Knowledge of stroke among stroke patients and their relatives in Northwest India
Correspondence Address: Source of Support: None, Conflict of Interest: None PMID: 16804258
Source of Support: None, Conflict of Interest: None
Background: The knowledge of warning symptoms and risk factors for stroke has not been studied among patients with stroke in developing countries. Aims: We aimed to assess the knowledge of stroke among patients with stroke and their relatives. Settings and Design: Prospective tertiary referral hospital-based study in Northwest India. Materials and Methods: Trained nurses and medical interns interviewed patients with stroke and transient ischemic attack and their relatives about their knowledge of stroke symptoms and risk factors. Statistical Analysis: Univariable and mulivariable logistic regression were used. Results: Of the 147 subjects interviewed, 102 (69%) were patients and 45 (31%) were relatives. There were 99 (67%) men and 48 (33%) women and the mean age was 59.7±14.1 years. Sixty-two percent of repondents recognized paralysis of one side as a warning symptom and 54% recognized hypertension as a risk factor for stroke. In the multivariable logistic regression analysis, higher education was associated with the knowledge of correct organ involvement in stroke (OR 2.6, CI 1.1- 6.1, P =0.02), whereas younger age (OR 2.7, CI 1.1-7.0, P =0.04) and higher education (OR 4.1, CI 1.5-10.9, P =0.005) correlated with a better knowledge regarding warning symptoms of stroke. Conclusions: In this study cohort, in general, there is lack of awareness of major warning symptoms, risk factors, organ involvement and self-recognition of stroke. However younger age and education status were associated with better knowledge. There is an urgent need for awareness programs about stroke in this study cohort.
Keywords: Stroke, awareness, patients, knowledge, risk factor, India.
Despite recent advances in stroke therapy, the majority of stroke patients do not seek immediate medical attention. Even in developed countries like USA,, UK and France there is a lack of knowledge among stroke patients about warning symptoms and risk factors. In a multi-centre survey in USA, over one-half of patients at increased risk for stroke were unaware of their risk factors.
Intravenous (IV) recombinant tissue plasminogen activator (rt-PA) is being used for acute ischemic stroke in India.,, Knowledge about stroke warning symptoms and risk factors are essential for the patients to effectively utilize the thrombolytic therapy for acute stroke. There are no studies from India and other developing countries regarding stroke patients' knowledge about warning symptoms and risk factors. We aimed to assess the knowledge of stroke warning symptoms and risk factors among stroke patients and their relatives.
The Stroke section, Department of Neurology in a tertiary referral centre in Northwest India, conducted this study and the study period was 15 months, from June 2002 to September 2003. Stroke and transient ischemic attack (TIA) patients admitted through emergency department and their relatives formed the subjects. Trained nurses and medical interns conducted a standardized structured interview with open-ended questions. The questionnaire was adapted and modified from a previous survey among general public conducted in Northwest India. The first section of the questionnaire gathered demographic information. Education was categorized into illiterates, primary (below 5th standard), secondary (6th standard to 12th standard) and college education. Income was classified into upper (Indian National Rupees > 5000 per month) and lower (Indian National Rupees < 5000 per month) income groups. In section two, the time of onset of stroke, patients or relatives' initial reaction to stroke symptoms, onset to arrival time to the hospital, were noted. The time of onset of stroke was defined as the time neurological deficit was first noticed by the patient or a relative or caregiver. When the symptoms were first noted on awakening, the time of awakening was recorded as the time of onset because it represents the time when medical help could be sought.,,, The third section consisted of questions to explore the knowledge of organ affected in stroke, warning symptoms and risk factors for stroke. There were options for multiple responses, in questions concerning warning symptoms and risk factors for stroke. The type of stroke (ischemic or hemorrhagic), neurological findings and patients' risk factors were documented in the fourth section of the questionnaire. When the stroke patients were unable to participate because of speech and language involvement or altered sensorium, one of the relative was interviewed. The questions were asked during a one-to-one interview in the local vernacular language (either Punjabi or Hindi). No attempt was made to prompt the respondents by suggesting answers directly. The questionnaire was pretested using a sample of 25 patients. Changes were made in the questionnaire to various terms that are used for "stroke" in the local languages Punjabi and Hindi. This was done entirely to differentiate heart attack and stroke in the local language. To minimize the in-hospital stroke education, all subjects were interviewed within 48 hours of admission.
