- Email: [email protected]
A community education program on atrial fibrillation: Implications of pulse self-examination on awareness and behavior
A Community Education Program on Atrial Fibrillation: Implications of Pulse Self-examination on Awareness and Behavior Frederick E. Munschauer III, MD,*,† Dena Sohocki, MBA,† Shannon Smith Carrow, MS,† and Roger L. Priore, ScD‡
Objectives: We postulated that community groups with older demographics could be taught to find and characterize their pulse rhythm for the presence of an irregular pulse (IP), which may indicate atrial fibrillation, a major risk factor for stroke. Methods: We conducted 281 community group education sessions involving 6203 attendees. Awareness objectives were to demonstrate that: (1) group education was effective in establishing awareness that an IP may indicate atrial fibrillation; and (2) this message was retained at follow-up. Behavioral objectives were to: (1) assess ability of participants to find and characterize their pulse rhythm; (2) regularly monitor pulse rhythm; and (3) act with medical appropriateness upon detecting an IP. Results: Of 6203 attendees, 4322 were older than 50 years and consented to participate. Of these consenting participants, 73.2% found their radial pulse and 91% characterized the rhythm (regular, 72.1%; irregular, 11%; undetermined, 7.9%). Telephone follow-ups on 1839 participants were performed at 30 to 60 days to assess durability of message and action taken. At follow-up, 89.1% remembered that an IP is potentially a risk factor for stroke, and 70.3% had taken their pulse since the program. Of those who discovered a new IP, 38% sought medical assessment. Conclusion: Community education programs focusing on pulse self-examination are effective in improving awareness that an IP may be a surrogate indicator of stroke risk. Such programs may lead to improved awareness of atrial fibrillation, subsequent behavioral changes, and stroke prevention. Key Words: Atrial fibrillation— community education—pulse—stroke. © 2004 by National Stroke Association
Atrial fibrillation (AF) is a major public health problem. Some 2.3 million US citizens have AF and, of these, 90,000 will have a stroke yearly. The presence of AF is the single greatest risk factor for ischemic stroke. Approximately 20% of all ischemic strokes occur in the setting of AF.1,2
AF becomes a larger health risk factor with advancing age. Below 50 years of age, estimates on prevalence range from 0.1% to 0.5%. Over the age of 55 years, the prevalence increases significantly with each decade, to approximately 9% to 14% of the population older than 85 years.
From the *Department of Neurology and †The Jacobs Neurological Institute, School of Medicine and Biomedical Sciences; ‡Department of Social and Preventative Medicine, School of Public Health and Health Professions, State University of New York at Buffalo, Buffalo, New York. Received May 11, 2004; accepted August 3, 2004. Check Your Pulse America Investigators: Washington, DC: Joseph Robinson, MD, and Thomas DeGraba, MD; Seattle, WA: Kyra Becker, MD; Tampa Bay, FL: James Pope, MD; Atlanta, GA: Mark Silverman, MD, and Michael Frankel, MD; Phoenix, AZ: Victor Salazar, MD, and Barry Hendin, MD; Detroit, MI: William O’Neil,
MD, and Steven Levine, MD; Philadelphia/Delaware Valley, PA: Dara Jamieson, MD, and Veronica Covalesky, MD; Chicago, IL: Philip Gorelick, MD, and Dan Fintel, MD; Buffalo/Rochester, NY: Frederick E. Munschauer III, MD. Address correspondence to Frederick E. Munschauer III, MD, Department of Neurology, State University of New York at Buffalo, The Jacobs Neurological Institute, Buffalo General Hospital, 100 High St, Buffalo, NY 14203. E-mail: [email protected]. 1052-3057/$—see front matter © 2004 by National Stroke Association doi:10.1016/j.jstrokecerebrovasdis.2004.08.001
208
Journal of Stroke and Cerebrovascular Diseases, Vol. 13, No. 5 (September-October), 2004: pp 208-213
PULSE SELF-EXAMINATION
Further, the prevalence of AF is increasing on an ageadjusted basis.3,4 The risk of stroke in the setting of AF also increases with age, compounding the health implications of AF, particularly in the population older than 50 years. In the Framingham cohort for participants age 50 to 59 years, the prevalence of stroke associated with AF was 6.7%, but increased to 36.2% in participants age 80 to 89 years. Strokes occurring in the setting of AF are twice as likely to be fatal as ischemic stroke from other causes. For survivors, functional outcomes in AF-associated strokes are worse than those of other causes and AFassociated strokes are more likely to be recurrent.5 Fortunately, if AF is identified, anticoagulation is associated with a 68% reduction in the risk of stroke.6 Unfortunately, many individuals with AF are unaware that they have the condition. In one community-based