1007

Globally, stroke is the third commonest cause of mortality1 and the fourth leading cause of disease burden2. In 2005, stroke accounted for nearly 5.7 million deaths

Feasibility study of stroke surveillance: Data from Bangalore, India

D. Nagaraja*, G. Gururaj**, N. Girish**, Samhita Panda*, A.K. Roy, G.R.K. Sarma† & R. Srinivasa*

Departments of *Neurology & **Epidemiology, National Institute of Mental Health & Neuro Sciences

†Department of Neurology, St. John’s Medical College & Hospital & *Department of Neurology M.S. Ramaiah Medical College & Hospital, Bangalore, India

Received June 10, 2008

Background & objectives: Stroke is a leading cause of deaths, and disabilities in India. Reliable and good quality data on epidemiological characteristics of stroke are essential to plan, implement and evaluate stroke prevention and control programmes. A feasibility study was undertaken in Bangalore to examine the possibility of establishing stroke surveillance and to develop methodology for a larger programme.

Methods: The study adapted WHO STEPs-STROKE methodology to collect data on hospitalized stroke events and fatal stroke events in the city of Bangalore. In STEP I, Information was collected from 1,174 stroke patients in three large hospitals and were followed till discharge and 28 days; outcome was measured as death or disability. Stroke cases fulfilling diagnostic criteria, evaluated by neurologists and CT/MRI confirmed cases were included. Brief information on major risk factors was collected from all stroke patients / family members and from medical records by trained research officers. In STEP II, death records of Bangalore Mahanagara Palike for 2004 (n=23,312) were analyzed to identify stroke related deaths.

Results: Using WHO-STEPs approach, a methodology was developed for stroke surveillance in a geographically defined population. By STEP 1 method - 7 per cent of medical and 45 per cent of neurological admissions were due to stroke with a fatality rate of 9 per cent at hospital discharge and 20 per cent at 28 days. With a mean age of 54.5 (± 17.0) yr and male preponderance, nearly half had one or more risk factors. Weakness or paresis (92%) was the commonest presentation and ischaemic stroke was most frequent (73.8%). One third of total stroke patients were dependent at both discharge and 28 day follow up. By STEP II method the proportional mortality rate for Bangalore city was observed to be 6 per cent and more than 50 per cent of total stroke deaths had occurred in 10 major hospitals.

Interpretation & conclusions: The present study has shown that stroke surveillance is possible and feasible. Institution based (hospitals and vital registry data) stroke surveillance supplemented with periodical population based information can provide comprehensive information on vital aspects of stroke like mortality, risk factors, disability and outcome. There is a need to develop stroke surveillance in a phased manner along with mechanisms to apply data for prevention and control programmes.

Key words Chronic disease - developing countries - National Health Programme - stroke - stroke surveillance - WHO STEPS-Stroke

and 16 million cases of first ever strokes worldwide2. More than three fourths of these occur in low and middle income countries (LMICs) of the world3. Countries

Indian J Med Res 130, October 2009, pp 396-403

396

with large populations (like India, China, Brazil and Russia) contribute for greater numbers of deaths and disability due to stroke4-7. In 2001, pooled estimates for India showed that the prevalence of stroke was 203 per 100,000 population above 20 years (resulting in about 1 million cases) and contributing for 1.2 per cent of the total deaths (approximately 102,000 deaths)8. The Indian Council of Medical Research (ICMR) estimates indicated that amongst the non communicable diseases (NCDs), stroke contributes for 41 per cent of deaths and 72 per cent of disability adjusted life years (DALYS)9. The ongoing socio-demographic and epidemiological transition in India has a key influence on the burden of stroke and the problem is likely to increase in the coming years. The socio-economic impact due to stroke on individuals, families and health care institutions is enormous. The economic loss due to heart disease, stroke and diabetes between the years 2006 to 2015 is estimated to be approximately 9 billion international dollar loss of national income in India5.

