See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/375059863

Stroke incidence, mortality, subtypes in rural and urban populations in five

geographic areas of India (2018-2019): results from the National Stroke
Registry Programme

Article in The Lancet Regional Health - Southeast Asia · October 2023

DOI: 10.1016/j.lansea.2023.100308

CITATIONS READS

0 66

28 authors, including:

Sukanya Rangamani Deepadarshan Huliyappa

National Centre for Disease Informatics and Research (NCDIR),Indian Council of M… ICMR National Centre for Disease Informatics and Research
18 PUBLICATIONS 96 CITATIONS 4 PUBLICATIONS 20 CITATIONS

SEE PROFILE SEE PROFILE

Soumyadarshan Nayak
MKCG Medical College and Hospital
24 PUBLICATIONS 72 CITATIONS

SEE PROFILE

All content following this page was uploaded by Soumyadarshan Nayak on 29 October 2023.

The user has requested enhancement of the downloaded file.

 Articles

Stroke incidence, mortality, subtypes in rural and urban

populations in five geographic areas of India (2018–2019):
results from the National Stroke Registry Programme
Sukanya Rangamani,a Deepadarshan Huliyappa,a Vaitheeswaran Kulothungan,a Sankaralingam Saravanan,b P. K. Murugan,b Radha Mahadevan,b
Chelladurai Rachel Packiaseeli,b Esakki Bobby,b Kandasamy Sunitha,b Ashok Kumar Mallick,c Soumya Darshan Nayak,c Santosh Kumar Swain,c
Manoranjan Behera,c Bhaskar Kanti Nath,d Abhijit Swami,d Amit Kumar Kalwar,d Bijush Difoesa,d Vijay Sardana,e Dilip Maheshwari,e
Bharat Bhushan,e Deepika Mittal,e Rameshwar Nath Chaurasia,f L. P. Meena,g K S Vinay Urs,a Rahul Rajendra Koli,a Natesan Suresh Kumar,a and
Prashant Mathura,∗
a
ICMR-National Centre for Disease Informatics and Research, II Floor of Nirmal Bhawan, ICMR Complex Kannamangala Post, Bengaluru,
562 110, India
b
Tirunelveli Medical College, Tirunelveli, 627011, India
c
SCB Medical College & Hospital, Cuttack, Behera Colony, Mangalabag, Cuttack, Odisha, 753001, India
d
Silchar Medical College, Beside Indian Post, Ghungoor, Masimpur, Silchar, Assam, 788014, India
e
Govt Medical College, MBS Hospital, Nayapura, Kota, Rajasthan, 324001, India
f
Department of Neurology, Institute of Medical Sciences, Banaras Hindu University, Varanasi, Uttar Pradesh, 221005, India
g
Institute of Medical Sciences, Banaras Hindu University, Varanasi, Uttar Pradesh, 221005, India

Summary The Lancet Regional

Health - Southeast
Background Increasing stroke burden in India demands a long-term stroke surveillance framework. Earlier studies in
Asia 2023;▪: 100308
India were urban-based, short term and provided limited data on stroke incidence and its outcomes. This gap is
Published Online XXX
addressed by the establishment of five population-based stroke registries (PBSRs) of the National Stroke Registry
https://doi.org/10.
Programme, India. This paper describes stroke incidence, mortality and age, sex, and subtypes distribution in the 1016/j.lansea.2023.
five PBSRs with urban and rural populations. 100308

Methods First-ever incident stroke patients in age group ≥18 years, resident for at least one year in the defined
geographic area, identified from health facilities were registered. Death records with stroke as the cause of death from
the Civil Registration System (CRS) were included. Transient ischemic attack (TIA) was excluded. Three PBSRs
(Cuttack, Tirunelveli, Cachar) included urban and rural populations. PBSRs in Kota and Varanasi were urban areas.
The crude and age-standardized incidence rate (ASR) by age, sex, and residence (urban and rural), rate ratios of ASR,
case fatality proportions and rates at day 28 after onset of stroke were calculated for years 2018–2019.

Findings A total of 13,820 registered first-ever stroke cases that included 985 death certificate-only cases (DCOs) were
analysed. The pooled crude incidence rate was 138.1 per 100,000 population with an age-standardized incidence rate
(ASR) of 103.4 (both sexes), 125.7 (males) and 80.8 (females). The risk of stroke among rural residents was one in
seven (Cuttack), one in nine (Tirunelveli), and one in 15 (Cachar). Ischemic stroke was the most common type in
all PBSRs. Age-standardized case fatality rates (ASCFR) per 100,000 population for pooled PBSRs was 30.0
(males) and 18.8 (females), and the rate ratio (M/F) ranged from 1.2 (Cuttack) to 2.0 (Cachar).

Interpretation Population-based registries have provided a comprehensive stroke surveillance platform to measure
stroke burden and outcomes by age, sex, residence and subtype across India. The rural–urban pattern of stroke
incidence and mortality shall guide health policy and programme planning to strengthen stroke prevention and
treatment measures in India.

*Corresponding author.
E-mail addresses: director-ncdir@icmr.gov.in (P. Mathur), sukanya.r@icmr.gov.in (S. Rangamani), deepadarshan.h@icmr.gov.in (D. Huliyappa),
vaitheeswaran.k@icmr.gov.in (V. Kulothungan), drsaravananneuro@gmail.com (S. Saravanan), dr_murugan2002@yahoo.com (P.K. Murugan),
radha_m@tvmc.ac.in (R. Mahadevan), rachelseeli72@gmail.com (C. Rachel Packiaseeli), bobbyguru74@gmail.com (E. Bobby), sunitha_s@tvmc.ac.in
(K. Sunitha), drashokkumarmallik@gmail.com (A.K. Mallick), Soumya_bapu@yahoo.co.in (S.D. Nayak), drmedsantoshswain@gmail.com (S.K.
Swain), doctor_manoranjan@rediffmail.com (M. Behera), drbknath@gmail.com (B.K. Nath), drabhijitswami@gmail.com (A. Swami), amit.kal-
war01@gmail.com (A.K. Kalwar), difoesabijush@gmail.com (B. Difoesa), vsard13@gmail.com (V. Sardana), dilipsoni2004@hotmail.com (D.
Maheshwari), drbushan90@yahoo.com (B. Bhushan), drmittaldeepika@gmail.com (D. Mittal), goforrameshwar@gmail.com (R.N. Chaurasia), drla-
litmeena@gmail.com (L.P. Meena), vinay.urs@icmr.gov.in (K.S. Vinay Urs), rahulr.k@icmr.gov.in (R.R. Koli), suresh.ku@icmr.gov.in (N. Suresh
Kumar).

www.thelancet.com Vol ▪ ▪, 2023 1

 Articles

Funding The National Stroke Registry Programme is funded through the intramural funding of the Indian Council of
Medical Research, Department of Health Research, Ministry of Health and Family Welfare, India.

Copyright © 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Stroke incidence; Subtypes; Mortality; Population-based stroke registry; Rural; Urban; India

Research in context
Evidence before this study onset of stroke through multiple methods. This is a first of its
Reliable population level data on stroke incidence & mortality kind study to include data on fatal events (recorded in Civil
is necessary for stroke surveillance and health programme Registration System) and non-fatal non-hospitalized events of
planning. The evidence available till date from population- stroke. The study has produced robust estimates on incidence
based stroke registries and community based surveys was of first-ever stroke and mortality in rural and urban
from urban cities with little data from few rural populations in populations.
India. A most recent systematic review on incidence,
Implications of all the available evidence
prevalence, and case fatality of stroke in India concluded that
The findings from this study provides comparable data on
further high-quality evidence was required that used WHO
incidence & mortality of stroke in different regions of the
STEP wise approach to stroke surveillance with longitudinal
country. It describes the burden in these areas and provides
data collection from populations across India. There is need
baseline evidence for planning of stroke prevention activities
for robust estimates on burden of stroke and its subtypes by
and its control strategies. The evaluation of urban/rural
age, sex and residence, to guide policy makers and strengthen
difference in burden of stroke and availability of imaging
stroke care services in India.
facilities for diagnosis of stroke shall envisage proper resource
Added value of this study allocation and healthcare planning. The National Stroke
The five population-based stroke registries established in five Registry Programme shall establish the stroke surveillance
regions of India under ICMR-NCDIR’s National Stroke Registry system and support the ‘National Programme for prevention
Programme, provides reliable estimates on the incidence and and control of noncommunicable diseases (NP-NCD)’ in
mortality of first-ever stroke in defined populations. The prevention and control of stroke in India. The evidence from
registry model integrates the STEPS approach to stroke India shall support global efforts in burden estimations and in
surveillance through case-finding in hospitals and imaging monitoring of the NCD targets towards attaining SDG of
centres with follow-up data on casefatality on 28 days after health and well-being.