screening program designed to detect abnormal cardiac rhythms, approximately 3.3% of the general population was found to have either paroxysmal or chronic AF and was unaware of this condition.7,8 Compounding the problem of AF as an underdiagnosed condition is a lack of awareness in the general population of this common condition or of its medical consequences. Even if the condition is diagnosed, only about 40% to 60% of affected individuals are on effective therapy.9 Thus, AF is a relatively common medical condition associated with high morbidity and mortality, and is underdetected in the community. AF is also associated with effective, inexpensive therapies to reduce the risk of stroke. Large community-based screening programs to detect AF are, therefore, justified if the screening procedure is effective and is associated with both low risk and low cost. A very irregular pulse (IP) is widely recognized as a clinical manifestation of AF. Screening populations for an IP as a surrogate indicator of AF could be a safe screening procedure if individuals could reliably find their pulse, characterize the rhythm, and retain the knowledge that an IP may represent a risk factor for stroke. In a previous study, we demonstrated that individuals from the community could be taught to find and characterize the rhythm of their pulse and to correctly classify the pulse of another as being either regular or irregular.10 The study paradigm required individual instruction by a health care professional. Under such supervision, greater than 90% of individuals could find their own pulse and that of another person. Using disease state models with and without AF, participants could also distinguish between a regular pulse and an IP with a sensitivity and specificity greater than 73%. This study validated the concept that members of the community could be taught that an IP may indicate a risk factor for stroke and that individuals drawn from the community could find and appropriately characterize their pulse and that of another.
209
In the current study, we sought to expand the previous educational paradigm from individual instruction to a group learning experience and to test the persistence of the message. The goals of the study were to achieve a durable awareness surrounding the medical implications of AF and a sustained behavior of monitoring the rhythm of pulse. The specific awareness aims were to determine whether members of community groups older than age 50 years could learn that: (1) an IP may indicate AF; and (2) AF is a strong risk factor for stroke. Specific behavioral goals were to determine whether members of community groups older than age 50 years could: (1) find their radial pulse; (2) characterize the rhythm of their pulse; (3) monitor prospectively the rhythm of their pulse; and (4) act appropriately on detection of an IP. Additionally, we sought to determine whether these behavioral and awareness messages were durable over a 30- to 60-day follow-up.
Methods We conducted a 10-city clinical demonstration project over a period of 12 months consisting of 281 communitybased educational programs directed at increasing awareness of an IP as a risk factor for stroke. The educational content and materials used at the sessions were developed in a series of pilot sessions where the effectiveness of conveying messages was assessed. Pilot participants were polled at two and 12 weeks after the sessions to determine the durability of the message. Based on this pilot study, the final session contents and materials were adopted. Community sites were selected to be both ethnically and geographically diverse. An emphasis was made at identifying senior centers, senior housing, community centers, faith sites, and some work sites. Hospital, clinic, or stroke survivor groups; health fairs; and groups of individuals living in assisted living situations were specifically excluded. In each city, a physician was appointed as site principal investigator and 10 to 25 nurse educators were identified. Nurses were selected on the basis of previous experience in community education programs and a training session was held for the nurse educators. Background information about the public health implications of AF was presented. Additionally, the nurse educators were instructed on a standardized, validated technique to identify and characterize the radial pulse, and the standardized community program was then presented. The community program included a slide kit and an audio compact disk demonstrating regular pulse versus IP. This audio program, and scripted statements, provided examples of a very IP as compared with a pulse with occasional extra systoles. “Check Your Pulse” reminder cards were also distributed. Participants were urged to take their pulse at least monthly and, if an IP
F.E. MUNSCHAUER ET AL.