Comprehensive information on stroke mortality, morbidity, risk factors and disability are required for planning and implementing programmes in prevention, management and rehabilitation3,10. Previous studies on stroke have documented the burden of stroke across the country. Banerjee & Das reviewing the epidemiology of stroke in India observe that the age adjusted prevalence rates vary between 250 to 350 per 100,000 population in different studies11. The incidence study from eastern India has shown the adjusted annual incidence (per 100,000 persons) of stroke to be 124 in rural area12

and 145 in urban area13. As stroke shares many risk factors with other NCDs, studies on hypertension14 also point to the greater burden of stroke in India. Several studies have elucidated the characteristics of stroke (type, mortality rates, complication, risk factors, etc.) and brought out the need for preventive measures in India15,16. In addition, stroke registries in Hyderabad17, Mumbai7, Chennai18, and Bangalore19 have delineated the treatment and rehabilitation components for better management of stroke. However, these initiatives have differed in methodology, been center specific and undertaken on different populations of varying sizes, focused on individual components and have not been continuous in nature (mostly one time effort). Systematic, uniform and continuous data / evidence can lead to stroke prevention, management and rehabilitation programmes in India20. This type of surveillance data will also help in monitoring and evaluating the impact of ongoing and future initiatives in this area.

Stroke surveillance provides inputs for planning services, monitoring trends, assessing the burden of problem, identifying changing profile and pattern of risk factors, and to monitor and evaluate the impact of intervention21. WHO has recommended STEPS approach for stroke surveillance and has examined feasibility of implementation in LMICs, including India7,16,22,23 to assess the magnitude and profile of stroke, design appropriate interventions and monitor their effectiveness21.

The present study examines the feasibility of developing a stroke surveillance programme, develop methodology and examine various issues of a larger programme in the Indian context. The specific objectives were to develop a strategy for establishment of a population based stroke surveillance system, implement the same and delineate mechanisms for a feasible system in a wider geographic area. The study thus explored the possibility of collecting uniform data on burden and other salient characteristics of stroke like mortality, morbidity, socio-demographic correlates, risk factors, patterns and disability.

Material & Methods

WHO STEPS stroke surveillance aims at collecting information on stroke events by a uniform and standardized method in hospital and community settings. Information is collected on both fatal and non fatal events and also hospitalized and non hospitalized events in different phases or steps (referred to as Step 1, Step 2 and Step 3). Step 1 aims at collecting information on stroke patients admitted to hospitals, Step 2 identifies fatal stroke events in the community and step 3 estimates community based non fatal stroke events21. Pooling of data from all 3 steps can provide complete and total information on both public health and clinical aspects of stroke, thus, helping to develop integrated programmes. Experience of several centres is vital to develop a larger uniform approach for the country to develop uniformity in procedures and practices16, 23. Due to resource constraints, the present study attempted implementation of steps 1 and 2 only. The Institutional Ethical Committee at National Institute of Mental Health and Neurosciences (NIMHANS), Bangalore, approved the study protocol.

Stroke events in hospitals (STEP 1): Three major referral treatment centres for stroke were selected from amongst the hospitals in Bangalore: NIMHANS, an apex public sector institution for neurological services; St. Johns Medical College Hospital (SJMCH) and M.S.

NAGARAJA et al: BANGALORE STROKE SURVEILLANCE 397

availability of services round the clock, presence of medical and neurological services, good recording systems and willingness to participate, the 3 hospitals were chosen for the feasibility study.

Data collection was undertaken during the 6 month period of March 1 to August 31, 2005. The definitions as recommended in the STEPS stroke manual were followed (Appendix). All 3 hospitals followed uniform methods in data capture procedures. The data capture forms were translated, pre-tested and validated in the local vernacular language (Kannada) for field use prior to beginning of the study. The focus of information gathering was on socio-demographic characteristics, past history, risk factors, details pertaining to the acute stroke event, diagnosis and classification of stroke, management, complications, status (death or disability) at discharge and at 28 day follow up. The risk factors were as reported by patients or attendants and supplemented with data from medical records. Information on diagnosis, complications, management and outcome was extracted from medical records.

