Public Health Surveillance

Goals
 Define surveillance, explain
surveillance systems
 Describe basic surveillance
techniques by person, place, time
 Touch on importance of
standardization
 Provide overview of how to present
surveillance data
What Is Surveillance?
 Centers for Disease Control and
Prevention (CDC): epidemiologic
surveillance is “ongoing systematic
collection, analysis, and interpretation
of health data essential to the
planning, implementation, and
evaluation of public health practice,
closely integrated with the timely
dissemination of these data to those
who need to know.”
• Surveillance provides the information for
descriptive epidemiology, which is
• Person (age, sex, description)
• Place (where)
• Time (dates, hours, days, months, years)
“In public health,
we can’t do
anything
without
surveillance.
That’s where
public health
begins.“
Why Is Surveillance
Important?
 Collecting data is merely one step
 Critical goal is to control and/or
prevent diseases
 Any data collected must be organized
and carefully examined
 Any results need to be communicated
to public health and medical
communities
Why Is Surveillance
Important?
 Vital to communicate results
 During potential outbreak so public
health and medical communities can
help with disease prevention and control
efforts
 During non-outbreak times to provide
information about baseline levels of
disease

Baseline provides information to public health
officials monitoring health at community
level, serves as reference in future outbreaks
 Why Do We Use
Public Health Surveillance?

• Because it can
– estimate the size of a health problem
– determine where an illness is occurring geographically
– portray the natural history of a disease
– detect epidemics or define a problem
– generate hypotheses in research
– monitor changes in infectious agents
– detect changes in health practices
– facilitate emergency planning
 Who Performs
Public Health Surveillance?

• Epidemiologists
• Medical professionals (doctors, nurses, clinics)
• Pharmacies
• Health insurance providers
• Emergency responders (9-1-1 centers, emergency
management technicians)
• Public health departments (local, state, federal)
 Where Do We Get
Public Health Surveillance Data?

• Vital records
– Hospital records, death certificates, birth records

• Surveys
–
Schools, doctors, insurance companies

• Environmental monitoring systems

– Water or air quality

• Animal health data

– Veterinarians, farms, food manufacturing
 How Are Public Health Surveillance
Data Collected?
• Voluntary reporting
– Individual persons reporting (a person, doctor, or

hospital)
– Media reports

– Unusual health occurrences

• Higher purchases of specific pharmaceuticals;

excessive school absences

• Public health initiated surveys
– Calling, completing forms, reviewing public records
 How Are Public Health Surveillance
Data Collected? con’t.

• Mandatory reporting
• Some diseases are required to be reported to the local
and state health departments (reportable diseases).
– Highly communicable (transmittable) diseases

• E.g., chickenpox

– High morbidity or mortality rates

• E.g., Ebola virus disease

– Strong public interest

• E.g., methicillin-resistant Staphylococcus aureus

(often referred to as MRSA)

 Public Health Surveillance
Types

• Different ways of collecting surveillance data fall

into 3 categories
– Passive

– Active

– Syndromic
 Passive Surveillance

• Laboratories, physicians, or others regularly report cases

of disease or death to the local or state health
department

• Examples
– A doctor’s office reports 2 cases of measles
– A nursing home reports an unusual number of older
patients with unexplained rashes
 Active Surveillance

• Local or state health departments initiate the collection of

information from laboratories, physicians, health care
providers, or the general population
– Achieves more complete and accurate reporting than
passive surveillance
– Example: Youth Risk Behavior Surveillance surveys
 Syndromic Surveillance

• The ongoing, systematic collection, analysis,

interpretation, and application of real-time indicators for
disease that allow for detection before public health
authorities might otherwise identify them
• Example: Hospital admittance records
 What Does an Ideal
Surveillance System Look Like?

• Simple
• Timely
• Representative
• Flexible
• Sensitive
• Strong predictive value
• Acceptable to
– The public

– Health care providers

• Cost-effective
 Purposes of
Public Health Surveillance
• Surveillance has been used for the following types of
outbreaks:
– Food poisoning
– Cancer clusters
– Health threats after natural or human-made disasters
• Cholera in Haiti after the 2010 earthquake
• Air quality after the September 11, 2001, attacks
 Purposes of
Public Health Surveillance, con’t.