Introduction population-based registries (PBSR) (rural and urban) to

Stroke is one of the leading non-communicable diseases establish stroke surveillance system in India. The main
(NCDs) causing significant death and disability in In- objective of PBSR was to generate reliable measure-
dia.1 Population based studies on incidence and mor- ments of stroke incidence and mortality.4 Disease reg-
tality of stroke conducted over the last three decades istries collect data on a continuous basis on first-ever
(1990–2020) in India were of short duration and pre- stroke, its subtypes and outcomes at day 28 after onset
dominantly in urban populations.2 The global burden of of stroke. This paper presents the measurements of
diseases study (GBD) showed state level variation in stroke incidence and mortality and its age, sex, and
stroke incidence and DALYs rate of stroke linked to the subtypes distribution in five population-based stroke
demographic and epidemiological transition in states in registries, with urban -rural differences for the period
India.1 Globally, higher stroke incidence and poor out- 2018–2019.
comes have been found in rural residents as compared
to urban residents. This is linked to higher prevalence of
stroke risk factors in rural areas.3 Literature on stroke Methods
incidence, its subtypes, mortality and access to stroke PBSRs were established in five geographical areas
services in rural India is limited.2 Thus longitudinal across different regions of India covering a population
studies in urban and rural populations are necessary to of 1 million and above. These include A. Registries with
generate evidence on stroke burden in different regions urban and rural areas—PBSR-Cuttack (east), PBSR-
of India so as to aid planning for preventive and curative Cachar (north-east), PBSR-Tirunelveli (south) and B.
stroke services.2 In this context, the National Stroke Registries with urban areas only—PBSR-Kota (west),
Registry Programme, India had initiated five PBSR-Varanasi (north) (Fig. 1). The census definitions

2 www.thelancet.com Vol ▪ ▪, 2023

 Articles

Fig. 1: Area and population (2018–2019) of five population based stroke registries (PBSRs), India.

of rural and urban composition and its population were registration system (CRS). The PBSR team of field in-
used to characterize each PBSR area. The PBSR centres vestigators collected data on first-ever incident stroke
are major hospitals for stroke management in these cases in age ≥18 years who were residents for at least
geographic areas. Registry study design ensured that one year before the diagnosis of stroke, in the defined
data was continuously collated from facilities (called as geographic area. The residential address was obtained
Sources of registration -SoR) that refer, diagnose, or from the patient registration slips, medical records or by
treat stroke patients. These included hospitals, nursing interview of care givers and categorised based on the
homes, clinics, general physicians, imaging centres, urban ward/area/town or the rural village of the regis-
physiotherapy and rehabilitation centres, and the civil tered case. Transient ischemic attacks (TIAs), traumatic

www.thelancet.com Vol ▪ ▪, 2023 3

 Articles

intracranial haemorrhage, symptoms due to trauma, stroke and stroke subtype. Following verification by the
coma of systemic vascular origin, vascular dementia, respective registry team, the duplicate records were
poisoning were excluded in the registry. The PBSR team deleted and tagged to the record that was retained for
reviewed hospital admission & discharge data, medical analysis. Duplicates occurred due to multiple registra-
records from all departments (emergency, medicine, tion of same patient from different sources or health
neurology, radio diagnosis, physiotherapy, outpatient facilities or recurrent stroke in same patient.
and referral registers) of the PBSR hospital and other The data of 2018 and 2019 of the five PBSRs were
private and public hospitals. In addition, all Computed finalized following multiple iterative process of verifi-
Tomography (CT) and Magnetic Resonance Imaging cation of data quality and analysed to measure the
(MRI) imaging reports, death register, death certificates following: crude and age standardized rate by age, sex,
were scrutinized for identification of stroke patients. and residence (urban and rural), rate ratios of age
Details regarding date of onset of stroke, clinical find- standardized incidence (ASR), case fatality proportions,
ings at onset, imaging (CT/MRI) findings, risk factors, and rates. ASRs were calculated using the direct method
and diagnosis of type of stroke were collected on hos- by obtaining the age specific rates and applying these
pital admission or attendance of stroke patients at the rates to the standard population of that age group.7 The
SoRs by data abstraction on a core form. Data on vital world standard population was used to normalise the
status was collected at day 28 after onset of stroke. In- differences across populations with different age struc-
hospital death details were noted from medical record ture to calculate the age standardised rate (ASR) and
and medical certificates of cause of death. If details were expressed per million population. Standard error and
collected through telephone call or though house visits, 95% confidence limits of ASRs were calculated using
vital status was noted, and cause of death was recorded the Poisson approximation. Standardised Rate Ratio of
from death certificate or medical records shared by the ASR by sex group with 95% confidence limits was
family during the follow-up. determined. ASR for urban and rural populations with
Death certificates from the offices of the civil regis- standardised rate ratio, and ASRs by subtypes of stroke
tration system in the respective areas were verified for were calculated. Case fatality rate per 100,000 popula-
cause of death mentioned as ‘stroke’ or synonyms such tion (crude and age standardized) by age and sex group
as ‘cerebrovascular accident (CVA), brain hemorrhage, were calculated for all PBSRs. Cumulative risk and Rate
brain attack, cerebral infarction’ etc. These were ratio were calculated.8 Cumulative risk explains the
matched with registered incident cases to avoid dupli- probability/likelihood that an adult of age ≥18 years in a
cation. The matched records were reviewed and population will develop stroke irrespective of other
confirmed by the field team. Through a follow-back competing causes of death.
process of unmatched death records, any new incident
stroke cases identified in hospitals that had not been Role of the funding source
registered earlier, were included in the stroke registry. Indian Council of Medical Research has funded the
The remaining death certificates with stroke as under- establishment and implementation of the five
lying cause of death that were not reported by any population-based stroke registries and the coordinating
hospital or SoR in the PBSR area were counted as unit at ICMR-NCDIR. The sponsor of the study had no
‘Death certificate only (DCO)’cases. role in writing the manuscript, and in the decision to
Detailed inclusion and exclusion criteria, data submit the paper for publication.
collection tools, abstraction, verification, transmission
processes have been earlier described in the methodol- IEC approval
ogy paper of the National Stroke Registry Programme, The Population-based stroke registries have been
India.3 Data collected for years 2018–2019 had been approved by the respective Institutional Ethics Com-
transmitted to ICMR-NCDIR through an online soft- mittee of the institute implementing the PBSR; and the
ware portal (https://stroke.ncdirindia.org/).5 The data overall project of establishing the PBSRs is approved by
sets were processed for quality checks like missing the IEC of ICMR-NCDIR.
fields, date range, and consistency errors of one or many
variables. The comprehensive verification of data and
indicators of quality of data have been described in the Results
report of the PBSRs.6 These included verification of A total of 13,820 first-ever stroke cases were registered
residence, completeness of data, consistency between in the five PBSRs during 2018–2019. These included
subtypes, imaging, diagnosis, ICD 10 codes, and infor- 985 death certificate only cases (DCOs) from the CRS.
mation on vital status on day 28 after onset of stroke. The characteristics of the registered cases of first-ever
Duplicate records were identified using a ‘de duplica- stroke by age, sex, residence, risk factors, imaging, and
tion’ software application that listed probable duplicates subtype of stroke is described in Table 1. In three PBSRs
using predefined criteria of similar names, age, sex, with rural and urban populations, the proportion of
address, date of onset of stroke, date of diagnosis of rural registered cases was higher than urban residents.