Trained research officers visited various divisions of the hospital (casualty and emergency, intensive care unit, male and female wards (both medical and neurological), stroke unit and medical records department for identifying stroke patients on a daily basis. The diagnosis of stroke was based on neurological evaluation supported by CT / MRI. Every identified patient was followed up daily till discharge or death. Among those discharged alive, 664 (61.9%) subjects (residents of Bangalore city) were followed up at 28 days post discharge to identify post hospital deaths and to ascertain disability status. The modified Rankin’s scale which measures independence and functional level of the individual was used for assessing disability21.

Stroke mortality in the community (STEP 2) : Death records for the period January 1, 2004 to December 31, 2004 maintained by Bangalore Mahanagara Palike (BMP) - Central administrative office of the city were analysed. The Central Statistical Office receives mortality information from all 100 ward health offices, including government hospitals and BMP referral maternity hospitals. Each death is registered only after obtaining information in Form 4 (institutional deaths) or form 4A (non institutional deaths). A data recording format was developed to extract available information from death certificates which included age, gender,

Appendix

The following definitions were used in the study as recommended in WHO STEPS stroke manual21

Stroke: A focal (or at times global) neurological impairment of sudden onset, and lasting more than 24 h (or leading to death) and of presumed vascular origin, excludes transient ischaemic attack (TIA) and haemorrhage and symptoms caused by trauma.

Definitive focal signs can be accepted as indicative of a stroke but must have lasted for more than 24 h and the signs must be of a presumed vascular origin and include: Unilateral or bilateral motor impairment and or sensory impairment, non-fluent speech, half sided impairment of visual fields, diplopia, conjugate deviation, acute onset apraxia or ataxia or perception of deficit,

Ischaemic stroke: Caused by sudden occlusion of arteries supplying the brain, either, due to a thrombus at the site of occlusion or formed in another part of the circulation. The diagnosis is based on clinical examination along with neuroimaging findings.

Intra-cerebral haemorrhage: Bleeding from one of the brain’s arteries into the brain tissue. Usually sudden in onset and the diagnosis is based on clinical presentation and confirmed by neuroimaging.

Sub-arachnoid haemorrhage: Characterized by arterial bleeding in the space between the two meninges; sudden onset of severe headache or unconsciousness or both. Diagnosis is established by neuroimaging methods.

Unspecified stroke: To be used if no diagnostic examination (like brain imaging ) was performed and diagnosis based on clinical examination alone.

Fatal stroke: Stroke resulting in death within 28 days after the onset.

Non fatal stroke: Patients who survive 28 days after the onset of stroke.

Current tobacco use: Patient currently using any form tobacco including those who have stopped using less than 3 months before the stroke event.

Current alcohol use: Patient currently using any type of alcohol including those who have stopped using less than 1 year before the stroke event (This definition was used in the Bangalore study but is not mentioned in the WHO guidelines).

Diabetes mellitus: Either diagnosis of diabetes pre-stroke or patient self report, or use of anti-diabetic medication. Includes diagnosis of diabetes during current hospital stay.

Hypercholesterlaemia: Reported pathological elevated plasma total or LDL cholesterol levels or use of lipid lowering medication.

Ramaiah Hospitals (MSRMH), two major tertiary care institutions attached to medical colleges. A preliminary exploration was carried out in identified hospitals and based on larger registration of stroke patients,

398 INDIAN J MED RES, OCTOBER 2009


immediate cause of death, antecedent cause of death, and other significant conditions associated with death. Stroke specific search words like weakness of limbs, paralysis, stroke, cerebrovascular accident, and hemiplegia was listed out after a pilot validation study of death records in one of the wards. Research officers trained in manual record search by visual sighting using key search words identified and recorded all probable deaths due to stroke21. At the second level, a trained physician again reviewed these records for cause of death and those due to stroke were included. In addition to a retrospective mortality record analysis of one full calendar year, a prospective analysis for the study reference period was also undertaken using similar method.