• Infectious disease outbreaks

– Ebola
– Malaria
– Influenza
– Severe acute respiratory syndrome (SARS)
– Chronic health problems
– Obesity
– Cardiovascular diseases
– Asthma
How to Conduct
Surveillance
 Surveillance data allow description
and comparison of patterns of
disease by person, place, and time
 Several ways to describe and
compare patterns, from
straightforward presentations to
statistically complex analyses
 Will concentrate on simple
techniques
How to Conduct
Surveillance: Person
 When available, demographic
characteristics such as gender, age,
race/ethnicity, occupation, education level,
socio-economic status, sexual orientation,
immunization status can reveal disease
trends
 Example: looking at Streptococcus pneumoniae,
a common cause of community-acquired
pneumonia and bacterial meningitis, examining
distribution of cases by race provides important
information about burden of disease in different
populations
How to Conduct
Surveillance: Person –
Numbers and Rates
 Table 1 shows data
collected on
Streptococcus
pneumoniae from
CDC Emerging
Infections Program
Network, a
surveillance program
that collects data
from multiple counties
in 10 US states (4)
How to Conduct
Surveillance: Person –
Numbers and Rates
 Data show majority of
cases reported among
whites
 Can draw only limited
conclusions because
race not recorded for
684 cases (15%)
 Shows only number of
reported cases, not rate
 Total number of
individuals by race
needed to determine if
there is a
disproportionate burden
of disease among races
How to Conduct
Surveillance: Person –
Numbers and Rates
 Table 2 shows
same data with
2006 population
estimates of total
number of
persons in each
racial category
used to calculate
disease rates (4)
How to Conduct
Surveillance: Person –
Numbers and Rates
 While Table 1 showed
that whites had the
highest number of
cases, Table 2
indicates that the rate
of disease was highest
among blacks
 Using rates, stratifying
by race provides
information about
disease burden in
different populations
that would not be
apparent from total
case numbers
More on Rates
 Rates—A rate is “an expression of the
frequency with which an event occurs in a
defined population”
 Using rates rather than raw numbers is
essential to compare different classes of
persons or populations at different times or
places. (5)

Rate = number of events in a specified

period
average population during the
period
How to Conduct
Surveillance: Place
 Best to characterize cases by place of
exposure rather than by place at which cases
reported
 The two may differ and place of exposure is
more relevant to epidemiology of a disease
 Example: travelers on a cruise ship exposed to a
disease just prior to disembarking but become
symptomatic and are diagnosed after return to
various home locations
 Example: person exposed to disease in small rural
town but referred to tertiary care center 100 miles
away where disease is diagnosed and reported
How to Conduct
Surveillance: Place –
Presenting Data
 Data by geographic location can be
presented in a table
 Also helpful to use maps to facilitate
recognition of spatial associations in data
 See FOCUS Volume 5, Issue 2: Mapping for
Surveillance and Outbreak Investigation for
discussion of maps and visual presentation of
information
 Inferential analysis can also be done using
multilevel modeling, other statistical
methods
How to Conduct
Surveillance: Place –
Modeling Resources
 Modeling of surveillance data by place is
described below.
 Resources for further information:
 Centers for Disease Control and Prevention.
Resources for creating public health maps.
http://www.cdc.gov/epiinfo/maps.htm. Updated
August 14, 2008. Accessed August 22, 2008.
 Clarke KC, McLafferty SL, Tempalski BJ. On
epidemiology and geographic information
systems: A review and discussion of future
directions. Emerg Infect Dis. 1996; 2(2):85-92.
How to Conduct
Surveillance: Place – Spot
Maps
 Spot maps: maps on which a dot or symbol
marks a case of disease
 Made by indicating exposure locations of
reported cases of disease on hard copy map
with pins or colored pen
 Or with geographic information systems (GIS)
 Computer programs designed for storing,
manipulating, analyzing, and displaying data in a
geographic context
 Very useful for mapping surveillance data by place
 Epi Map (part of Epi Info™) can be downloaded for free
at http://www.cdc.gov/epiinfo to assist with map
making
How to Conduct
Surveillance: Place – Spot
Maps
 Example: spot map used to
show geographic spread of
cases in 1995 outbreak of
toxoplasmosis thought to
be associated with a
municipal water system in
British Columbia, Canada
(5)
 Spot maps show
geographic distribution of
cases but not population
size at each location, so
should not be used to
assess disease risk
How to Conduct
Surveillance: Time
 Compare number of cases reported in
time period of interest (weeks, months,
years) to number of cases reported
during similar historical period
 Usually a delay (sometimes months to
years) between disease onset and date
when disease is reported, so preferable
to use date of onset, if available, rather
than date of report
How to Conduct
Surveillance: Time – Line
Graphs
 Especially helpful for examining data
not likely to have much short term
variation
 Example: there is limited variation in
number of AIDS cases reported each month
 Provide valuable qualitative
information; disease outbreaks often
obvious from visual inspection of data,
may not require a quantitative analysis
How to Conduct
Surveillance: Time – Line
Graphs
 Example of line graph
using fabricated
data: reported cases
of Salmonella
typhimurium for 2-
year time intervals
from 1974 to 2002
 Spike in 1994
indicating outbreak of
S. typhimurium
obvious without
quantitative analysis
How to Conduct
Surveillance: Time –
Incidence Rates
 May use line graph to plot incidence rates
 Incidence rate is number of new cases that
occur during a specified time interval in a
population at risk for developing the disease