210
was noted at either the group session or afterward, participants were instructed to consult a physician. A presession and postsession questionnaire assessing session effectiveness and demographic information, and a program evaluation form, were obtained. Participants were asked to record whether they could find their pulse after the session and to characterize the rhythm, and told that they would be contacted in 30 to 60 days. The durability of both the awareness and behavioral objectives of the program was evaluated by follow-up telephone interview. Of those who remembered attending the program, the interview began with an unprompted request to discuss the key program messages. Awareness was assessed on the basis of whether the participant mentioned any of the following key words or phrases: stroke prevention, AF, IP, or learning to take your pulse. The structured interview then specifically prompted the participant to see if they remembered the primary learning objective of the program: that an IP may be a risk factor for stroke. Behavioral objectives were assessed directly by asking the participant: (1) if they detected an irregular rhythm at the program, did they consult a physician; (2) had they taken their pulse since the program; (3) if so, what was the rhythm; and (4) if they thought they had an IP did they consult a physician.
Table 1. Demographics Characteristics of consenting participants Age (y) 50-65 66-70 71-75 76-80 81⫹ Sex Male Female Race Caucasian African American Hispanic Asian Native American Other Education Grades 1-8 Some high school Graduated high school Some college College degree Postgraduate
Percentage
19.6 14.5 21.6 20.6 23.7 24.4 75.6 71.5 24.0 1.8 1.1 1.0 .6 6.0 13.5 34.2 25.7 11.9 8.7
Statistical Analysis Study forms were constructed using Teleform (Verity, Sunnyvale, CA) and scanned into a database (Access, Microsoft Corp., Redmond, WA); appropriate queries were resolved. Groups were evaluated at the baseline session and at follow-up using parametric statistical comparisons to determine if there were differences in key study variables across groups. All statistical analysis was done using software (SPSS Inc., Chicago, IL). Univariate logistic regression analyses were performed on the following dependent variables: (1) could the participant
remember without prompting at follow-up two or more major points of the program; (2) could the participant remember with prompting at follow-up that an IP was a risk for stroke; and (3) had the participant taken his or her pulse since the program. The independent variables for these analyses were the participant demographic variables, stated health problems, knowledge of pulse characteristics, and group session characteristics. Each variable was entered into a separate analysis. Forward stepwise logistic regression was then used to build a multivariate model for each dependent variable starting with variables with a P ⱕ .05 in the univariate analysis.
Results
Figure 1. Breakdown of total participants according to age, consent, and follow-up.
Some 6203 attended the group educational sessions. Of these, 4322 were older than 50 years and signed informed consent (Figure 1). There were significantly greater numbers of female than male participants reflected in the sex distribution at the principal community places such as senior centers and faith sites. The age, sex, race, and education distribution of participants responding at the educational sessions is shown in Table 1. Table 2 shows the participants at each of the types of community groups. At the beginning of the sessions, participants were questioned as to whether or not they had AF. Of participants, 8% responded as having AF, 41% reported that
PULSE SELF-EXAMINATION
211
Table 2. Number of consenting participants by community group Group type
N
Percentage
Senior center Faith site Senior residence Service organization Not indicated Work site Total
1517 1027 642 569 347 220 4322
35.1 23.8 14.9 13.1 8.0 5.1
they did not have it, and 51% did not respond. Before the session, only 61% of participants had ever taken their pulse, some 37% had never taken their pulse, and the remainder were either not sure or did not answer. Of those who did take their pulse, 26% reported that at some time they had noticed an irregularity. With the group instructional paradigm, 73.2% of participants could find their own pulse, whereas 16.9% either could not or were not sure, with 9.9% not responding. Of those who could find their pulse, 72.1% thought it was regular, 11% believed that it was either irregular or very irregular, 7.9% could not characterize the rhythm of the pulse, and 9% did not respond.