Step 3 focusing on nonfatal events in the community was not included in the present programme due to resource constraints. Data analysis was done using Epi-Info24 and SPSS25. Only major descriptive findings are provided in this communication to highlight the type and nature of information that will be available from a surveillance approach.

Proportions and rates were used to describe and compare the results.

Results

Stroke events in the hospitals: A total of 1,174 cases were registered from the three hospitals during the 6 month period. Stroke constituted 4 per cent (1.5 to 7.5%) of casualty registration, 5 per cent (1 to 18%) of inpatient registrations, 7 per cent of medical admissions and 45 per cent (22 to 60%) of neurology registrations with rates varying across institutions.

The mean age of stroke patients was 54.5 (± 17.0) yr, with two thirds (65.6%) being 50+ and 18 per cent below 40 yr (Fig.). A greater preponderance was seen among men (67%) with a male to female ratio of 2:1.

Higher proportion of women below 30 yr was due to cortical venous thrombosis (CVT). The observed difference between age and sex and occurrence of stroke was statistically significant (P<0.01). Nearly half of stroke subjects were without any education (N=535, 46%) and employed in non government or unorganized sectors (N=525, 44.8%), while more than three fourths were married (N=931, 79.3%)), with a near equal number of patients residing within (N=664, 56.6%) and outside (N=510, 43.4%) Bangalore.

One third (N=393, 33.5%) of subjects reported directly and remaining were referred either from government hospitals (28.6%) or private health care institutions (36%). Of the latter, approximately 69 per cent referrals came from private nursing homes, 11 per cent from general practitioners or a family practitioner. Amongst those residing in the city, 44 per cent reached a definitive hospital later than 48 h, while similar figure for those residing outside the city was 56 per cent from rural areas. The commonest presentation of stroke was weakness or paresis (92%) with the left and the right side of the body being involved in almost equal proportions (46.2 and 47.4%, respectively). Speech and/or language disturbances and disturbed consciousness was reported by 73.9 and 14.7 per cent of the patients, respectively (Table I). A CT scan was done for 1085 (92%) of the registered patients; amongst them 610 (52%) was within 24 h but was delayed (24 h to 7 days) in 468

Fig. Age-sex distribution of stroke patients (% hospitalised patients).

Table I. Neurological presentation of the stroke subjects and diagnosis (n=1174)

n %Source of referral: Direct or self referral 393 33.5 Government hospital 336 28.6 Nursing home 289 24.6 Private teaching hospital 72 6.1 General practitioner 47 4.0 Others 37 3.2Time interval since stroke onset and definitive care:

Less than 24 h 320 27.324 to 48 h 343 29.2> 48 h 511 43.5

Presenting symptoms:Weakness or paresis 1081 92.1

Left 499 46.2Right 513 47.4Both 69 6.4

Problems with speech 868 73.9Disturbed consciousness 173 14.7Diagnostic investigations: CT scan 1085 92.4 MRI scan 163 13.9 Carotid ultrasound 162 13.8

(40%) of the cases. The certainty of diagnosis of stroke was very high in 92 per cent of the cases. Ischaemic stroke was the most frequent (73.8%), followed by Intra-cerebral hemorrhage (13.7%), cortical venous thrombosis (8.7%) and subarachnoid haemorrhage (1.4%). Depending on the type of institution, majority of the cases were first admitted into the emergency rooms (63.5%) and subsequently managed by either neurology and / or medical unit.

