How Effective Are “Community” Stroke Screening Programs at Improving Stroke Knowledge and Prevention Practices?
Results of a 3-Month Follow-Up Study
Background and Purpose— Community stroke screening is a commonly used prevention strategy to identify and educate those at risk. Although the goal of this approach is to reduce the overall occurrence of stroke, its long-term benefit remains unknown. The purpose of this study was to determine whether attendance of a stroke screening changes knowledge or prevention practices in persons at risk for stroke 3 months later.
Methods— A stroke screening event was held following the National Stroke Association guidelines, with health screening, counseling, and education. Knowledge about stroke was measured by a questionnaire before and after the event. At 3 months, attendees identified at risk for stroke were contacted by telephone to determine their retained knowledge and any specific actions taken as a result of the health counseling.
Results— At 3 months, 78 persons were contacted. Knowledge of stroke warning signs increased from 59% to 94% after screening but decreased to 77% at 3 months. At 3 months, 73% had done nothing to change their health practices.
Conclusions— Community stroke screening has modest effects on health behavior, knowledge of stroke risk factors, and warning signs.
Improving community knowledge of stroke risk factors and stroke symptoms is a critical factor in improving access to preventative stroke treatments and acute stroke intervention. Recent efforts to educate the public have increased knowledge of warning signs but did not significantly improve risk factor recognition.1 Community stroke screenings are a commonly used strategy to improve recognition of stroke warning signs and risk factors, but their benefits in changing health behaviors remain unknown. The goal of this study was to determine whether attendance of a stroke screening changes knowledge or prevention practices in persons at risk for stroke 3 months later.
Subjects and Methods
Subjects consisted of a convenience sample of participants who attended an advertised community stroke screening conducted in the spring of 2000. Funding for the screening program was provided by an unrestricted educational grant from the National Stroke Association and Bristol-Myers Squibb. Human subjects approval was granted by the sponsor agency, Mercy Healthcare Sacramento and California State University Sacramento.
The screening program followed the guidelines and format for stroke screening recommended by the National Stroke Association (NSA).2 Participants completed the NSA self-risk assessment that detailed knowledge of stroke, medical history, lifestyle, and current prevention practices. The participants who consented to participate in the follow-up study supplied a contact number for follow-up. Healthcare professionals manned screening stations for blood pressure measurement, carotid bruit detection, cholesterol measurement, and atrial fibrillation screening.
Blood pressure was measured by a registered nurse using the Omron automatic blood pressure cuff. A manual cuff was available in the event that a digital pressure could not be obtained. Atrial fibrillation screening was accomplished by cardiac auscultation to detect irregular heartbeat. If an irregular rhythm was detected, a 3-lead ECG was performed to evaluate for the presence of atrial fibrillation. Heart rhythms were first interpreted by a telemetry-certified registered nurse and confirmed by a second reader who was a physician. A stroke neurologist performed all carotid screening by placing a stethoscope over the region of the carotid artery during the expiratory phase and requesting that the subject hold his or her breath. The carotid artery was auscultated for a pulsation bruit related to flow. Cholesterol measurements were performed by certified personnel with the use of the Cholestek measurement device.
Personalized health counseling that addressed individual risk factors was offered to all participants. Counseling focused on follow-up with primary care providers for those at risk, and a plan was devised to modify risk factors through medication compliance, dietary changes, smoking cessation, and exercise. A written prevention plan was provided detailing necessary lifestyle changes, and educational material about specific risk factors was available. Before leaving the screening event, participants were asked to complete a second questionnaire that measured knowledge of stroke.
Descriptive data from the results of screening and risk factor profiles were examined to determine baseline knowledge of stroke, demographics, baseline lifestyle characteristics, medical history, and prevention practices. Those found to have 1 or more modifiable risk factors for stroke were designated as high risk. Participants in the high-risk group who supplied their telephone number were contacted at 3 months to assess recall of risk of stroke, interval change in behavior to reduce risk factors, and knowledge of stroke. The follow-up telephone calls were conducted by trained healthcare professionals and followed a prescribed format. Follow-up by telephone was attempted 3 times before subjects were declared lost to follow-up. The collated descriptive data from 39 654 subjects at other screenings were made available by the NSA for comparison of demographics.
Statistical analysis was performed with the use of SPS version 11.5. Patients took a 3-question quiz testing the ability to recognize the symptoms of a stroke. The same quiz was taken at 3 different times: before receiving education on identifying the signs of a stroke, immediately after this education, and 3 months after this education. Each individual question was scored as either a 1 for correct or a 0 for incorrect. An overall score of 0, 1, 2, or 3 correct answers could be achieved on each of the 3 tests. A Friedman ANOVA by ranks test was used to test for differences in overall quiz score performance by subjects at each of the 3 time periods. The McNemar change test was subsequently used to determine the specific testing periods and questions where changes in performance occurred.