Awareness Outcomes The mean time from conduct of the session to telephone follow-up was 54 days. Of those who agreed to follow-up, 6.8% did not remember attending the program. Follow-up data excludes those who did not remember the program. On follow-up, for the unprompted awareness outcome assessing how many key points were remembered, 5.8% could remember none, 58.8% remembered 1, 31.2% remembered 2, 3.8% remembered 3, and
0.4% remembered all 4 key points. In a univariate analysis of the main unprompted awareness measure of whether the participant remembered two or more points of the program, the variables that correlated significantly were: (1) education; (2) whether the participant could take his own pulse; (3) whether the participant could identify the rhythm of his pulse; (4) race; (5) type of group; (6) city; and (7) history of AF (Table 3). The variables that were simultaneously significant in a multivariate model were: (1) race (P ⬍ .004); (2) city (P ⬍ .01); (3) group type (P ⬍ .02); and (4) whether the participant knew the rhythm of his pulse (P ⬍ .04). For the prompted key awareness point that an IP can lead to a stroke, 89.1% of participants responded correctly, whereas 10.9% did not. In the multivariate analysis the significant variables predicting incorrect awareness (P ⬍ .05) were: (1) age ⬎ 65 years; and (2) city. Participants also retained the knowledge that an IP should be evaluated by a health care professional. At follow-up, 83% said that they should seek medical evaluation if they had an IP, but that it was not an emergency. Some 14.4% recognized that they should seek medical advice, but incorrectly thought that a call to 911 was the most appropriate action.
Behavioral Outcomes On follow-up, 70.3% of participants stated that they had checked their pulse since the program, while 29.7% had not. In the univariate analysis of this primary behavior outcome measure, significant variables were: (1) education; (2) whether the participant could take his own pulse; (3) hypertension before the study; (4) number of health problems; (5) group size; (6) type of group; and (7) city (Table 3). The multivariate analysis, with the dependent variable being whether the participant had taken his or her pulse since the program, depended on whether the
Table 3. Correlation between participant characteristics and awareness and behavior variables at follow-up Participant remembered 2⫹ main points (awareness)
Participant took own pulse since program (behavioral)
Characteristic
P value
P value
Age Age ⬎65 y Race Education City Type of group Group size Participant can take own pulse Participant identified rhythm of pulse History of atrial fibrillation History of hypertension
.3519 .7638 ⬍ .0001 .0005 ⬍ .0001 ⬍ .0001 .4328 .0001 .0069 .0329 .6883
.0858 .5615 .4966 .0122 ⬍ .0001 .0006 .0061 ⬍ .0001 .0564 .3395 .0032
F.E. MUNSCHAUER ET AL.
212
participant could take his pulse, the group size, and the city (P ⬍ .001 for all). A total of 174 participants who had either identified an IP at the time of the session or discovered an IP since the session were available at follow-up. Of these, 62.1% took no action (Table 4). As a result of the sessions, a total of 10 participants (0.5%) were identified and diagnosed by their physicians as having AF. An additional 18 (1%) were found to have a benign arrhythmia, with three (0.2%) reporting an arrhythmia that required some investigation. As a result of seeking medical attention, an additional 11 (0.6%) were found to have an unrelated medical condition requiring either investigation or therapy.