With regard to risk factors, 13 per cent reported a previous history of stroke, 48 per cent were hypertensive, 23 per cent reported to be diabetic, 33 per cent used tobacco and 25 per cent consumed alcohol (Table II). Nearly 1 in 5 patients had both hypertension and diabetes (18.5%). A family history of stroke was reported by 7.3 per cent of patients. While 23 per cent did not have any risk factor, 29 per cent had one risk factor and 30.0 and 11.6 per cent, had 3 or more than 3 risk factors, respectively.

Among the hospital registered patients, 102 out of 1174 died during their hospital stay resulting in case fatality rate of 8.7 per cent. Of those patients discharged alive (1072), 664 patients could be followed up to 28 days and 81 (12.2%) patients had died during the reference period. Assuming the same fatality rate amongst those who could not be followed up (50 out of 408), the total number of deaths in the series was 233, with an overall case fatality rate of 19.8 per cent. Among the survivors (1072 patients), one-third (32.5%) were discharged home and remaining were referred to other hospitals (67.5%). At the time of discharge, one third of the patients (31.3%) were dependent on others [either being not able to walk - 25.7% or having other severe disabilities - 5.6%]. This overall proportion changed marginally (from 31.3 to 29.7%) during the 28 day follow up. Further, it was observed that nearly one out of two patients continued to have some disability even at 28 days after discharge (Slight disability-25.3% and moderate disability but able to walk-26.8%) (Table III).

Stroke mortality in the community: From a total of 23,312 available death records at Bangalore Mahanagara Palike, 1,690 (7.5%) deaths could be categorized as probably due to stroke. The physician review of the given cause(s) of death excluded non stroke deaths and resulted in 83 per cent of deaths classified as definitively due to stroke. The proportional mortality rate of stroke in the city of Bangalore for the year 2004 was thus calculated to be 6.0 per cent. The overall crude stroke mortality rate was 326 per million, with rates being higher in men (388 vs. 257 per million population). Nearly 97 per cent of stroke deaths occurred in health care institutions; it was observed that nearly 50 per cent of stroke deaths occurred in 10 major hospitals alone. Nearly 62 per cent of the stroke deaths were amongst men with a male: female ratio of 2:1. The age specific mortality rate revealed that stroke deaths increased from 30 yr and reached a peak around 65 yr (45%) with 20 per cent in less than 45 yr.

Continuing to explore the feasibility of a stroke surveillance programme, a total of 2049 deaths records were analyzed for the period March 1, 2005 to August 31, 2005 in the three health ranges attached to 3 study hospitals. Nearly 85 per cent (76 to 95%) of the stroke deaths reported from the three ranges were from the three study hospitals and 7.9 per cent (6 to 11%) of the deaths in these 3 ranges were due to stroke.

Discussion

Previous studies on stroke11,12, earlier neuroepidemiological surveys11,26 and stroke registries7,17-19 have identified the burden and select characteristics of stroke but there have been several gaps in continuous and comprehensive information needed for planning and implementing integrated programmes. Most studies have been one time efforts and not continuous, single hospital based, while the methodology and focus of information gathering has varied. Hospital registries have estimated risk

Table II. Major risk factors for stroke*

Risk factors n=1174 %Previous history of stroke 154 13.1Hypertension 563 48.0Tobacco use 383 32.6Alcohol use 295 25.1Diabetes mellitus 271 23.1Atrial fibrillation 114 9.7Family history of stroke 86 7.3*As reported by patients and/or ascertained from medical records

Table III. Disability status of the hospitalized stroke subjects (Modified Rankin Scale, %) At discharge

(n=1072)At 28 day follow up(n=583)

No symptoms (0) 1.4Symptoms present but not related to stroke and no disability (1)

24.4 16.8

Slight disability (2) 6.5 25.3Moderate disability able to walk (3) 37.8 26.8Moderate disability not able to walk (4) 25.7 26.8Severe disability (5) 5.6 2.9


factors and selective fatality rates but failed to provide information on mortality for the larger population. The present study has shown that it was possible to collect minimal, good quality, vital data in different hospital settings and also from the city vital registry division. Information on characteristics of stroke, risk factors, disability and outcome and mortality rates were gathered by a uniform methodology. Supplementing this information with periodical data from non fatal events in community can help in comprehensive and total understanding of stroke events in defined populations.

