Effects of Urban Community Intervention on 3-Year Survival and Recurrence After First-Ever Stroke
Background and Purpose— For the past 2 decades, stroke has been a principal cause of death in China, and stroke incidence tends to increase with the increase of stroke-related risk factors. The purpose of this study was to evaluate the effects of urban community-based intervention on 3-year survival and recurrence after first-ever stroke.
Methods— Two communities with a registered population of ≈50 000 each were selected as either intervention or control communities in Beijing during 1991 to 2000. Comprehensive intervention measures including the management of high-risk population and the health education of whole community population were regularly implemented. Then the influence of community intervention on 3-year survival and recurrence after initial stroke was evaluated.
Results— Within 3 years, 41.85% of 736 patients in the intervention community died whereas 40.34% of 818 patients in the control community died. Of 223 cases from the intervention community, 26 (11.66%) had a recurrent stroke within 3 years versus 52 (20.80%) of 250 cases from the control community. The statistical difference was found. Compared with the control community, the death risk of first-ever stroke in the intervention community decreased by 26% (relative risk [RR]=0.74; 95% confidence interval [CI]: 0.61 to 0.89; P=0.002); especially, that of hemorrhagic stroke decreased by 39% (RR=0.61; 95%CI: 0.46 to 0.81; P=0.001). Compared with the control community, the recurrence risk of first-ever stroke from the intervention community decreased by 42% (RR=0.58; 95% CI: 0.34 to 1.00; P=0.048).
Conclusion— Community intervention may be effective and beneficial to the recurrence prevention and survival improvement of stroke, especially hemorrhagic stroke.
Stroke is a chronic noncommunicable epidemic with high rates of incidence, mortality, and disability. For the past 2 decades, stroke has been a principal cause of death in China. The WHO MONICA Project showed that China has one of the higher incidence and mortality rates of stroke in the world.1 The national surveys of hypertension and diabetes showed that the prevalence of hypertension increased from 7.7% in 1979 to 11.4% in 1991, and the prevalence of diabetes increased from 1% in 1980 to 3.2% in 1996.2,3 Stroke incidence tends to increase with the increase of stroke-related risk factors. The Beijing Neurosurgical Institute firstly initiated the Chinese Seven-City Project of Stroke Prevention during 1987 to 1990.4 Thereafter, the Comprehensive Community-based Intervention Trial of Stroke was expanded to larger urban population in Beijing, Shanghai, and Changsha, 3 cites with higher incidence of stroke during 1991 to 2000.5,6 The new concept of prevention and control toward chronic noncommunicable diseases emphasizes primary and secondary prevention, and the secondary prevention of stroke recurrence may achieve twice the results for half the efforts of the primary prevention. The beneficial effects of secondary preventive measures after stroke, such as antihypertensive, antiplatelet, or anticoagulant therapy and surgical treatment of severe symptomatic carotid artery stenosis, have been well established in clinical trials.7 However, little evidence of the secondary prevention for health care medicine is available. The purpose of this study was to evaluate the effects of urban community-based intervention on 3-year survival and recurrence after first-ever stroke.
Materials and Methods
Two geographically separated communities, each with a population of ≈50 000 and matched for age and gender, were selected as either intervention or control communities in Beijing during 1991 to 2000. In the intervention community, 738 cases with first-ever stroke were identified during 1991 to 1997. Meanwhile, 820 cases were reported in the control community. Two cases in each community were excluded because of no report of follow-up outcome within 3 years. The characteristics of subjects of the 2 communities are summarized in Table 1, which lists the subjects used for survival observation during 1991 to 1997 and recurrent follow-up study during 1996 to 1997.