Approximately 400 people participated in the screening program, and 186 completed the questionnaire and selected screening exams. Key demographic features of the community participants are compared with the NSA national sample in the Table. Our population is similar except for an increased representation of Asians.
One hundred thirteen participants were found to have at least 1 modifiable risk factor for stroke. Recognition of the signs and symptoms of stroke before the screening was poor, with 23% of participants selecting chest pain as a sign of stroke and 31% selecting shortness of breath. Twenty-two percent believed that a stroke occurred in the heart, and another 8% were unable to decide where a stroke occurred. Most participants were able to identify weakness or numbness as signs of stroke, but only 47% would call 911 if stroke symptoms developed.
A test conducted after education and screening (n=110) showed dramatic increases in stroke symptom recognition and overall knowledge. Ninety-four percent could recognize symptoms of stroke, defined as unilateral weakness or numbness, difficulty talking or understanding, or difficulty with vision. Ninety-eight percent reported that they would call 911, but these improvements in stroke knowledge were incompletely retained at 3 months (Figure).
Seventy-eight high-risk subjects were contact at 3 months. Only 19% recalled being informed of risk for stroke, even though those found to be at risk received counseling and received a written plan and advice to follow up with their primary care provider. Twenty-seven percent made a lifestyle change. Despite the emphasis on physician follow-up, only 9% saw a doctor. Sixty-four percent made no change at all. None of the smokers stopped smoking.
Statistical analysis demonstrated significant differences in ability to recognize risk factors over time. The results of the Friedman ANOVA by rank indicated that there were significant differences in each time point (r2=35.48, χ2, P<0.001).
In our experience, stroke screenings tend to attract a highly motivated older population containing both the “worried well,” who seem highly interested in staying healthy, and the “worried sick,” who are looking for free health advice. Surprisingly, the results suggest that a minority of these motivated individuals screened for stroke and advised of their risk factors change their behavior and even fewer follow up with their physician. Multiple studies confirm the difficulty in changing lifestyle among those at risk for cardiovascular disease.3–5 The response of 36% (27% modified risk plus 9% saw physician) of high-risk patients to either seek medical care or modify risks is encouraging. Unfortunately, the majority of patients still failed to take steps to control risk factors.
One of the limitations of this study was the convenience sample design. The selection bias of “worried well” may have created trends in the data reflecting those who were more motivated to change their lifestyle and improve their health. Results may have been less favorable if a random sample had been screened. Because all screenings are voluntary, a convenience sample is likely a reasonable estimate of the population who attend. Another limitation identified is the small sample size and number of patients lost to follow-up. Larger samples are required to determine whether our data can be replicated.
The advertised stroke screening attracted at-risk patients, with more than half of the sample exhibiting at least 1 modifiable risk factor. Very few smokers attended the screening. These findings support previous research that demonstrated that those in the highest-risk groups do not attend screenings.6 Those attracted to this health screening were older adults with a mean age of 73 years. The older age of participants may have played a role in the poor recall of advice and retention of knowledge.
While health counseling was provided during the screening, stroke screening personnel did not reinforce the healthcare advice after the screening. One community-based blood pressure screening program found that minor modifications in educational material and assistive prompts increased compliance with physician follow-up to 94% in those found to be hypertensive. All patients who saw a physician did so within 1 month of the screening (78% with minimal prompting and 95% after intervention).7 These data suggest that there may be a critical period to reinforce teaching and change health behavior.
The most promising role for prevention in current medical practice lies in changing health behavior before disease develops.8 Improved communication between stroke screening teams and physician providers may boost follow-up and build on the small gains measurable at community-based stroke screenings and health fairs.
We acknowledge Catholic Healthcare West Sacramento Region, Bristol-Myers Squibb, and the National Stroke Association, who graciously provided unrestricted grant funding for this important project. We would also like to acknowledge the dedicated community healthcare providers who volunteered to provide the health screenings.
- Received April 30, 2003.
- Revision received June 17, 2003.
- Accepted July 23, 2003.
National Stroke Association. Stroke Prevention Screening Guide. Englewood, Colo; National Stroke Association; 1996.
Ebrahim S, Smith GD. Systematic review of randomized controlled trials of multiple risk factor interventions for preventing coronary heart disease. BMJ. 1997; 314: 1666–1674.
Joseph LN, Babikian VL, Allen NC, Winter MR. Risk factor modification: the experience of the stroke clinic. Stroke. 1999; 30: 16–20.
Cronin JA, Bowen M. Comparison of risk factors for coronary artery disease among attenders and non-attenders at a screening programme. Br J Gen Pract. 1993; 43: 375–377.
Kane ML, Iwata BA, Kane DF. Temporal effects of prompting on acceptance and follow-up in a community-based hypertension screening program. J Community Psychol. April 12 1984; (2): 164–172.
US Preventative Services Task Force. Guidelines From Guide to Clinical Preventative Services, Second Edition. Baltimore, Md; Williams & Wilkins; 1996.