The present study registered a larger number of CT confirmed stroke cases in the limited period of 6 months. The observation that nearly 6 - 8 per cent of total deaths in the city and 20 per cent of stroke deaths occurring during or soon after discharge (case fatality rate of 20%) indicate that stroke is a major public health problem. Nearly 5 to 10 per cent of the hospital patients and half of neurology patients were due to stroke. Majority of stroke deaths and hospitalizations occurred amongst the 50 to 65 yr old population and were predominantly among men. The 28 day follow up showed that 55 per cent of the stroke survivors had moderate to severe disabilities. The methodology and mechanism evolved by the current study (integrating with existing systems - vital registration system and hospitals), collecting data by use of existing and available records, ensuring completeness of data and simplicity of the study instrument can serve as guidelines for initiating stroke surveillance in India. The need for strengthening mortality information needs no overemphasis and is critical to identify fatal stroke events in the community.

At the city level, 97 per cent of the stroke deaths occurred in institutions, with 50 per cent of these from 10 major hospitals alone. Thus, identification and inclusion of select large health care institutions will ensure wider coverage along with inclusion of large number of patients in a given and defined geographic area. It is desirable to begin with lesser number of institutions, establish quality control mechanisms and expand in phases. These centres can serve as sentinel centres for stroke surveillance. In India, private sector provides care for nearly 70 per cent of health care needs27, and in the present study nearly 50 per cent of deaths occurred in smaller hospitals or nursing homes and all these were in the private sector. Thus, it is important to involve these in the overall programme for stroke management and prevention.

Nearly three fourths (77%) of the study group had one or more known risk factors for stroke with one in five patients being both hypertensive and diabetic. Mendis et al28 have reported that in middle and low income countries one in two patients has at least two or more risk factors for either ischaemic heart disease or stroke and a significant number of patients do not receive appropriate medication. Evidence from high income countries reveal that modifying risk factors remains a major strategy for decreasing the stroke burden2,5,,29. As many risk factors are common between stroke and other NCDs, risk factor prevention programmes would contribute to a greater extent in reducing the burden of stroke and other NCDs1,2,30,31. The observations from the current study critically reflects on need for building integrated programmes and surveillance will be helpful in strengthening these activities. NCD risk factor surveillance is already under progress and it would be helpful to include stroke surveillance in these centers.

However, certain challenges need to be overcome for establishing stroke surveillance and from the early stages itself. Methods to include cases based on physician / neurologist evaluation (neurologists and CT facilities may not be available everywhere) and inclusion of private hospitals and diagnostic centers need to be evolved in a systematic way. With availability of CT facilities in urban areas and in districts, case identification and diagnostic ascertainment would become easier. Identifying units in hospitals where stroke patients are likely to be included (emergency room, outpatients, inpatients, ICU) and utilizing different types of hospital records ensured that all cases were included in the present study, and this needs to be developed in all centers included for surveillance in future.

Step 3 which aims at inclusion of non fatal community stroke events was not included in present feasibility study. However, some of the centres have initiated this and their experience has shown that it is possible23.

Stroke surveillance needs to be planned and implemented in a phased manner, beginning with medical college hospitals, select urban hospitals and district hospitals in rural areas. Stroke surveillance can be initiated in centers already identified for NCD risk factor surveillance along with appropriate use of information technology through web based approaches. This integrated approach is feasible and better suited to the Indian context to develop programmes in India and other LMICs in the coming


years3. The success and sustainability of stroke surveillance programme depends on simplicity of data collection process, utilization and feedback to all partners for action33,32.