The comprehensive intervention measures of stroke included both the management of stroke-related high-risk population (hypertension) and the health education for whole community population. The community residents older than 35 years (inclusive) were encouraged to participate in the community-based stroke prevention in the course of high blood pressure (HBP) census. In each resident committee, HBP counseling was provided, and the identified patients with HBP received regular blood pressure (BP) measurement and were advised to take appropriate preventive and therapeutic measures. The mean management rates of HBP patients trimonthly followed-up were 77.1% in phase I (1991 to 1995) and 85.1% in phase II (1996 to 2000). The BP of 37.8% of the HBP patients in phase I and 45.8% in phase II were <160/95 mm Hg for at least 6 months every year. The survival patients with stroke were also regularly managed and followed-up along with the HBP patients every year during 1991 to 2000. The systematic management of diabetes patients and smokers was not initiated until phase II.
To advocate healthy lifestyles among the urban residents, eg, limiting salt intake, exercising more often, quitting smoking, and refraining from alcoholic drinks, the health education and promotion programs were regularly made available to the residents living in the intervention community. Flyers and booklets on health knowledge of stroke were distributed to every household in the intervention community every 3 months during the periods of 1991 to 2000. Bulletins for the same purpose were also issued every 3 months. Three or four lectures were provided to the directors of local residential committees, volunteer residents, and patients with HBP every year. Audiovisual facilities and materials were also used. The aforementioned materials explain the role of hypertension, coronary heart diseases, and diabetes in stroke risk and describe the related risk factors of these diseases. The local health, education, and administration authorities also participated in the work described. According to a sampling survey, >90% of residents were influenced by the health education and promotion implemented in the community.
Disease Surveillance and Follow-up
An active surveillance system was established to identify and ascertain any patient with an incident or recurrent stroke. The details of this system were described in earlier works of similar surveillance.4,8 The standards of diseases surveillance were consistently applied in both intervention and control communities. The information collected were incidence, mortality, and recurrence of stroke. During 1991 to 1997, all patients of first-ever stroke in both the intervention and control communities were annually monitored and followed-up for death cause information within 3 years; during 1996 to 1997, the patients of first-ever stroke from the 2 communities were also monitored for stroke recurrence information within 3 years. All possible stroke cases were first reported to medical workers in local health care station who in turn notified us. The diagnosis was verified by neurologists or neurosurgeons according to medical history, neurological examination, and computed tomography/magnetic resonance imaging (CT/MRI) diagnosis. In addition, the official records of death certificates were reviewed annually.
For each case, the starting point of follow-up was the incident date of first-ever stroke, and the endpoint events were the all-cause death and the first recurrence after first-ever stroke within 3 years, which were reported annually. Patients who did not die or have recurrent stroke were censored at either the date of their last follow-up interview or their date of death. The low rate of out-migration after stroke (1.41%) and the unique system of household registration and cancellation in Chinese urban community made it possible that these endpoint events of the patients after initial stroke be readily recorded. For the stroke patients migrated out of the 2 communities, we obtained the outcome information by other methods such as telephone and mail.
Diagnosis Criteria for Stroke and Its Recurrence
Stroke was defined as rapidly developing clinical signs of local or global loss of cerebral function lasting for >24 hours (or leading to death before then) with no apparent cause other than cerebral ischemia or hemorrhage. By definition, cases of transient ischemic attacks (TIA) were excluded. The recurrent stroke was defined as a new neurological deficit that fit the definition of stroke, and the recurrent event after the initial stroke was stipulated that stroke should occur at least 4 weeks after the preceding event. The rate of CT/MRI diagnosis in this study was 84.68%.
Definition of Stroke-Related Diseases
The risk factors were defined as follows: hypertension (reported systolic blood pressure ≥160 mm Hg, reported diastolic blood pressure ≥95 mm Hg, patient’s self-report of hypertension or use of antihypertension drugs), TIA (a temporary disturbance in brain function resulting from a temporary blockage of the brain’s blood supply within 24 hours), cardiac diseases (history of myocardial infarction, coronary artery disease, congestive heart failure, arrhythmia, or valvular heart disease), diabetes mellitus (fasting blood glucose level ≥7.8 mmol/L, patient’s self-report of diabetes, or use of antidiabetes drugs), hyperlipidemia (reported fasting total cholesterol ≥5.72 mmol/L, reported low-density lipoprotein (LDL) ≥3.64 mmol/L, reported high-density lipoprotein (HDL) ≤0.91 mmol/L, reported triglyceride ≥1.70 mmol/L, patient’s self-report of hyperlipidemia, or use of antihyperlipidemic drugs).