Discussion This clinical demonstration project establishes that community-derived groups biased toward older demographics can be taught that an IP may be a surrogate indicator for a major risk factor for stroke, AF. The community education sessions were effective in achieving a sustained improvement in awareness. At the end of the training sessions, 93% of the participants had identified an IP as a potential risk factor for stroke; on follow-up at a mean of 54 days, 89% had retained the message. Multivariate analysis demonstrated that the educational level of participants significantly correlated with the ability to retain this message. Further, the program achieved a sustained behavioral change in monitoring pulse rhythm. Before the session, only 61% of participants had ever taken their pulse. After the group educational session, 73% of participants could take their pulse. At follow-up, 70.3% of participants reported that they had taken their pulse in the interim based on the instructional paradigm emphasized during the session. During the course of the program, 174 participants discovered an IP. By the follow-up telephone call, an additional 9% believed that their pulse was irregular. A limitation of this study is that only 43% of consenting participants could be reached for follow-up. This could potentially bias the results pertaining to durability of awareness and behavioral change. Whether the IP de-
Table 4. Action taken by participants on discovery of an irregular pulse Action
N
Percentage
Did nothing Physician knew Called physician Visited physician Total
108 11 16 39 174
62.1 6.3 9.2 22.4
tected during, or subsequent to, the sessions represented AF or another cause of an IP, such as frequent premature atrial or ventricular contractions or even exaggerated sinus arrhythmia, was beyond the scope of this study. Therefore, this trial does not establish the sensitivity or specificity of pulse self-examination in determining AF electrocardiographically. The paradigm was simply to consult a physician if the participant found an IP. The concerns that this recommendation would result in unnecessary medical evaluations for benign conditions proved to be unjustified. On telephone follow-up, only 22.4% of all participants with an IP ever sought medical evaluation. We, therefore, think that the program was less successful at motivating participants to seek medical attention than it was at producing a durable change in awareness and pulse-monitoring behavior. This clinical demonstration project represents the first attempt to educate community groups of an older demographic that an IP may be a surrogate indicator for AF, a major risk factor for stroke. Sudlow and coworkers performed a community-based trial where nurses screened for AF by palpation of the radial pulse in 916 participants older than 65 years.11 If an IP was detected, the diagnosis of AF was confirmed by electrocardiography. In their study, the sensitivity of AF as detected by radial pulse examination by nurses was more than 90% with a specificity of 71%. In a previous study, we demonstrated that nurses could instruct participants in both pulse selfexamination and pulse examination in disease state model. Further, participants were able to determine whether the pulse of a disease state model was regular or irregular with a sensitivity and specificity both more than 70%. Thus, members of the community at large can be taught to examine the pulse for the presence of an irregular rhythm. Such paradigms fulfill the sensitivity and specificity criteria required for effective screening evaluations. The current trial results extend the pulse selfexamination paradigm by validating that group instruction is effective in achieving both a sustained awareness that an IP is a risk factor for stroke and a behavioral change in pulse self-examination can be learned through group instruction. We believe that this clinical demonstration project effectively increases the awareness of AF as a risk factor for stroke. At baseline, 66% of participants responded that they agreed an IP may represent AF, which can lead to a stroke. By follow-up this percentage increased to nearly 90%. Unlike other risk factors such as hypertension, smoking, diabetes, and obesity, the public is not aware that an IP may indicate a major risk factor for stroke. Because 20% of all ischemic stroke is associated with AF, programs to increase awareness that an IP may indicate AF is an important public health educational objective. Whether the technique of pulse self-examination taught in this study improves diagnosis and appropriate therapy was not an objective of this trial. However, extension of the pulse self-examination par-
PULSE SELF-EXAMINATION
adigm resulted in an increased awareness of AF and its relationship to stroke. Radial pulse self-examination may be analogous to breast self-examination. Whether breast self-examination results in increased early detection of malignant lesions is controversial.12 However, public campaigns emphasizing monthly breast self-examination have resulted in increased use of mammography, particularly in individuals at high risk.13 To date, little public health education has been directed to AF as a significant risk factor for stroke. Community-based instruction increasing the awareness of the link between an IP and a treatable risk factor for stroke might well result in improved detection of AF and should be further developed.