4 www.thelancet.com Vol ▪ ▪, 2023

 Articles

Cuttack (n = 3226) Cachar (n = 2493) Tirunelveli (n = 3730) Kota (n = 2347) Varanasi (n = 2024)
mean [SD]/n (%) mean [SD]/n (%) mean [SD]/n (%) mean [SD]/n (%) mean [SD]/n (%)
Age in years Mean (SD) 64.0 [13.7] 59.5 [13.1] 62.3 [12.8] 60.5 [15.5] 62.1 [14.0]
Age group-Male Male Male Male Male Male
18–29 21 (1.1) 19 (1.2) 23 (1.0) 55 (3.8) 22 (1.8)
30–44 142 (7.5) 147 (9.1) 202 (9.1) 175 (12.0) 100 (8.0)
45–59 450 (23.6) 591 (36.8) 702 (31.7) 436 (29.8) 322 (25.6)
60–74 814 (42.8) 659 (41.0) 938 (42.3) 546 (37.3) 562 (44.7)
75+ 477 (25.1) 192 (11.9) 350 (15.8) 251 (17.2) 251 (20.0)
Age group-Female Female Female Female Female Female
18–29 21 (1.6) 16 (1.8) 13 (0.9) 25 (2.8) 19 (2.5)
30–44 88 (6.7) 95 (10.7) 99 (6.5) 95 (10.7) 64 (8.3)
45–59 338 (25.6) 305 (34.5) 353 (23.3) 201 (22.7) 204 (26.6)
60–74 569 (43.0) 327 (36.9) 738 (48.7) 357 (40.4) 319 (41.6)
75+ 306 (23.1) 142 (16.0) 312 (20.6) 206 (23.3) 161 (21.0)
Gender
Male 1904 (59.0) 1608 (64.5) 2215 (59.4) 1463 (62.3) 1257 (62.1)
Female 1322 (41.0) 885 (35.5) 1515 (40.6) 884 (37.7) 767 (37.9)
Place of residence
Urban 1381 (42.8) 369 (14.8) 1419 (38.0) 2347 (100.0) 2024 (100.0)
Rural 1845 (57.2) 2124 (85.2) 2311 (62.0) NA NA
Risk factors
Diabetes 844 (26.2) 289 (15.9) 985 (26.4) 559 (23.8) 707 (35.1)
Hypertension 2420 (75.0) 1239 (67.9) 1504 (40.3) 1377 (58.7) 1297 (64.1)
Current tobacco use 2009 (62.4) 970 (53.8) 861 (23.1) 756 (32.2) 385 (19.3)
Imaging studiesa
CT 2642 (81.9) 1810 (72.6) 2303 (61.7) 1326 (56.5) 1710 (84.5)
MRI 45 (1.4) 2 (0.1) 823 (22.1) 851 (36.3) 123 (6.1)
Both CT and MRI 39 (1.2) 1 (0.0) 112 (3.0) 160 (6.8) 14 (0.7)
Type of strokeb
Ischemic 2435 (75.5) 1160 (46.5) 3163 (84.8) 1962 (83.6) 1222 (60.4)
Haemorrhagic 665 (20.6) 667 (26.8) 430 (11.5) 377 (16.1) 710 (35.1)
Undetermined 126 (3.9) 666 (26.7) 137 (3.7) 8 (0.3) 92 (4.5)
Deaths
Deaths within 28 days of onset 540 (16.7) 1028 (41.2) 664 (17.8) 286 (12.2) 767 (37.9)

NA: Data not available. CT, Computed Tomography; MRI, Magnetic Resonance Imaging. aCases registered from ‘Death certificates only’ are excluded. bCases registered from
‘Death certificates only’ are included as ‘undetermined stroke’.

Table 1: Characteristics of registered cases of first-ever stroke in five Population-based Stroke Registries (PBSRs), India, 2018–19.

Imaging of brain was done in most stroke cases (72.7% stroke was one in 12 (both sexes), one in ten (males),
in Cachar in north-east India to 99.6% in Kota in west and one in 15 (females) (Fig. 2) from the pooled data of
India). Venous strokes were included in the subtype of PBSRs.
ischemic stroke (Table 1) as a small proportion (0.1% in Incidence rates by age group in rural and urban
Cuttack and Varanasi, 0.5% Tirunelveli, and 3% in Kota) residents are described in Table 3. The incidence rates
(numbers not shown separately). (crude and ASR for adults ≥18 years) was two times
The incidence rates in the five PBSRs and pooled higher in rural Cuttack (ASR 191.7), Tirunelveli (ASR
incidence rate by age and sex group, age standardised 163.3) and Cachar (ASR 93.9) as compared to the urban
rate (ASR), rate ratio of ASR by sex group have been residents of these three PBSRs. The risk of stroke
described in Table 2. The age specific incidence rate was among rural residents was one in seven (Cuttack), one
highest in 75+ years group. The ASR in females ranged in nine (Tirunelveli), and one in 15 in Cachar (Fig. 3).
from 59.5 (Cachar) to 109.4 (Cuttack), and in males Table 4 described incidence rates by stroke subtypes,
ranged from 102.8 (Cachar) to 144.6 (Cuttack). The age standardized incidence rates in males and females
pooled crude rate was 138.1 per 100,000 population with for all stroke subtypes, and rate ratio of ASR in each of
ASR of 103.4 (both sexes), 125.7 (males), and 80.8 (fe- the PBSRs. The ASR (≥18 years) for ischemic stroke
males). The cumulative risk of occurrence of first ever ranged from 39.6 in Cachar to 96.6 per 100,000 in

www.thelancet.com Vol ▪ ▪, 2023 5

 Articles

1396.9 (1271.5–1522.3) 966.9 (858.6–1075.3) 699.2 (600.3–798.1) 506.2 (423.0–589.5) 1044.9 (935.5–1154.4) 714.4 (635.2–793.7) 1254.4 (1099.2–1409.6) 885.2 (764.3–1006.1) 1037.7 (909.3–1166.1) 696.7 (589.1–804.4) 1091.9 (1037.0–1146.8) 752.6 (708.7–796.5) 916.1 (881.2–951.0)
Cuttack. Undetermined stroke was highest in Cachar

139.1 (131.8–146.4) 187.7 (181.8–193.6)

109.1 (106.2–112.0) 138.1 (135.8–140.4)

103.4 (100.7–106.1)
541.1 (527.2–555.0)
37.8 (35.7–39.9)
Both sexes

(20.6 per 100,000 population).

6.9 (6.0–7.8)
Multiple methods were used to follow up to ascertain
vital status on day 28 after onset of stroke. The most

–
common method of follow up was telephone call
430 (412.5–447.5)

80.8 (77.4–84.2)
27.9 (25.3–30.5)

(74.3%). Follow up was missing or unknown in 1.6% of

5.7 (4.5–6.9)
Female

registered cases (Supplementary Figure S1). In 38.6%

registered cases, vital status was ascertained between 28
and 30 days of onset of stroke and median duration to
166.1 (162.6–169.6)
376.6 (335.3–417.9) 651.5 (630.0–673.0)
233.5 (224.3–242.7)
Pooled PBSRs

125.7 (121.5–129.9)

complete follow up was 36 days.