The data were input into computer with the software package of Foxbase+. The Kaplan–Meier method was used to estimate the probability of survival and recurrence. The difference of survival and recurrence probability between the intervention and control communities was tested with the log-rank method. The Cox proportional hazards model was used to examine the risk factors as independent determinants of survival and recurrence rates. Two hundred twenty-eight subjects, 85 of which had missing values and 143 of which had no positive time, with survival time <0.1 day, were automatically excluded from the survival analysis. Eighty-six censored cases before the earliest event in a stratum were automatically excluded in the recurrence analysis of the Cox proportional hazards model. The analyzed risk factors were age, sex, community intervention, types of stroke, histories of hypertension, TIA, cardiac diseases, mellitus diabetes and hyperlipidemia, symptoms/signs of vomiting, headache, coma, hemiplegia, diplopia, aphasia, hemianopsia, dysesthesia, meningeal irritation, vertigo (disorders of gaits), and dysarthria (difficulty in swallowing, enunciation). The information on the factors was collected at the onset of first-ever stroke. In the multifactor survival analysis, histories of diabetes and hyperlipidemia were not included in the model because of lack of information for them during 1991 to 1995. All statistical calculations were performed by computer using SPSS software (SPSS Inc). P<0.05 were considered statistically significant.
Survival Rates of Stroke
Within 3 years, 41.85% of 736 patients in the intervention community died whereas 40.34% of 818 patients in the control community died. No significant difference of the survival probability between the intervention and control communities was found (P=0.5445, Figure 1). In terms of stroke types, the cumulative mortalities of ischemic, hemorrhagic, and unspecified stroke respectively were 29.25% (141/482), 64.13% (143/223), and 77.42% (24/31) in the intervention community within 3 years, and 24.95% (136/545), 72.13% (176/244), and 62.07% (18/29) in the control community. The survival probability of ischemic stroke was significantly higher than that of hemorrhagic stroke in both intervention and control communities (Figure 2). The survival probability of hemorrhagic stroke in the intervention community was higher than that in the control community. The difference was marginally significant (P=0.0505).
Recurrence Rates of Stroke
Twenty-six (11.66%) out of the 223 cases in the intervention community and 52 (20.80%) out of the 250 cases in the control community had a recurrent stroke within 3 years. The statistical difference of the recurrent rates between the two communities was significant (P=0.0291, Figure 3). As seen in Figure 3, the difference became apparent and significant from the second year.
In all the risk factors studied, age, community intervention, history of hypertension, stroke type, coma, aphasia, and vertigo were found to be the independent predictors of death after first-ever stroke. Community intervention was a protective factor for survival of stroke, especially of hemorrhagic stroke. Coma was found to be a strong predictor of both hemorrhagic and ischemic strokes (Tables 2 and 3⇓).
In all the risk factors studied, community intervention, TIA, vomiting, and hemianopsia were found to be the independent predictors of recurrent event after first-ever stroke. Community intervention was a protective factor against recurrence (Table 2).
Stroke is characterized by high rates of case fatality and recurrence. The cumulative death rate in this study was 41.1% within 3 years, which was close to 43.4% reported in Sweden.7 The cumulative recurrence rate of both fatal and nonfatal strokes was 16.5%, which was close to 16.6% reported by the South London Stroke Register.9 Compared with ischemic stroke, the cumulative death rate of hemorrhagic stroke was higher.