We compared the knowledge of stroke between different groups; early arrivers (who arrived in hospital less than £6 hours after the onset of stroke) versus late arrivers (who arrived in hospital more than >6 hours of the onset of stroke) and between patients and relatives.
All statistical analysis was performed using SPSS software version 10.05 (SPSS Inc. Chicago, Illinois). c tests were used to assess the univariable relationship between components of stroke knowledge, warning symptoms, risk factors and demographic variables. Multivariable logistic regression was used to assess the predictors of knowing a single correct response to various questions. Variables included in the model were age (less or more than 60 years), gender, religion (Hindus vs. others), education (lower [illiterates and primary] vs. higher [secondary and college education]), income (upper vs. lower) and place of residence (urban vs. rural). Variables were eliminated in a stepwise backward fashion if they failed to reach significance ( P <0.05) until a final model resulted. Finally, odds ratios (OR) and 95% CI were generated for all the terms in the final models. c 2sub test and Fisher's exact tests were used to compare the knowledge of stroke between different groups. The hospital research committee had approved this study and informed consent was obtained from the patient or the relative.
Of the 173 subjects contacted, 147 subjects consented to participate in the study. One hundred and two (69%) patients and 45 (31%) relatives were interviewed. The stroke type was ischemic in 96 (65%), hemorrhagic in 39 (27%) and TIA in 12 (8%). There were 99 (67%) men and 48 (33%) women. The mean age was 59.7 ± 14.1 years (range 23-95 years). The demographic characteristics of the study cohort was similar to that of the population of Punjab, except for the elderly age group (>60 yrs) and the place of domicile [Table - 1]. The proportion of elderly subjects was higher (52%) than the population of Punjab (9%). Sixty-two percent of the study cohort were from urban area as compared to the population of Punjab (34%).
Subjects' initial reaction to stroke symptoms
One-hundred and seven subjects (73%) were not aware that the initial symptoms were due to stroke. Fifty-six (38%) subjects reached the hospital immediately. The rest consulted a private doctor 86 (59%) or called a relative 5 (3%).
Organ affected in stroke
Only one third of the subjects interviewed recognized that the organ injured in stroke is brain [Figure - 1]. In the univariable analysis higher education (OR 2.3, CI 0.16-3.1, P =0.02) was associated with a better knowledge about the organ injured in stroke [Table - 2]. In the multivariable logistic regression analysis, higher education remained a significant factor of knowing the correct organ involved in stroke [Table - 3].
The most common warning symptom described by the subjects was paralysis of one side of the body 91(62%) [Figure - 2]. Fifty (34%) subjects correctly identified one symptom, 42 (29%) identified two symptoms, only 12 (8%) knew three or more symptoms, while 43 (29%) did not know even a single warning symptom. In univariable analysis, younger age (OR 2.4, CI 1.1-5.1, P =0.04) and higher education (OR 3.3, CI 1.5-6.9, P =0.006) correlated with knowing at least one warning symptom [Table - 2]. In multivariable logistic regression, both younger age and higher education remained significant [Table - 3].
Hypertension 79 (54%) and diabetes 45 (31%) were the two most common risk factors identified by the study cohort [Figure - 3]. Only 55 (37%) of the subjects could identify one risk factor correctly, 31 (21%) subjects two risk factors and only 19 (13%) 3 or more risk factors. In the univariable analysis, higher education (OR 2.0, CI 0.98 - 4.1, P = 0.03) correlated with better knowledge of at least one risk factor [Table - 2]. However, in the multivariable logistic regression analysis, none of the variables reached any statistical significance.
Risk factor recognition among high risk individuals
There were 120 (82%) subjects who had various risk factors for stroke. A high proportion of subjects with hypertension 64/93 (69%) and diabetes 36/50 (72%) were able to correctly identify their risk factors for stroke. However, a very low proportion of cohort who had other risk factors such as heart disease 4/18 (22%), dyslipidemia 11/38 (29%) and smoking 13/36 (36%) could recognize their own risk factors for stroke.