1.56 (1.48–1.64)
47.6 (44.2–51.0)
8.1 (6.8–9.4)

Age standardized case fatality rates (ASCFR) per

Male

100,000 population for pooled PBSRs was 30.0 (males)

and 18.8 (females). The ASCFR ranged from 14.9 (Kota)
135.9 (117.2–154.5)

97.8 (90.9–104.7)
74.2 (65.9–82.4)

to 43.4 (Cachar) in males and 11.1 (Kota) to 28.8 (Vara-

25.5 (19.2–31.7)
6.9 (3.8–10.0)

nasi) in females. The rate ratio of ASCFR (males/fe-

Female

males) ranged from 1.2 in Cuttack to 2.0 in Cachar

(Table 5). Age specific case fatality rates were maximum
597.2 (547.8–646.5)
200.7 (178.8–222.6)

145.9 (137.9–154.0)
112.9 (103.1–122.7)

in age group of ≥75 years for all types of stroke in all

38.5 (30.9–46.0)

1.52 (1.32–1.75)
6.8 (4.0–9.7)

PBSRs (Tables 5 and 6). Proportion of case fatality was

Varanasi

maximum within first week after onset of stroke (53.5%)

Male

and 17% died on same day of onset of stroke in all PBSRs

418.8 (375.4–462.3)
116.8 (100.7–132.9)

(Supplementary Figure S2). Majority were ischemic

98.6 (92.1–105.1)
80.5 (72.1–88.9)
34 (27.1–40.8)
7.4 (4.5–10.4)

(56.7%) and haemorrhagic (39.4%) (Supplementary

Female

Figure S2). In Cachar, the type of stroke was in order

of haemorrhagic (57%), ischemic (38%), and undeter-
mined stroke (5%) (Supplementary Figure S3). The
460.9 (427.6–494.1) 636.2 (582.8–689.5)
135.8 (121.7–150.0) 230.8 (209.1–252.5)

134.9 (128.1–141.7) 150.8 (143.1–158.6)

130.5 (119.9–141.1)
58.9 (50.2–67.6)
14.6 (10.7–18.4)

1.62 (1.42–1.85)

ASCFR for ischemic stroke ranged from 5.2 per 100,000

Table 2: Incidence rate of first-ever stroke (per 100,000 population) by age group & sex in PBSRs, India, 2018–19.

population (Cachar) to 17.8 (Varanasi). The ASCFR for

Male
Kota

haemorrhagic stroke ranged from 4.7 (Tirunelveli) to

13.5 (Varanasi) (Table 6).
27.6 (22.2–33.0)

82 (75.5–88.4)
4.3 (2.0–6.7)

Discussion
Female

This study covered five geographical populations of one

to two million in India to provide measurements on
217.2 (199.7–234.7) 123.8 (109.9–137.7) 281.2 (260.4–302.0)
545.8 (504.1–587.4) 289.6 (258.2–321.0) 679.6 (636.1–723.1)

205.9 (197.3–214.5)
136.2 (127.3–145.0)

incidence and mortality of stroke by age, sex, and resi-

1.66 (1.50–1.84)
Tirunelveli

58 (50.0–66.0)
7.5 (4.4–10.6)

dence (urban and rural). The crude incidence rate of

first-ever stroke ranged from 96.6 to 187.6 per 100,000
Male

population and crude case fatality rate ranged from 15.3

to 46.6 per 100,000 population. Rate ratio of age
69.4 (64.8–73.9)
59.5 (53.3–65.7)
22.7 (18.1–27.2)
3.4 (1.7–5.1)

adjusted incidence of stroke was two times higher in

Female

rural as compared to the urban population in Cuttack,

Cachar, and Tirunelveli. The cumulative risk of devel-
Abbreviations: ASR, Age Standardised Rate; M, Male/F, Female.
123 (117.3–129.4)
103 (94.8–110.7)

oping stroke ranged from 1 in 7 in rural Cuttack to 1 in

34.7 (29.1–40.3)

1.73 (1.52–1.96)
4.1 (2.3–6.0)

15 in rural Cachar. Higher incidence rates in males was

Cachar
Male

observed as compared to females in age 30 and above

(rate ratio ranged from 1.32 to 1.73). This translated to a
189.1 (169.0–209.3)
604.1 (554.5–653.7)

156.6 (148.2–165.1)
109.4 (100.1–118.7)

cumulative risk of developing stroke of 1 in 10 among

32 (25.3–38.7)
7.9 (4.5–11.3)

males and 1 in 15 among females age ≥18 years in the

Incidence rate and 95% CI
Female

pooled population. The risk of stroke was higher in

males as compared to females in all registries (Table 2
224.9 (204.1–245.6)

144.6 (134.4–154.9)
802.4 (747.2–857.5)

and Fig. 1). The crude incidence rates of ischemic

217.4 (207.6–227.1)
50.8 (42.4–59.1)

Rate Ratio 1.32 (1.18–1.48)

stroke ranged from 45 (Cachar) to 143.9 per 100,000

8.1 (4.6–11.5)
Cuttack

(Tirunelveli) in age ≥18 years in both sexes. Subarach-

Male

noid haemorrhage (SAH) accounted for 1–2% of all

haemorrhages,6 and therefore incidence rates of intra-
group

of ASR
30–44

60–74
45–59
18–29
Age

(M/F)

cerebral haemorrhage (ICH) and SAH have been pre-

≥18
ASR
75+

sented as single group as ‘haemorrhagic stroke’. The

6 www.thelancet.com Vol ▪ ▪, 2023

 Articles

Fig. 2: Cumulative risk of stroke in males and females (≥18 years) in respective PBSR and pooled data of all PBSRs, India, 2018–2019.

Age group Cuttack Cachar Tirunelveli

Urban Rural Urban Rural Urban Rural
Incidence rate and 95% CI
18–29 4.9 (2.6–7.1) 13.0 (7.8–18.2) 0.0 (0.0–0.0) 4.7 (3.2–6.3) 3.8 (1.7–5.9) 8.8 (5.2–12.4)
30–44 21.3 (16.6–25.9) 74.9 (62.9–86.9) 10.8 (6.4–15.2) 33.9 (29.4–38.4) 24.6 (19.9–29.3) 67.9 (58.4–77.4)
45–59 139.3 (124.8–153.9) 312.4 (283.0–341.8) 76.0 (62.0–90.1) 204.1 (189.8–218.4) 120.4 (108.2–132.6) 332.1 (307.1–357.1)
60–74 546.4 (502.4–590.5) 1052.5 (979.2–1125.9) 224 (188.3–259.7) 485.4 (452.4–518.3) 349.6 (322–377.2) 849.7 (798.6–900.8)
75+ 973.1 (869.1–1077.2) 1701.6 (1543.9–1859.4) 494.7 (388.3–601.1) 625.7 (548.3–703.1) 644.2 (572.4–716.0) 1150.7 (1030.7–1270.7)
≥18 120.5 (114.1–126.8) 295.9 (282.4–309.4) 56.6 (50.8–62.3) 107.3 (102.7–111.9) 108.8 (103.1–114.5) 253.8 (243.5–264.1)
ASR 94.3 (86.5–102.2) 191.7 (177.9–205.5) 43.4 (36.4–50.4) 93.9 (87.5–100.2) 67.6 (62.1–73.1) 163.3 (152.9–173.7)
Rate Ratio of 2.0 (1.81–2.28) 2.2 (1.87–2.50) 2.4 (2.17–2.69)
ASR(R/U)
Abbreviations: ASR, Age Standardised Rate; R, Rural; U, Urban.

Table 3: Incidence rate of first-ever stroke (per 100,000 population) by age group & sex in rural and urban population in three PBSRs, India, 2018–19.

incidence rate of all haemorrhagic stroke was highest in with exclusion of the DCO cases (data not presented).
Varanasi (43.1). The pattern of incidence of Stroke incidence (ASR) was higher in males as
ischemic > haemorrhagic > undetermined stroke was compared to females for both ischemic and haemor-
similar in all registries, in rural and urban areas,6 and rhagic stroke (Table 4).

Fig. 3: Cumulative risk of stroke in both sexes (≥18 years) by Urban and Rural residence in three PBSRs, India, 2018–2019.

www.thelancet.com Vol ▪ ▪, 2023 7

 Articles

Abbreviations: ASR, Age Standardised Rate; I, Ischemic stroke; H, Haemorrhagic stroke (includes Intracerebral and Subarachnoid Haemorrhage); U, Undetermined stroke [includes registered stroke cases from Death Certificates with stroke as cause of
The incidence rates were similar to rates reported in

902.8 (830.2–975.4) 231.0 (194.3–267.7) 56.2 (38.1–74.3) 338.7 (290.3–387.1) 140.5 (109.3–171.7) 122.5 (93.4–151.6) 719.2 (659.4–779.0) 76.5 (57.0–96.0) 62.2 (44.6–79.8) 905.7 (816.0–995.4) 147.9 (111.7–184.1) 2.3 (0.0–6.8) 568.8 (500.8–636.8) 253.7 (208.3–299.1) 48.6 (28.7–68.5)
1.8 (0.0–3.8) 298.1 (272.8–323.4) 169.4 (150.3–188.5) 25.2 (17.8–32.6)
0.3 (0.0–0.8)