The health education of community populations and the management of high-risk populations have achieved great success in reducing the level of stroke-related risk factors and lowering the morbidity and mortality of cardiovascular diseases as documented in a number of reports.4–6,10,11 However, there is still no report about the effect of community intervention on the survival and recurrence rates of first-ever stroke. In this community-based trial, 2 kinds of strategies, for the whole population and the high-risk population, respectively, were taken. The former strategy for the whole population centered on health education. The latter strategy for the high-risk population was on the management of patients with hypertension, diabetes, and stroke. Multifactor analysis revealed that the recurrence risk of first-ever stroke in the intervention community was 42% lower in comparison with the control community (RR=0.58; 95% CI: 0.34 to 1.00; P=0.048). The death risk of first-ever stroke in the intervention community was 26% lower than that in the control community (RR=0.74; 95% CI: 0.61 to 0.89; P=0.002); the risk of death from hemorrhagic stroke was 39% lower (RR=0.61; 95% CI: 0.46 to 0.81; P=0.001) compared with control community. It was justifiable that the reduced risk of recurrence and death were associated with the intervention.
However, this community-based trial has a few shortcomings. Because subjects were not randomized to intervention versus control, there were some systematic differences in comorbidities (eg, TIA, cardiac diseases, diabetes, and different types of stroke) between the intervention and control patients. Such an imbalance created a confounding effect between the intervention and the unbalanced covariates.
Selection bias brought by the exclusion of some subjects from the Cox model might exist in this study, although the exclusion was not artificial. To eliminate the selection bias of this study was obviously a dilemma, because subjects with no positive time always exist in the population, and no stipulation on the recurrent event of stroke would make it difficult to judge whether it was recurrent stroke. Because of the unbalanced baseline comorbidity characteristics and the exclusion of a large number of subjects from the Cox model, one should be cautious in making definite inferences according to the results of present study. Of course, the results will be testified by more studies in the future.
A meta-analysis of 18 cohort studies of eastern Asian populations has shown that hypertension is a major risk factor of stroke, including ischemic and hemorrhagic strokes, and the linkage between blood pressure and stroke is particularly stronger in the eastern Asian population than in western population; also, the linkage between blood pressure and hemorrhagic stroke is stronger than the link between blood pressure and ischemic stroke.12 This study verified that the community intervention and history of hypertension were prognostic predictors influencing the survival probability of stroke. It is interesting that the community intervention was not as effective against the ischemic stroke as against the hemorrhagic stroke. We presume that the management of hypertension and the crucial treatment of critical stroke might improve the survival situation of stroke, especially hemorrhagic stroke. The community management of hypertension is especially suitable to the eastern Asian population with a higher proportion of hemorrhagic stroke cases.
TIA is a risk factor leading to complete stroke and stroke recurrence. Approximately one-third of people who have had at least 1 TIA will have an ischemic stroke. However, some studies found that nearly all cases of TIA could finally lead to recurrence or complete stroke.13 This study showed that the history of TIA was a strong prognostic factor of stroke recurrence, suggesting that TIA prevention and treatment is important to the prevention of stroke recurrence.
We also found that age, stroke types, and symptoms of vertigo, coma, and aphasia affected the survival status of first-ever stroke, as was reasonable from a clinical perspective. Similar findings were reported in other studies.7,14,15
Cardiac diseases, diabetes, and hyperlipidemia were also found to be major risk factors for stroke. Their effects on stroke recurrence were well-documented.7,9 However, in the present study, the stroke recurrence remained unexplained by those conventional risk factors. As for the influence of vomiting and hemianopsia on recurrence, its mechanism is still unknown and will be elucidated in future studies.
The out-migration rate in this study was 1.41%. The rate of CT/MRI diagnosis reached up to 84.68%. The data reliability was high because of quality-control measures in the course of the project, the high rate of neuroimaging diagnosis, and the low rate of out-migration. If the difference of surveillance strengths did exist in the 2 communities, the strength of surveillance in the intervention community was probably more powerful than that in the control community because of the closer contact between workers and community residents. Thus the effect of intervention would be underestimated because of bias brought by the different strengths of surveillance.
