of countries, and a special chapter on the
main factors driving life expectancy gains.
Consult this publication on line at http://dx.doi.org/10.1787/health_glance-2017-en.
This work is published on the OECD iLibrary, which gathers all OECD books, periodicals and statistical databases.
Visit www.oecd-ilibrary.org for more information.
isbn 978-92-64-28039-7
81 2017 30 1 P
9HSTCQE*ciadjh+
You might also like
- Health at A Glance Europe 2010 1st Edition Oecd Updated 2025100% (2)Health at A Glance Europe 2010 1st Edition Oecd Updated 2025171 pages
- OCDE - Health at A Glance 2019 - 4dd50c09-En PDFNo ratings yetOCDE - Health at A Glance 2019 - 4dd50c09-En PDF243 pages
- MOD1 LO1 Pop Health Diffs - WHO Stats For SDGs in 2018No ratings yetMOD1 LO1 Pop Health Diffs - WHO Stats For SDGs in 201856 pages
- Report-World Health Statistics 2019-Monitoring Health For The SDGsNo ratings yetReport-World Health Statistics 2019-Monitoring Health For The SDGs132 pages
- Iii. Quality of Life: © OECD 2011 Compendium of OECD Well-Being IndicatorsNo ratings yetIii. Quality of Life: © OECD 2011 Compendium of OECD Well-Being Indicators17 pages
- OECD Health Statistics 2015 Frequently Requested DataNo ratings yetOECD Health Statistics 2015 Frequently Requested Data341 pages
- Global Health - 4th Edition Ebook Download100% (14)Global Health - 4th Edition Ebook Download15 pages
- 9789241565486-engINFORME Estadísticas Sanitarias Mundiales Que Publica Anualmente La OMSNo ratings yet9789241565486-engINFORME Estadísticas Sanitarias Mundiales Que Publica Anualmente La OMS116 pages
- THE Cost of Staying Healthy: Semiannual Report of The Latin America and The Caribbean RegionNo ratings yetTHE Cost of Staying Healthy: Semiannual Report of The Latin America and The Caribbean Region76 pages
- Determining Health Expectancies 1st Edition Jean-Marie Robine Available Full ChaptersNo ratings yetDetermining Health Expectancies 1st Edition Jean-Marie Robine Available Full Chapters156 pages
- Global Health: Diseases, Programs, Systems, and Policies. ISBN 128412262X, 978-1284122626100% (16)Global Health: Diseases, Programs, Systems, and Policies. ISBN 128412262X, 978-128412262623 pages
- Curated en 831551468252030532 Text 328300PAPER0AFRHDWPS1No142No ratings yetCurated en 831551468252030532 Text 328300PAPER0AFRHDWPS1No142151 pages
- Informe de La OMS Sobre Pobreza y Salud Publica 2016No ratings yetInforme de La OMS Sobre Pobreza y Salud Publica 201656 pages
- Determinants of Health - 250204 - 162542No ratings yetDeterminants of Health - 250204 - 16254255 pages
- Health at A Glance Organization For Economic Co-Operation and Development - Downloadable PDF 2025No ratings yetHealth at A Glance Organization For Economic Co-Operation and Development - Downloadable PDF 202552 pages
- Global Health - 4th Edition No-Wait DownloadNo ratings yetGlobal Health - 4th Edition No-Wait Download14 pages
- Comparing Health System Performance in OECD CountriesNo ratings yetComparing Health System Performance in OECD Countries14 pages
- From Evidence Into Action: Opportunities To Protect and Improve The Nation's HealthNo ratings yetFrom Evidence Into Action: Opportunities To Protect and Improve The Nation's Health28 pages
- Global Health - Diseases, Programs, Systems, and Policies 4th Edition50% (2)Global Health - Diseases, Programs, Systems, and Policies 4th Edition993 pages
- The Oxford Handbook of Health Economics 9780199238 250320 185236No ratings yetThe Oxford Handbook of Health Economics 9780199238 250320 185236992 pages
- SPH200 - Fall2025 - Tuesday Lecture - Unit 3 - All ModulesNo ratings yetSPH200 - Fall2025 - Tuesday Lecture - Unit 3 - All Modules34 pages
- The Essential Guide To Primary Care Procedures 2nd Edition Complete Volume DownloadNo ratings yetThe Essential Guide To Primary Care Procedures 2nd Edition Complete Volume Download14 pages
- Hand Hygiene Audit To Study Adherence To Hand Hygiene in A Tertiary Care HospitalNo ratings yetHand Hygiene Audit To Study Adherence To Hand Hygiene in A Tertiary Care Hospital4 pages
- Mx2pem-03222024 Pemeresult 03222024 20240322 125904 0000No ratings yetMx2pem-03222024 Pemeresult 03222024 20240322 125904 00001 page
- NURS FPX 6618 Assessment 2 Mobilizing Care For An Immigrant PopulationNo ratings yetNURS FPX 6618 Assessment 2 Mobilizing Care For An Immigrant Population5 pages
- Economics Evaluation of Health Care ProgrammeNo ratings yetEconomics Evaluation of Health Care Programme16 pages
- Planned Medical Treatment Abroad - Your EuropeNo ratings yetPlanned Medical Treatment Abroad - Your Europe3 pages
- Occupational Therapy Clinical Practice Guideline - Home Oxygen Therapy For Adult Patients - 3No ratings yetOccupational Therapy Clinical Practice Guideline - Home Oxygen Therapy For Adult Patients - 392 pages
- William T. O'Donohue - Jeanne Wendel - Teresa D. Serratt - Understanding Healthcare Economics - Managing Your Career in An Evolving Healthcare System (2018)No ratings yetWilliam T. O'Donohue - Jeanne Wendel - Teresa D. Serratt - Understanding Healthcare Economics - Managing Your Career in An Evolving Healthcare System (2018)287 pages
- Source Answer Legal Basis: Questions) Overtime Work. Work May BeNo ratings yetSource Answer Legal Basis: Questions) Overtime Work. Work May Be4 pages
- NURS FPX 6016 Assessment 3 Data Analysis and Quality Improvement Initative ProposalNo ratings yetNURS FPX 6016 Assessment 3 Data Analysis and Quality Improvement Initative Proposal5 pages
- Improving Inpatient Medication Dispensing With AnNo ratings yetImproving Inpatient Medication Dispensing With An9 pages
- Catalbas, Joshua Galura 13588176 80031716 11,000.00 07/15/2023 09/08/2023No ratings yetCatalbas, Joshua Galura 13588176 80031716 11,000.00 07/15/2023 09/08/20234 pages
- Stroke Rehabilitation Framework 2022 Update FinalNo ratings yetStroke Rehabilitation Framework 2022 Update Final29 pages
- Health System Reforms in Five Sub-Saharan African Countries That Experienced Major Armed Conflicts Wars During 1990 2015 A Literature ReviewNo ratings yetHealth System Reforms in Five Sub-Saharan African Countries That Experienced Major Armed Conflicts Wars During 1990 2015 A Literature Review16 pages