In conclusion, despite limitations of shorter duration along with resource constraints, the present study has shown that establishing population based stroke surveillance for fatal and non fatal events is both possible and feasible and needs to develop specific mechanisms to overcome challenges in India. With the recognition of stroke as a major public health problem, there is a need for long term, continuous, and good quality information to monitor trends, identify risk factors, develop - implement and evaluate interventions.

Acknowledgment Authors acknowledge the financial grant received from World Health Organization, South East Asia Regional Office, New Delhi (ICP GPE 002) and Indian Council of Medial Research, New Delhi. The authors thank Drs Ruth Bonita and Thomas Truelson from WHO, Geneva, for technical guidance, encouragement and support in undertaking this study and the entire field staff involved in data collection (Dr Kavita R, Junior Scientific Officer in Epidemiology, Dr Pandian D, Assistant Professsor of Psychiatric Social Work, Lakshminarayan KL, Narayanaswamy and Dr Thandavamurthy PS from Bangalore Mahanagara Palike and the field team of Dr Lavanya G; Shriyut Basavaraju KS, Manjunath DP, Shanmukha CG, Srinivasamurthy KG , Suresh C, Gangaraju, Saniwaram Reddy and Ananda G).

References1. Warlow C, Sudlow C, Dennis M, Warlow C, Sudlow C, Dennis

M, et al. Stroke. Lancet 2003; 362 : 1211-24.2. Strong K, Mathers C, Bonita R. Preventing stroke: saves lives

around the world. Lancet Neurol 2007; 6 : 182-7.3. Feigin VL. Stroke in developing countries: can the epidemic

be stopped and outcomes improved. Lancet Neurol 2007; 6 : 94-7.

4. Feigin VL. Stroke epidemiology in the developing world. Lancet 2005; 365 : 2160-1.

5. World Health Organization. The economic impact of chronic diseases in preventing chronic diseases: a vital investment. World Health Organization, Geneva; 2005.

6. Liu M, Wu B, Wang W, Lee L, Zhang S, Kong L. Stroke in China: Epidemiology, prevention and management strategies. Lancet Neurol 2007; 8 : 456-64.

7. Dalal P, Bhattacharjee M, Vairale J, Bhat P. UN Millennium development goals: Can we halt the stroke epidemic in India. Ann Indian Acad Neurol 2007; 10 : 130-6.

8. Anand K, Chowdury D, Singh KB, Pandav CS, Kapoor SK. Estimation of mortality and morbidity due to strokes in India. Neuroepidemiology 2001; 20 : 208-11.

9. Indian Council for Medical Research. Stroke. In: Assessment of the burden of non communicable diseases: Final project

report. New Delhi: Indian Council of Medical Research; 2004. p. 18-22.

10. Indrayan A. Forecasting vascular disease cases and associated mortality in India. In; Burden of disease in India, Vol 2, NMCH background papers, National Commission on Macroeconomics nd Health, Ministry of Health and Family Welfare, Government of India; 2005. p. 197-215.

11. Baneerjee TK, Das SK. Epidemiology of strokes in India. Neurol Asia 2006; 11 : 1-4.

12. Bhattacharya S, Saha SP, Basu A, Das SK. A 5 years prospective study of incidence, morbidity and mortality profile of stroke in a rural community of eastern India. J Indian Med Assoc 2005; 103 : 655-9.

13. Das SK, Banerjee TK, Biswas A, Roy T, Raut DK, Mukherjee CS, et al. A prospective community-based study of stroke in Kolkata, India. Stroke 2007; 38 : 906-10.

14. Gupta R. Trends in hypertension epidemiology in India. J Hum Hypertens 2004; 18 : 73-8.

15. Banerjee TK, Roy MK, Bhoi KK. Is stroke increasing in India – preventive measures that need to be implemented? J Indian Med Assoc 2005; 103 : 162-4.