5.6 (4.5–6.7)
earlier population based studies in urban (140)9 and

5.0 (2.9–7.0)
4.3 (2.9–5.7)

3.6 (1.7–5.4)
1.2 (0.3–2.1)
5.2 (2.7–7.7)
rural Ludhiana (162.8 per 100,000),10 116.4 in urban and
U

119.4 per 100,000 in rural Trivandrum,11 urban Kolkata

(123.15),12 and urban Mumbai (145 per 100,000).13 The
69.9 (60.6–79.2)

43.1 (39.9–46.3)

23.6 (19.0–28.3)
42.1 (36.1–48.1)
33.2 (29.4–37.1)
11.2 (8.3–14.1)
ASR in the five PBSRs ranged from 81.9 to 127.7
2.2 (1.0–3.4)

annually which was lower compared to Asian countries

H

like Singapore (164.5),14 China (246.8)15 or European

0.6 (0.0–1.4) 94.3 (83.5–105.1)

0.3 (0.0–0.7) 56.8 (51.8–61.8)

countries like Sweden (165)16; and higher as compared
6.1 (3.8–8.5) 65.8 (58.4–73.3)
5.6 (3.4–7.8) 47.0 (40.4–53.5)
0.3 (0.0–0.8) 19.8 (15.9–23.7)

0.4 (0.1–0.7) 74.3 (70.1–78.5)

0.0 (0.0–0.0) 4.3 (2.6–6.0)
Varanasi

to countries like Japan (69.8)17 and Malaysia.18 Very few

studies have reported age standardized incidence of
I

stroke in rural residents, which has increased from

123.5 (rural Bengal),19 138 (rural Trivandrum)11 in the
early 2000s to 197.6 (rural Ludhiana)10 in 2018. The
U

latter rates were similar to incidence rates among rural

491.7 (466.5–516.9) 50.3 (42.2–58.4) 20.1 (15.0–25.2) 444.3 (412.7–475.9) 81.8 (68.2–95.4)

16.9 (14.2–19.6)
110.4 (100.6–120.2) 19.7 (15.6–23.8)
105.1 (100.4–109.8) 20.2 (18.2–22.2)
144.0 (131.6–156.4) 31.9 (26.1–37.7)

14.0 (10.5–17.5)

residents in Tirunelveli (163.3) and Cuttack (191.7).

7.5 (5.3–9.7)
2.1 (1.0–3.2)

Incidence rates were higher in rural areas as compared

H

to urban areas at each age group in the three registries

(Table 3). Stroke incidence, mortality and risk factors for
88.5 (82.4–94.7)

66.2 (58.7–73.8)
39.0 (33.9–44.1)
9.1 (6.9–11.3)

stroke were higher in rural areas than urban areas in

other studies in India,20,21 and in countries like China,15
Kota

and the USA.3 Increase in incidence with age was

I

observed in both sexes similar to the increasing age

115.3 (107.2–123.5) 16.3 (13.2–19.3) 4.6 (2.9–6.2)
12.4 (10.5–14.2) 3.9 (2.9–4.9)
0.1 (0.0–0.4)

3.3 (2.0–4.6)
0.2 (0.0–0.5)

143.9 (138.9–148.9) 19.6 (17.8–21.4) 6.2 (5.2–7.2)

169.6 (158.3–180.9) 32.2 (27.3–37.1) 5.3 (3.3–7.3)

specific incidence estimated by the GBD study in India.1

The GBD study also showed increase in stroke inci-
U

dence and DALYs rates (1990–2019) and variation of

8.8 (6.7–10.9)
1.0 (0.2–1.8)
7.2 (5.2–9.2)

DALYs rates of 5.5 times between the states.1 Stroke

Table 4: Incidence rate of first-ever stroke subtype (per 100,000 population) by age group in PBSRs, India, 2018–19.

incidence rates in the younger age group of 18–29 was

H

almost similar in males and females in Cuttack, Cachar

91.4 (86.4–96.3)

69.9 (63.9–75.8)
35.2 (30.8–39.6)
Tirunelveli

and Varanasi, a pattern that should be monitored in the

4.8 (3.1–6.5)

future. Earlier studies have shown incidence rates

higher in males than females9,11 except for few excep-
I

tions in Kolkata,12 and rural Ludhiana.10 The sex differ-

546.4 (513.6–579.2) 132.9 (116.7–149.1) 27.6 (20.2–35.0) 236.7 (217.0–256.4) 110.4 (96.9–123.9) 74.9 (63.8–86.0)
59.2 (52.6–65.8)

25.8 (23.8–27.8)
20.6 (18.1–23.0)
26.7 (22.8–30.7)
14.0 (11.0–16.9)
11.4 (9.1–13.7)
2.2 (1.3–3.1)

entials of incidence rates (men > women) in all age

groups may also have a gendered dimension of access to
U

care. A nationwide study (National Sample Survey

47.8 (41.8–53.8)

28.4 (24.2–32.6)
25.9 (23.9–27.9)

14.6 (11.5–17.6)
21.7 (19.1–24.3)

organisation (NSSO) 2017 & 2018, India) of services

9.1 (7.1–11.1)
0.6 (0.1–1.1)

availed for stroke revealed that women sought care in

public hospitals as compared to men who accessed
H

private hospitals. Men stayed longer in hospitals and

65.8 (58.8–72.8)

39.6 (36.0–43.2)
45.0 (42.4–47.6)

31.0 (26.4–35.6)
47.7 (42.2–53.2)

overall expenditure (medical and rehabilitation) for

8.2 (6.3–10.1)
1.0 (0.4–1.6)
Cachar

stroke management was higher among men as

compared to women.22There is a need for continuous
I

robust data on stroke burden based on residence, state,

4.9 (3.0–6.9)
0.4 (0.0–0.9)

7.3 (6.0–8.6)
4.9 (3.5–6.3)
4.9 (3.0–6.7)
2.2 (1.0–3.4)
5.5 (3.1–7.9)

and region wise, to plan for stroke care services that are
grossly deficient in India.23
U

In this study, hypertension was the most common

110.6 (101.6–119.6) 29.2 (24.6–33.8)
151.3 (138.9–163.7) 51.2 (44.0–58.4)

141.6 (136.0–147.2) 38.7 (35.8–41.6)

26.2 (23.1–29.4)

26.2 (23.1–29.4)
Incidence rate and 95% CI

risk factor reported across all registries ranging from

9.7 (7.1–12.3)
1.0 (0.2–1.8)

40.3% to 75% among stroke patients which was lesser

when compared to studies in south and north India
H

(83–89%).11,20 Similar observations were seen for dia-

96.6 (90.5–102.6)

81.4 (73.4–89.5)
29.6 (25.1–34.1)

betes and tobacco use. There may be some under-

6.7 (4.5–8.9)
Age group Cuttack

reporting of risk factors of stroke in our registries as it

was based on data abstraction from patient medical re-
I

cords. The National NCD risk factor monitoring survey

Females-ASR
Males-ASR

in India (2017–2018), revealed the national population

death].
30–44

60–74
45–59
18–29

≥18
ASR
75+

prevalence of raised blood pressure (28.5%) and glucose

8 www.thelancet.com Vol ▪ ▪, 2023

 Articles

(9.3%) among adults aged 18–69 years, with higher

223.7 (199.7–247.7)
prevalence in urban areas as compared to rural re-

(29.0–36.0)

(81.5–97.5)

(17.2–20.4)
(24.3–27.1)
spondents.24 The survey also captured the gaps in risk

(0.8–1.8)

(5.1–7.5)

89.5

18.8
32.5
factor awareness, its management, and control cascade.

25.7
6.3
1.3
F
Nearly 50% of those who were aware of their raised

Pooled PBSRS
blood pressure status, were on treatment and less than

(309.0–370.2)
(133.4–153.6)
50% had their BP under control.25 Similarly, less than

(38.0–41.4)
(50.2–59.0)

(28.0–32.1)

(1.51–1.68)
(8.3–11.3)
(1.2–2.4)
half (45.8%) who were aware of their raised blood

339.6
143.5
54.6

30.0
39.7
9.8
1.8

1.6
M
glucose levels, were on treatment and only one third had
their blood glucose under control.26 In India, poor
awareness and management of risk factors are signifi-

9.6 (5.8–13.4)

(239.6–383.6)
(105.6–154.2)
(44.6–68.6)

(34.0–42.6)

(23.7–33.9)
cant contributing factors for increasing incidence of

(1.2–5.4)

129.9

311.6
56.6

28.8
38.2
cardiovascular diseases like stroke.