213
5.
6.
7.
8.
9.
References 1. Benjamin EJ, Wolf PA, D’Agostino RB, et al. Impact of atrial fibrillation on the risk of death: The Framingham heart study. Circulation 1998;10:946-952. 2. Kilander L, Andrén B, Nyman H, et al. Atrial fibrillation is an independent determinant of low cognitive function: A cross-sectional study in elderly men. Stroke 1998; 29:1816-1820 (suppl). 3. Wolf PA, Benjamin EJ, Belanger AJ, et al. Secular trends in the prevalence of atrial fibrillation: The Framingham study. Am Heart J 1996;131:790-795 (suppl). 4. Go AS, Hylek EM, Phillips KA, et al. Prevalence of diagnosed atrial fibrillation in adults. National implica-
10.
11.
12.
13.
tions for rhythm management and stroke prevention: The anticoagulation and risk factors in atrial fibrillation (ATRIA) study JAMA 2001;285:2370-2375 (suppl). Lin HJ, Wolf PA, Kelly-Hayes M, et al. Stroke severity in atrial fibrillation: The Framingham study. Stroke 1996; 27:1760-4 (suppl). Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation: Analysis of pooled data from five randomized controlled trials. Arch Intern Med 1994;154:1449-1457. Ruigomez A, Johansson S, Wallander MA, et al. Incidence of chronic atrial fibrillation in general practice and its treatment pattern. J Clin Epidemiol 2002;55:358-363 (suppl). Somerville S, Somerville J, Croft P, et al. Atrial fibrillation: A comparison of methods to identify cases in general practice. Br J Gen Pract 2000;50:727-729 (suppl). Munschauer FE, Priore RL, Hens M, et al. Thromboembolism prophylaxis in chronic atrial fibrillation. Stroke 1997;28:72-76. Munschauer FE, Hens M, Priore RL, et al. Screening for atrial fibrillation in the community: A multicenter validation trial. J Stroke Cerebrovasc Dis 1999;8:99-103 (suppl). Sudlow M, Rodgers H, Kenny RA, et al. Identification of patients with atrial fibrillation in general practice: A study of screening methods. BMJ 1998;317:327-328. Aubard Y, Genet D, Eyraud JL, et al. Impact of screening on breast cancer detection: Retrospective comparative study of two periods ten years apart. Eur J Gynaecol Oncol 2002;23:37-41 (suppl). Worden JK, Mickey RM, Flynn BS, et al. Development of a community breast screening promotion program using baseline data. Prev Med 1994;23:267-275 (suppl).
Objectives: We postulated that community groups with older demographics could be taught to find and characterize their pulse rhythm for the presence of an irregular pulse (IP), which may indicate atrial fibrillation, a major risk factor for stroke. Methods: We conducted 281 community group education sessions involving 6203 attendees. Awareness objectives were to demonstrate that: (1) group education was effective in establishing awareness that an IP may indicate atrial fibrillation; and (2) this message was retained at follow-up. Behavioral objectives were to: (1) assess ability of participants to find and characterize their pulse rhythm; (2) regularly monitor pulse rhythm; and (3) act with medical appropriateness upon detecting an IP. Results: Of 6203 attendees, 4322 were older than 50 years and consented to participate. Of these consenting participants, 73.2% found their radial pulse and 91% characterized the rhythm (regular, 72.1%; irregular, 11%; undetermined, 7.9%). Telephone follow-ups on 1839 participants were performed at 30 to 60 days to assess durability of message and action taken. At follow-up, 89.1% remembered that an IP is potentially a risk factor for stroke, and 70.3% had taken their pulse since the program. Of those who discovered a new IP, 38% sought medical assessment. Conclusion: Community education programs focusing on pulse self-examination are effective in improving awareness that an IP may be a surrogate indicator of stroke risk. Such programs may lead to improved awareness of atrial fibrillation, subsequent behavioral changes, and stroke prevention. Key Words: Atrial fibrillation— community education—pulse—stroke. © 2004 by National Stroke Association
Atrial fibrillation (AF) is a major public health problem. Some 2.3 million US citizens have AF and, of these, 90,000 will have a stroke yearly. The presence of AF is the single greatest risk factor for ischemic stroke. Approximately 20% of all ischemic strokes occur in the setting of AF.1,2
AF becomes a larger health risk factor with advancing age. Below 50 years of age, estimates on prevalence range from 0.1% to 0.5%. Over the age of 55 years, the prevalence increases significantly with each decade, to approximately 9% to 14% of the population older than 85 years.