Number of new cases may be used as a proxy for
overall disease occurrence
 Reporting incidence rates rather than numbers
particularly important if population has
changed in size or characteristics

Example: addition of towns to a surveillance region
has increased population size, or influx of migrant
workers has significantly changed the demographics


Incidence = (New Cases) / (Population x
Timeframe)
Standardization
 Rate made up of numerator and denominator
 Surveillance data often numerator data
(number of cases reported in time period)
 Utility of these raw numbers is limited because do
not take into account size of population or
distribution of demographic factors such as age or
gender
 Rates allow more meaningful comparisons
over time within a population, among
subpopulations, or between populations
 Rates take into account size of the population and
time period involved
Standardization
 Crude rates often calculated using surveillance data
 Number of events of interest (such as reported
cases of disease) for a specific period of time for
the entire population
 Only appropriate to compare crude rates if
populations are similar with respect to factors
related to disease of interest, such as age, gender,
race
 Example: would be inappropriate to compare rate of
prostate cancer in population with high proportion of
elderly men to rate in another population with mostly
young men, since risk of prostate cancer increases with
age
Standardization
 Standardization used to remove effects of
differences in confounding variables such as
age when comparing two or more populations
 Results in adjusted rates
 Is particularly useful when comparing rates in
different populations (e.g., comparing state data to
national data) when comparison of crude rates may
be misleading if populations differ on key variables
 Most common technique uses weighted
average rates specific to potential confounding
variables, based on specified distribution of
the variables
 Data Presentation
 Surveillance data must be presented in way that
is easy to understand and interpret
 Many ways to display surveillance data:
 Line graphs for displaying data by time
 Maps for presenting data in geographic context
 Graphical displays such as histograms, frequency
polygons, box plots, scatter diagrams, bar charts, pie
charts, or stem-and-leaf displays
 Spot or chloropleth maps (A choropleth map is a
thematic map that is used to represent statistical data
using the color mapping symbology technique.)
 Single/multivariable tables
Data Presentation
 The choice of a particular graph or table
depends on type of data, but presentation
should be simple and easy to follow
 Should provide all information necessary to
interpret the figure without referring to text
 Include concise title that describes subject or
disease, time, place (when relevant)
 Define any abbreviations or symbols
 Note any data exclusions
Data Presentation
 Additional
display
guidelines
for tables
and
graphs
Conclusion
 Surveillance is valuable epidemiologic
tool that can serve many purposes
 When surveillance data is collected,
analyzed, interpreted, reported
appropriately, these data can provide
important information about disease
patterns to inform public health
practice and policy