3.3
Diagnosis of type of stroke in a registry was based on
increasing levels of confirmation using clinical, imaging

(198.9–260.1)

(366.0–535.2)
records, and death certificates. Availability of imaging
Varanasi

(35.9–47.8)
(49.5–73.9)

(49.3–59.1)

(1.33–1.58)
(8.0–16.6)
(1.4–5.4)
(CT or MRI) were high in all PBSRs (72%–99.6%), and

450.6
229.5

41.8
54.3
61.7
12.3
3.4

1.5
M

was comparable to earlier studies that reported 38%–

Table 5: Case fatality rate of first-ever stroke at day 28 after onset of stroke (per 100,000 population) by age group & sex in PBSRs, India, 2018–19.
95% of imaging available for stroke.2 Cerebral venous
sinus thrombosis (CVST) (≤3%) reported in the regis-

(151.6–269.6)
(37.4–68.2)

(11.1–15.9)

(8.0–14.2)
tries was comparable to urban Ludhiana.9 Nonspecific (5.2–14.6)
(0.9–4.9)
(0.0–1.4)

210.6
52.8

13.5

11.1
clinical presentation and poor sensitivity of the initial
9.9
0.6

2.9
F

non contrast CT to detect CVST may have resulted in

lower reporting of CVST, as imaging such as CT or MR
(219.7–370.1)

Venography is required for its diagnosis.27 (14.4–19.6)

(16.4–30.2)

(32.2–61.0)

(11.3–18.5)

(1.17–1.54)
(3.3–8.9)
(0.1–2.1)
Kota

The most common subtype of stroke reported in

294.9
46.6

14.8
23.3

17.0
6.1

1.3
M

1.1

earlier studies was ischemic stroke (65–84%) followed

by intracerebral haemorrhage (ICH 11–35%).2 Cachar
reported the highest proportion of undetermined stroke
(161.5–246.1)
(67.2–95.2)

(21.5–27.3)

(11.8–17.2)
(11.3–21.1)
(0.0–1.0)

(1.4–5.2)

(26.7%) due to cases registered from death certificates

203.8

24.3
16.2

81.2

14.5

with ‘stroke’ as a cause of death. Cachar is the largest

0.3

3.3
F

PBSR by area with predominantly rural population, and

lesser number of imaging centres as compared to other
Tirunelveli

(267.0–389.8)
(88.6–123.0)

(1.44–1.89)
(32.7–39.9)
(35.9–52.3)

PBSRs. High proportion of undetermined stroke was

(20.2–27.7)
(0.0–0.0)

(4.1–9.7)

328.4
105.8

reported in an analysis of the nationwide insurance

36.3
44.1

23.9
6.9
0.0

1.6
M

(Ayushman Bharat scheme) claims data, which was

attributed to non-availability of standard recording of
(131.9–231.7)

variables in stroke management in many states.28

(72.0–106.8)
(45.6–64.0)

(23.5–29.1)

(18.2–25.7)
(6.8–12.8)
(0.4–2.6)

Stroke mortality is a key indicator of quality of care of

Abbreviations: ASCFR, Age Standardised Case Fatality Rate; M, Male; F, Female.
181.8
89.4
54.8

26.2

22.0
9.8
1.5

stroke. A total of 2296 cases died within 28 days of onset

F

of stroke across all registries, and among them 70.6%

died within first week. Similar findings were reported in
(176.0–226.6)

(247.0–379.4)
(90.5–114.5)

(38.3–48.5)
(49.2–57.2)
(12.9–20.7)

(1.76–2.21)

Trivandrum registry with 72.1% cases dying within 10

(1.4–4.6)
Cachar

102.5

201.3

313.2
16.8

days of stroke onset.11 Case fatality ranged from 12.2%

43.3
53.2
3.0

2.0
M

to 41.2%,6 comparable to case fatality reported at 28–30

days in earlier studies between 2003 and 2013
Case fatality rate and 95% CI

(180.5–287.1)
(80.6–121.2)

(19–41%),9–13,29 signifying that stroke mortality has been

(19.3–34.5)

(15.1–22.8)
(24.1–31.3)
(0.0–1.8)

(2.7–8.3)

stagnant if not increasing over the last two decades in

100.9

233.8
26.9

18.8
27.8
0.8

5.5
F

India. However, case fatality was higher as compared to

Singapore (7.9%),14 England (14%),30 and Sweden
(11.2%).31 The case fatality rate seen in Kota may
(259.3–379.1)
(106.1–150.1)

(18.8–26.9)
(19.4–33.6)

(31.0–38.8)
Cuttack

(1.13–1.28)

represent a lower rate as cause of death data is not

(0.0–1.9)

(1.9–6.7)

319.2
128.1
26.5

22.8
35.0

properly recorded in the Civil Registration system. The

0.8

4.3

1.2
M

case fatality rates were lower as compared to the stroke

mortality rates estimated in the GBD 2021 (50.2),1 the
Rate Ratio
of ASCFR

million death study (71.5),32 and a rural community

group

ASCFR
60–74
30–44

45–59
18–29
Age

(M/F)
≥18
75+

based study in central India (121.6),21 where latter two

www.thelancet.com Vol ▪ ▪, 2023 9

 Articles

studies had used verbal autopsy to ascertain cause of

Abbreviations: ASCFR, Age Standardised Case Fatality Rate; I, Ischemic stroke; H, Haemorrhagic stroke (includes Intracerebral and Subarachnoid Haemorrhage); U, Undetermined stroke [includes registered stroke cases from Death Certificates with
155.0 (124.9–185.1) 77.5 (56.2–98.8) 44.1 (28.1–60.1) 73.9 (51.3–96.5) 50.4 (31.7–69.1) 122.5 (93.4–151.6) 156.8 (128.9–184.7) 40.2 (26.1–54.3) 62.2 (44.6–79.8) 184.8 (144.3–225.3) 64.7 (40.7–88.7) 0.0 (0.0–0.0) 188.2 (149.1–227.3) 145.9 (111.5–180.3) 48.6 (28.7–68.5)
death. The major limitation in calculating cause specific

25.2 (17.8–32.6)

4.6 (2.6–6.5)
5.5 (4.4–6.6)
0.3 (0.0–0.8)

4.2 (2.9–5.6)
4.8 (2.4–7.2)

3.6 (1.7–5.4)
1.2 (0.3–2.1)
mortality rates is the poor cause of death data in the
medical records and low coverage (22.5%) of Medical
U

Certification of Cause of death (MCCD) in India.33

69.9 (57.6–82.2)
23.8 (18.4–29.2)

17.6 (15.6–19.6)

16.2 (12.4–19.9)
The registry design included cases admitted in hos-

13.5 (11.1–16.0)

10.6 (7.5–13.8)
0.5 (0.0–1.1)
3.5 (1.9–5.1)

pitals and outpatient cases (fatal and non-fatal) and data

from the CRS (Table 7). Major sources include public
H

(medical college hospitals, district and sub-district hos-

19.9 (13.2–26.6) 0.0 (0.0–0.0) 87.2 (73.5–100.9)
0.0 (0.0–0.0) 30.6 (24.4–36.8)

0.0 (0.0–0.0) 20.8 (16.6–25.0)

0.0 (0.0–0.0) 14.6 (10.9–18.2)
0.0 (0.0–0.0) 23.5 (21.2–25.8)
0.0 (0.0–0.0) 17.8 15.0–20.6)
pitals, primary health centres) [37–80% cases], and pri-
0.0 (0.0–0.0) 6.3 (4.1–8.5)
0.0 (0.0–0.0) 2.5 (1.2–3.8)
Varanasi

vate hospitals (tertiary care hospitals, nursing homes,

clinics) [20–47% cases]. Data was collected from imag-
I

ing centres, physiotherapy centres, alternative healing

centres, and local death registration offices. A small
U

proportion of non-fatal non-hospitalized events (as

described in the STEPS 3 of the WHO Stepwise
7.5 (4.7–10.3)