From the *Department of Neurology and †The Jacobs Neurological Institute, School of Medicine and Biomedical Sciences; ‡Department of Social and Preventative Medicine, School of Public Health and Health Professions, State University of New York at Buffalo, Buffalo, New York. Received May 11, 2004; accepted August 3, 2004. Check Your Pulse America Investigators: Washington, DC: Joseph Robinson, MD, and Thomas DeGraba, MD; Seattle, WA: Kyra Becker, MD; Tampa Bay, FL: James Pope, MD; Atlanta, GA: Mark Silverman, MD, and Michael Frankel, MD; Phoenix, AZ: Victor Salazar, MD, and Barry Hendin, MD; Detroit, MI: William O’Neil,
MD, and Steven Levine, MD; Philadelphia/Delaware Valley, PA: Dara Jamieson, MD, and Veronica Covalesky, MD; Chicago, IL: Philip Gorelick, MD, and Dan Fintel, MD; Buffalo/Rochester, NY: Frederick E. Munschauer III, MD. Address correspondence to Frederick E. Munschauer III, MD, Department of Neurology, State University of New York at Buffalo, The Jacobs Neurological Institute, Buffalo General Hospital, 100 High St, Buffalo, NY 14203. E-mail: [email protected]. 1052-3057/$—see front matter © 2004 by National Stroke Association doi:10.1016/j.jstrokecerebrovasdis.2004.08.001
208
Journal of Stroke and Cerebrovascular Diseases, Vol. 13, No. 5 (September-October), 2004: pp 208-213
PULSE SELF-EXAMINATION
Further, the prevalence of AF is increasing on an ageadjusted basis.3,4 The risk of stroke in the setting of AF also increases with age, compounding the health implications of AF, particularly in the population older than 50 years. In the Framingham cohort for participants age 50 to 59 years, the prevalence of stroke associated with AF was 6.7%, but increased to 36.2% in participants age 80 to 89 years. Strokes occurring in the setting of AF are twice as likely to be fatal as ischemic stroke from other causes. For survivors, functional outcomes in AF-associated strokes are worse than those of other causes and AFassociated strokes are more likely to be recurrent.5 Fortunately, if AF is identified, anticoagulation is associated with a 68% reduction in the risk of stroke.6 Unfortunately, many individuals with AF are unaware that they have the condition. In one community-based screening program designed to detect abnormal cardiac rhythms, approximately 3.3% of the general population was found to have either paroxysmal or chronic AF and was unaware of this condition.7,8 Compounding the problem of AF as an underdiagnosed condition is a lack of awareness in the general population of this common condition or of its medical consequences. Even if the condition is diagnosed, only about 40% to 60% of affected individuals are on effective therapy.9 Thus, AF is a relatively common medical condition associated with high morbidity and mortality, and is underdetected in the community. AF is also associated with effective, inexpensive therapies to reduce the risk of stroke. Large community-based screening programs to detect AF are, therefore, justified if the screening procedure is effective and is associated with both low risk and low cost. A very irregular pulse (IP) is widely recognized as a clinical manifestation of AF. Screening populations for an IP as a surrogate indicator of AF could be a safe screening procedure if individuals could reliably find their pulse, characterize the rhythm, and retain the knowledge that an IP may represent a risk factor for stroke. In a previous study, we demonstrated that individuals from the community could be taught to find and characterize the rhythm of their pulse and to correctly classify the pulse of another as being either regular or irregular.10 The study paradigm required individual instruction by a health care professional. Under such supervision, greater than 90% of individuals could find their own pulse and that of another person. Using disease state models with and without AF, participants could also distinguish between a regular pulse and an IP with a sensitivity and specificity greater than 73%. This study validated the concept that members of the community could be taught that an IP may indicate a risk factor for stroke and that individuals drawn from the community could find and appropriately characterize their pulse and that of another.