16. Dalal PM. Burden of stroke: Indian perspective. Int J Stroke 2006; 1 : 164-6.

17. Kaul S, Venkataswmy P, Meena AK, Sahay R, Murthy JMK. Frequency, clinical features and risk factors of lacunar infarction (Data from a Stroke registry from South India). Neurol India 2000; 48 : 116-9.

18. Arjundas D, Pandiyan U, Arjundas G, Baux H. Surveillance of stroke: WHO STEP-wise approach: A Chennai stroke Unit report. Ann Indian Acad Neurol 2007; 10 : 154-9.

19. Nagaraja D, Gururaj G, Jayakumar R, Subbakrishna DK. Summary report of the project “Stroke Registry in NIMHANS, NIMHANS/NEU/Reports/Stroke/2000 (unpublished report).

20. Dalal PM, Bhattacharjee M. Stroke epidemic in India: Hypertension – stroke control programme is urgently needed J Assoc Physicians India 2007; 55 : 689-91.

21. World Health Organization: The WHO STEPs Stroke Manual. The WHO STEPwise approach to stroke surveillance Geneva: World Health Organization; 2006.

22. Truelsen T, Heuschmann PU, Bonita R, Arjundas G, Dalal P, Damasceno A, et al. Standard method for developing stroke registers in low-income and middle-income countries: experiences rom a feasibility study of a stepwise approach to stroke surveillance (STEPS Stroke). Lancet Neurol 2007; 6 : 134-9.

23. Dalal PM, Bhattacharjee M, Vairale J, Bhat P. Mumbai stroke registry (2005-2006) – Surveillance using WHO steps stroke Instrument – challenges and opportunities. J Assoc Physicians India 2008; 56 : 675-80.

24. Epi Info: Database and statistics software for public health professionals. Version 3.3.2 February 9 2005, Centre for Disease Control. Available from: http://www.cdc.gov/epiinfo/, accessed on March 4, 2005.

25. SPSS (Statistical Package for Social Sciences) for Windows, Release 11.0.0; 2001.

26. Gourie-Devi M, Gururaj G, Satishchandra P, Subbakrishna DK. Prevalence of neurological disorders in Bangalore, India:


Reprint requests: Dr D. Nagaraja, Professor of Neurology & Director & Vice Chancellor, National Institute of Mental Health & Neuro Sciences, Bangalore 560 029, India e-mail: [email protected]

A community based study with an urban-rural comparison. Neuroepidemiology 2004; 23 : 261-8.

27. Government of India. National Health Accounts India 2001 – 2002, National Health Accounts Cell, Ministry of Health and Family Welfare, Government of India, New Delhi; 2005.

28. Mendis S, Abegunde D, Yusuf S, Ebrahim S, Shaper G, Ghannem H, et al. WHO study on prevention of recurrences of myocardial infarction and stroke (WHO-PREMISE). Bull World Health Organ 2005; 83 : 820-8.

29. Rothwell PM, Coull AJ, Giles MF, Howard S, Silver L, Bull L, et al. Change in stroke incidence, mortality, case-fatality, severity, and risk factors in Oxfordshire, UK from

1981 to 2004 (Oxford Vascular Study). Lancet 2004; 363 : 1925-33.

30. Goyal A, Yusuf S. The burden of cardiovascular disease in the Indian subcontinent. Indian J Med Res 2006; 124 : 235-44.

31. Gaziano TA. Reducing the growing burden of cardiovascular disease in the developing world. Health Aff (Millwood) 2007; 26 : 13-24.

32. Truelson T, Bonita R, Jamrozik K. Surveillance of stroke: a global perspective Int J Epidemiol 2001; 30 : s11-6.

33. World Bank, 2002: Public Health Surveillance tool kit. Available from: http://siteresources.worldbank.org/INTPH/Resources/376086-1133371165476/PHSurveillanceToolkit.pdf, accessed on December 9, 2007.


Documents

1007