We conclude that the community intervention may be effective and beneficial to the recurrence prevention and survival improvement of stroke, especially hemorrhagic stroke. The community management of hypertension may be especially suitable to the eastern Asian population with a higher proportion of hemorrhagic stroke cases. The study also suggests that the TIA prevention and treatment is important to the prevention of stroke recurrence.
This study was supported by the funds from the National Scientific and Technical Committee and the Ministry of Health, People’s Republic of China. The serial numbers of grants were, respectively, 85-916-01-03 from 1991 to 1995 and 96-906-02-20 from 1996 to 2000.
- Received September 12, 2003.
- Revision received January 20, 2004.
- Accepted February 11, 2004.
Thorvaldsen P, Kuulasmaa K, Rajakangas AM, Rastengte D, Sarti C, Wilhelmsen L. Stroke trends in the WHO MONICA Project. Stroke. 1997; 28: 500–506.
PRC National Blood Pressure Survey Cooperative Group. Prevalence and development trends of hypertension in China. Chinese J Hypertens. 1995; 3 (suppl): 7–13.
Xiang HD, Wu W, Liu CQ, Li K, Feng JG, Zhang YT, Wang FQ, Yan SL, Wang CJ, Xu YC, Xu DY, Fu ZZ, Liu ZY, Li TL, Bai J, Fu ZY, Wang KA. An epidemiological study on DM 1995–1996, in China. Chinese J Diabetes. 1998; 6: 131–133.
Fang XH, Kronmal RA, Li SC, Longstreth WT, Cheng XM, Wang WZ, Wu SP, Du XL, Siscovick D. Prevention of stroke in urban China: a community-based intervention trial. Stroke. 1999; 30: 495–501.
Wang WZ, Wu SP, Hong Z, Yang QD, Du XL, Huang MS, Liu YH, Jiang B. The change in incidence of stroke after a community-based intervention for nine years in three cities of China. Chin J Geriatr Cardiovasc Cerebrovasc Dis. 2002; 4: 30–33.
Wang WZ, Wu SP, Yang QD, Hong Z, Du XL, Liu YH, Huang MS, Jiang B. The change of mortality of stroke after a community-based intervention trial for nine years in three cities of China. Chin J Prev Contr Chron Noncommun Dis. 2002; 10: 49–51.
Elneihoum AM, Goransson M, Falke P, Janzon L. Three-year survival and recurrence after stroke in Malmo, Sweden: an analysis of stroke registry data. Stroke. 1998; 28: 2114–2117.
Cheng XM, Ziegler DK, Lai Y-HC, Li SC, Jiang GX, Du XL, Wang WZ, Wu SP, Bao SG, Bao QJ. Stroke in China. 1986 through 1990. Stroke. 1995; 26: 1990–1994.
Hillen T, Coshall C, Tilling K, Rudd AG, McGovern R, Wolfe CD. Cause of stroke recurrence is multifactorial: patterns, risk factors, and outcomes of stroke recurrence in the South London Stroke Register. Stroke. 2003; 34: 1457–1463.
Vartiainen E, Jonsilahti P, Alfthan G, Sundvall J, Pietinen P, Puska P. Cardiovascular risk factor changes in Finland, 1972–1997. Int J Epidemiol. 2000; 29: 49–56.
Vemmos KN, Bots ML, Tsibouris PK, Zis VP, Takis CE, Grobbee DE, Stamatelopoulos S. Prognosis of stroke in the south of Greece, 1 year mortality, functional outcome and its determinants: the Arcadia Stroke Registry. J Neurol Neurosurg Psychiatry. 2000; 69: 595–600.
Appelros P, Nydevik I, Viitanen M. Poor outcome after first-ever stroke: predictors for death, dependency and recurrent stroke within the first year Stroke. 2003; 34: 122–126.