5.7 (4.6–6.8)
4.8 (3.3–6.2)

4.3 (2.3–6.2)
0.6 (0.1–1.1)
2.4 (1.1–3.7)

5.1 (3.1–7.2)

approach) could have been missed out, as the registry

focused on cases that had some form of medical atten-
H

dance (imaging/medical doctor). In addition, patients

17.8 (13.0–22.6) 19.5 (14.5–24.5) 29.8 (21.6–38.0)

who had taken treatment in hospitals that were outside

9.4 (6.2–12.6)

9.6 (8.2–11.0)

9.7 (6.8–12.7)
8.2 (6.3–10.1)

6.8 (4.4–9.2)
0.3 (0.0–0.7)
2.1 (0.9–3.3)

the PBSR area may also have been missed. Details on all
Kota

risk factors of stroke (dyslipidemia, obesity, atrial

I

fibrillation, hormonal use etc.) may not be available in

4.6 (2.9–6.2)
3.9 (2.9–4.9)
0.1 (0.0–0.4)
0.2 (0.0–0.5)

Table 6: 28-days case fatality rate of first-ever stroke (per 100,000 population) by subtypes of stroke in PBSRs, India,2018–19.
3.3 (2.0–4.5)
6.2 (5.2–7.2)
5.5 (3.5–7.5)

all health facilities that provide data and thus

population-based attribution of risk factors to stroke is
U

not attempted in this study. Notwithstanding, the

12.0 (9.0–15)

6.3 (4.4–8.2)
0.0 (0.0–0.0)

7.6 (6.5–8.7)
4.7 (3.6–5.9)

3.3 (2.0–4.6)
3.0 (1.7–4.3)

methodology with standard definitions, process of data

abstraction and quality has adhered to the standard
H

criteria of a registry for stroke surveillance that is com-

parable.34 This has helped to provide reliable estimates
53.7 (45.4–62.0)

16.5 (14.8–18.2)

13.0 (10.3–15.8)
13.1 (10.0–16.2)

10.3 (8.6–12.0)
Tirunelveli

0.0 (0.0–0.0)

7.9 (5.9–9.9)
2.0 (1.0–3.0)

on incidence and mortality of stroke by demography,

residence, and subtypes in five geographical areas in
I

India that are comparable and can be monitored over

29.6 (22.0–37.2) 21.5 (15.0–28.0) 30.8 (23.7–37.9) 42.4 (34.1–50.7) 74.0 (63.0–85.0)
14.5 (11.2–17.8) 59.2 (52.6–65.8)

26.6 (22.7–30.6)
20.4 (18.0–22.9)

13.8 (10.9–16.7)
25.7 (23.7–27.7)

the years.
11.2 (8.9–13.5)
2.2 (1.3–3.1)

The evidence on stroke burden and mortality will be

useful to develop and monitor interventions in the five
U

populations. It will guide the hospitals treating stroke

10.5 (8.0–13.1)
8.5 (7.4–9.6)
7.4 (5.8–8.9)

4.0 (2.4–5.6)
0.1 (0.0–0.3)
1.9 (1.0–2.8)

patients to strengthen the diagnostic, curative and

follow-up services for stroke management, reduction of
H

disability, rehabilitation, and prevention of recurrent

5.7 (4.8–6.6)
0.0 (0.0–0.0)

6.2 (4.2–8.2)
6.2 (4.1–8.3)
0.2 (0.0–0.5)

stroke. Initiatives of the National Programme for pre-

5.2 (3.9–6.5)

4.2 (2.5–5.9)
Cachar

vention and control of non-communicable diseases (NP-

NCD) like the population-based risk factor screening
I

and management of hypertension and diabetes through

0.4 (0.0–0.9)

4.0 (2.0–6.0)

3.8 (2.6–5.0)
5.6 (4.5–6.7)

3.6 (2.0–5.2)
1.6 (0.5–2.7)

3.9 (2.2–5.7)

a comprehensive primary health care system and

monitored through IT platform called National NCD
U
Case fatality rate and 95% CI

portal shall address risk factor management and control.

9.8 (6.7–12.9)

8.9 (7.5–10.3)
0.0 (0.0–0.0)

6.2 (4.1–8.3)
5.9 (4.5–7.4)

5.6 (3.5–7.8)
1.3 (0.4–2.2)

The health and wellness centres, and Prime Minister’s

health insurance scheme under the Ayushman Bharat
H

scheme have the potential to address the availability of

64.4 (53.2–75.6)

16.9 (15.0–18.8)

13.0 (10.0–16.1)
12.9 (9.3–16.5)

health services for all NCDs. Recent studies have

11.3 (9.2–13.3)

stroke as cause of death].

Females-ASCFR 9.3 (6.7–12.0)
0.4 (0.1–0.9)
2.0 (0.8–3.2)
Cuttack

demonstrated the feasibility of training frontline health

workers in rural areas for detection and reporting of
I

stroke,10 and in secondary prevention of risk factors of

Males-ASCFR

stroke.35
group

ASCFR
30–44

60–74
45–59
18–29
Age

Conclusion: The study has provided clear lessons to

≥18
75+

strengthen primary prevention of risk factors and

10 www.thelancet.com Vol ▪ ▪, 2023

 Articles

Cuttack Cachar Tirunelveli Kota Varanasi

Sources Registered Sources Registered Sources Registered Sources Registered Sources Registered
cases cases cases cases cases
n % n % n % n % n % n % n % n % n % n %
Public Hospitals 6 9.4 2097 65.0 13 65.0 1885 75.6 28 20.7 1511 40.5 2 8.3 882 37.6 2 4.3 749 37.0
Private Hospitals 39 60.9 987 30.6 5 25.0 590 23.7 79 58.5 1761 47.2 8 33.3 1099 46.8 33 70.2 946 46.7
Imaging Centre 8 12.5 66 2.0 2 10.0 18 0.7 10 7.4 263 7.1 13 54.2 345 14.7 10 21.3 261 12.9
Others (Civil Registration System, Rehabilitation and 11 17.2 76 2.4 0 0.0 0 0.0 18 13.3 195 5.2 1 4.2 21 0.9 2 4.3 68 3.4
physiotherapy centres, Alternate healing centres)
64 100.0 3226 100.0 20 100.0 2493 100.0 135 100.0 3730 100.0 24 100.0 2347 100.0 47 100.0 2024 100.0

Table 7: Sources of registration in all PBSRs and number of registered cases (proportions) (%).