Comparison of knowledge of stroke between different subgroups
We compared the knowledge of stroke between early (n=61, 41%) and late arrivers (n=86, 59%) to the hospital and also between the patients (n=102, 69%) and the relatives (n=45, 31%). We did not find any differences between the two groups with regards to knowledge about organ involved in stroke, warning symptoms and risk factors for stroke.
This is the first study to examine the knowledge of stroke warning symptoms and risk factors among stroke patients in a developing country. The majority (73%) of subjects did not realize that the symptoms were due to stroke. Only a third of the study cohort correctly identified the brain as the affected organ in stroke. Even though our subjects had a better knowledge to some of the stroke warning symptoms and risk factors, only a smaller proportion of the subjects were able to recognize other warning symptoms and risk factors.
In this study self recognition of stroke symptoms by patients was lower (27%) when compared to what has been reported in other similar studies in developed countries, 36% in the Cincinnati study and 40% in the UK study, but was similar to other studies, 21% in the France study and 25% in the Indianapolis study.
The knowledge regarding the organ injured in stroke in our cohort was less (33%) when compared to the study in Cincinnati (49%). A small proportion of our subjects (2%) were unable to differentiate heart attack from stroke. In a recent survey conducted among the general public in Northwestern India, 10% of the respondents had difficulty in differentiating heart attack and stroke. Recognizing symptoms of stroke in the community is more difficult than recognizing heart attack, because stroke symptoms are more heterogeneous.
The majority of our respondents correctly answered paralysis of one side (62%) as a warning symptom and hypertension (54%) as a risk factor for stroke as compared with the study from Cincinnati (26% and 27%, respectively). However, only a minority of our subjects were able to identify other warning symptoms and risk factors.
We found a positive relationship between stroke knowledge and higher education. The plausible explanation is that, majority (62%) of our patients were from urban area and in India educational opportunities are more in the urban than in the rural areas. However, Kothari et al did not find any correlation between better awareness and education. Most studies on public awareness of stroke from developed countries have found that knowledge about stroke varies positively with education.,,
We acknowledge the limitations of our study. Firstly, we interviewed the family members when the patients were unable to participate, because of dysphasia or altered sensorium. The sample may not fully reflect the knowledge of stroke among stroke population. In India, the structure of the family is different from the developed world. Approximately 26% of the population in Punjab lives in a joint family system i.e. two or three families live together in a single house, with their parents. It is highly essential in an Indian family setup, for the relatives of stroke patients to be aware of the warning symptoms and risk factors. This is one principal reason why we included the relatives of stroke patients in our study. Moreover, knowledge of stroke did not differ between patients and relatives. Secondly, this is a hospital-based study with a small sample size and may not exactly represent the entire population of Punjab, hence the findings limits generalisability. Moreover, India is a vast country with diverse socio-cultural and linguistic practices. Our findings can not be extrapolated to other states in the country. Thirdly, the observations in the present study is not different from a previous study of awareness of stroke among the general public in Northwestern India. The added value of this study is that it provides new information for the first time, regarding the knowledge of stroke among stroke patients in a developing country. Moreover, this study would enable us to develop appropriate educational campaigns for the stroke patients and their relatives to improve the prevention and early treatment of stroke.
In conclusion, we observed a better knowledge in our study cohort, to some of the warning symptoms and risk factors. However, the awareness of other major warning symptoms, risk factors, organ involved in stroke and self-recognition of stroke symptoms were poor. Community based studies are required in the future including both urban and rural populations to confirm our findings. There is an urgent need to educate our patients and their relatives and such educational programs should also target high risk groups.
We sincerely thank Mrs. Soosamma Verghese and Ms. Dewinder Kaur for their assistance in data collection. We are grateful to Mr Douglas J Lincoln MBiostat, Royal Brisbane and Women's Hospital Research Foundation, Brisbane, for his help in the statistical analysis.
[Figure - 1], [Figure - 2], [Figure - 3][Table - 1], [Table - 2], [Table - 3]