strengthen stroke care services in these five populations. References

1 Singh G, Sharma M, Kumar GA, et al. The burden of neurological
Registries provide a comprehensive framework of stroke disorders across the states of India: the Global Burden of Disease
surveillance to assess the burden and risk factors, Study 1990–2019. Lancet Global Health. 2021;9(8):e1129–e1144.
monitor availability and accessibility of stroke care ser- 2 Jones SP, Baqai K, Clegg A, et al. Stroke in India: a systematic
review of the incidence, prevalence, and case fatality. Int J Stroke.
vices, measure outcomes by age, sex, residence, and 2021;17(2):132–140.
subtype in different parts of India. The rural vs urban 3 Howard G. Rural-urban differences in stroke risk. Prev Med.
risk of stroke shall guide health policy and programme 2021;152(Pt 2):106661.
4 Mathur P, Rangamani S, Kulothungan V, Huliyappa D, Bhalla BB,
to strengthen efforts for stroke prevention and control in Urs V. National stroke registry programme in India for surveillance
India. Strengthening the National Stroke registry pro- and research: design and methodology. Neuroepidemiology.
2020;54(6):454–461.
gramme shall be a useful investment to inform and 5 NCDIR | National stroke registry programme. Available from:
monitor stroke prevention and care. https://stroke.ncdirindia.org/.
6 ICMR-NCDIR. Stroke incidence and mortality: a report of the popu-
Contributors lation based stroke registries, India (National Stroke Registry Pro-
SR, DH, VK and PM contributed to the concept and design of the paper. gramme). Bengaluru,India: ICMR-NCDIR; 2021. Available from:
SR & PM developed the data analysis plan and SR, DH, VK, VU, RRK, https://ncdirindia.org/All_Reports/pbsrbook/default.aspx.
NS were involved in data acquisition, management, and statistical 7 Boyle P, Parkin DM. Statistical methods for registries- Ch. In:
analysis. SR & DH conducted the literature search and prepared the Jensen OM, Parkin DM, Maclennan R, eds. Cancer registration –
principles and methods. Lyon: IARC Scientific Publications; 1991:.
manuscript. SR, DH, VK, VU, RRK, NS and PM were involved in Muir CS, skeet RG; 95.
manuscript review, editing and approved the final version of the 8 Day NE. Cumulative rate and cumulative risk. In: Cancer incidence
manuscript. SR, DH, VK, VU, RRK, NS and PM were part of the central in five continents. Lyon: International Agency for Research on
coordinating unit of the population-based stroke registry. Cancer; 1992:862–864. IARC Scientific Publication; vol. VI.
SS, PKM, MR, CRP, EB, KS were involved in data acquisition, 9 Pandian JD, Singh G, Kaur P, et al. Incidence, short-term outcome,
manuscript review and editing. and spatial distribution of stroke patients in Ludhiana, India.
AKM, SDN, SKS, MB, were involved in data acquisition, manuscript Neurology. 2016;86(5):425–433.
10 Singh S, Kate M, Samuel C, et al. Rural stroke surveillance and
review and editing.
establishment of acute stroke care pathway using frontline health
BKN, AS, AKK, BD were involved in data acquisition, manuscript workers in rural northwest India: the Ludhiana experience. Neu-
review and editing. roepidemiology. 2021;55(4):297–305.
VS, DM, BB, DEM were involved in data acquisition, manuscript 11 Sridharan SE, Unnikrishnan JP, Sukumaran S, et al. Incidence,
review and editing. types, risk factors, and outcome of stroke in a developing country:
RNC and LPM were involved in data acquisition, manuscript review the Trivandrum Stroke Registry. Stroke. 2009;40(4):1212–1218.
and editing. 12 Das SK, Banerjee TK, Biswas A, et al. A prospective community
All authors have read and approved the final manuscript. based study of stroke in Kolkata, India. Stroke. 2007;38(3):906–910.
13 Dalal PM, Malik S, Bhattacharjee M, et al. Population- based stroke
survey in Mumbai, India: incidence and 28-day case fatality. Neu-
Data sharing statement
roepidemiology. 2008;31(4):254–261.
Aggregate data and summary tables of the individual registries is 14 National Registry of Diseases Office. Singapore stroke registry
available for access at the following url: https://ncdirindia.org/All_ annual report 2020. Available from: https://nrdo.gov.sg/publi
Reports/pbsrbook/default.aspx. The corresponding author to be con- cations/stroke. https://nrdo.gov.sg/docs/librariesprovider3/default-
tacted for any further requirements. document-library/ssr-web-report- 2020c544bb698cf04ad1aaaa7a147
2296132.pdf?sfvrsn=33b4f18a_0.
Declaration of interests 15 Wang YJ, Li ZX, Gu HQ, et al. China stroke statistics 2019: a report
The authors have no conflicts of interest to declare. from the national center for healthcare quality management in
neurological diseases, China national clinical Research center for
Acknowledgements neurological diseases, the Chinese stroke association, national
center for Chronic and non communicable disease control and
The authors acknowledge the contribution of all the sources of stroke
prevention, Chinese center for disease control and prevention and
cases of the five population-based stroke registries which contributed
institute for global neuroscience and stroke Collaborations. Stroke
data and the registry staff. Vasc Neurol. 2020;5(3):211–239.
16 Aked J, Delavaran H, Norrving B, Lindgren A. Temporal trends of
Appendix A. Supplementary data stroke epidemiology in southern Sweden: a population-based study
Supplementary data related to this article can be found at https://doi. on stroke incidence and early case- fatality. Neuroepidemiology.
org/10.1016/j.lansea.2023.100308. 2018;50(3–4):174–182.

www.thelancet.com Vol ▪ ▪, 2023 11

 Articles

17 Takashima N, Arima H, Kita Y, et al. Incidence, management and national NCD monitoring survey. Front Public Health. 2022;10:
short-term outcome of stroke in a general population of 1.4 748157.
million Japanese ― shiga stroke registry. Circ J. 2017;81(11):1636– 27 Leach JL, Fortuna RB, Jones BV, Gaskill-Shipley MF. Imaging of
1646. cerebral venous thrombosis: current techniques, spectrum of
18 Hwong WY, Ang SH, Bots ML, et al. Trends of stroke incidence findings, and diagnostic pitfalls. Radiographics. 2006;26(Suppl
and 28-day all-cause mortality after a stroke in Malaysia: a linkage of 1):S19–S41.
national data sources. Glob Heart. 2021;16(1):39. 28 ICMR-NCDIR. Patterns of stroke care in ABPM-JAY beneficiaries
19 Bhattacharya S, Saha SP, Basu A, Das SK. A 5 years prospective in empaneled hospitals of PM-JAY scheme. Bengaluru; Available
study of incidence, morbidity and mortality profile of stroke in a from: https://ncdirindia.org/All_Reports/Stroke/POSC_ABPM_
rural community of eastern India. J Indian Med Assoc. JAY.pdf.
2005;103(12):655–659. 29 Nagaraja D, Gururaj G, Girish N, et al. Feasibility study of stroke
20 Kaur P, Verma SJ, Singh G, et al. Stroke profile and outcome be- surveillance: data from Bangalore, India. Indian J Med Res.
tween urban and rural regions of Northwest India: data from 2009;130(4):396–403.
Ludhiana population based stroke registry. Eur Stroke J. 30 Bray BD, Paley L, Hoffman A, et al. Socioeconomic disparities in
2017;2(4):377–384. first stroke incidence, quality of care, and survival: a nationwide
21 Kalkonde YV, Deshmukh MD, Sahane V, et al. Stroke is the leading registry based cohort study of 44 million adults in England. Lancet
cause of death in rural gadchiroli, India: a prospective community- Public Health. 2018;3(4):e185–e193.
based study. Stroke. 2015;46(7):1764–1768. 31 Guéniat J, Brenière C, Graber M, et al. Increasing burden of stroke:
22 Vijayan B, Ramanathan M, Rangamani S, Joe W, Gopinathan S, the dijon stroke registry (1987–2012). Neuroepidemiology.
Mishra US. Treatment and rehabilitation of stroke patients in In- 2018;50(1– 2):47–56.
dia: a gendered analysis based on repeated cross sectional national 32 Ke C, Gupta R, Xavier D, et al. Divergent trends in ischaemic heart
sample surveys on health, 2014 and 2019. Health Care Women Int. disease and stroke mortality in India from 2000 to 2015: a na-
2021;0(0):1–18. tionally representative mortality study. Lancet Global Health.
23 Pandian JD, Sudhan P. Stroke epidemiology and stroke care ser- 2018;6(8):e914–e923.
vices in India. J Stroke. 2013;15(3):128–134. 33 Office of the Registrar General of India. Report of medical cause of
24 Mathur P, Kulothungan V, Leburu S, et al. National non- certification of death,2020. Vital statistics division, office of the
communicable disease monitoring survey (NNMS) in India: esti- registar general of India, Ministry of Home Affairs, India. Available
mating risk factor prevalence in adult population. PLoS One. from: https://censusindia.gov.in/nada/index.php/catalog/42681;
2021;16(3):e0246712. 2022.
25 Amarchand R, Kulothungan V, Krishnan A, Mathur P. Hyperten- 34 Sudlow CL, Warlow CP. Comparing stroke incidence worldwide:
sion treatment cascade in India: results from national non- what makes studies comparable? Stroke. 1996;27(3):550–558.
communicable disease monitoring survey. J Hum Hypertens. 35 Jain M, Pandian J, Samuel C, Singh S, Kamra D, Kate M. Multi-
2022;37:1–11. component short term training of ASHAs for stroke risk factor
26 Mathur P, Leburu S, Kulothungan V. Prevalence, awareness, management in rural India. J Neurosci Rural Pract. 2019;10(4):592–
treatment and control of diabetes in India from the Countrywide 598.

12 www.thelancet.com Vol ▪ ▪, 2023

View publication stats