209
In the current study, we sought to expand the previous educational paradigm from individual instruction to a group learning experience and to test the persistence of the message. The goals of the study were to achieve a durable awareness surrounding the medical implications of AF and a sustained behavior of monitoring the rhythm of pulse. The specific awareness aims were to determine whether members of community groups older than age 50 years could learn that: (1) an IP may indicate AF; and (2) AF is a strong risk factor for stroke. Specific behavioral goals were to determine whether members of community groups older than age 50 years could: (1) find their radial pulse; (2) characterize the rhythm of their pulse; (3) monitor prospectively the rhythm of their pulse; and (4) act appropriately on detection of an IP. Additionally, we sought to determine whether these behavioral and awareness messages were durable over a 30- to 60-day follow-up.
Methods We conducted a 10-city clinical demonstration project over a period of 12 months consisting of 281 communitybased educational programs directed at increasing awareness of an IP as a risk factor for stroke. The educational content and materials used at the sessions were developed in a series of pilot sessions where the effectiveness of conveying messages was assessed. Pilot participants were polled at two and 12 weeks after the sessions to determine the durability of the message. Based on this pilot study, the final session contents and materials were adopted. Community sites were selected to be both ethnically and geographically diverse. An emphasis was made at identifying senior centers, senior housing, community centers, faith sites, and some work sites. Hospital, clinic, or stroke survivor groups; health fairs; and groups of individuals living in assisted living situations were specifically excluded. In each city, a physician was appointed as site principal investigator and 10 to 25 nurse educators were identified. Nurses were selected on the basis of previous experience in community education programs and a training session was held for the nurse educators. Background information about the public health implications of AF was presented. Additionally, the nurse educators were instructed on a standardized, validated technique to identify and characterize the radial pulse, and the standardized community program was then presented. The community program included a slide kit and an audio compact disk demonstrating regular pulse versus IP. This audio program, and scripted statements, provided examples of a very IP as compared with a pulse with occasional extra systoles. “Check Your Pulse” reminder cards were also distributed. Participants were urged to take their pulse at least monthly and, if an IP
F.E. MUNSCHAUER ET AL.
210
was noted at either the group session or afterward, participants were instructed to consult a physician. A presession and postsession questionnaire assessing session effectiveness and demographic information, and a program evaluation form, were obtained. Participants were asked to record whether they could find their pulse after the session and to characterize the rhythm, and told that they would be contacted in 30 to 60 days. The durability of both the awareness and behavioral objectives of the program was evaluated by follow-up telephone interview. Of those who remembered attending the program, the interview began with an unprompted request to discuss the key program messages. Awareness was assessed on the basis of whether the participant mentioned any of the following key words or phrases: stroke prevention, AF, IP, or learning to take your pulse. The structured interview then specifically prompted the participant to see if they remembered the primary learning objective of the program: that an IP may be a risk factor for stroke. Behavioral objectives were assessed directly by asking the participant: (1) if they detected an irregular rhythm at the program, did they consult a physician; (2) had they taken their pulse since the program; (3) if so, what was the rhythm; and (4) if they thought they had an IP did they consult a physician.
Table 1. Demographics Characteristics of consenting participants Age (y) 50-65 66-70 71-75 76-80 81⫹ Sex Male Female Race Caucasian African American Hispanic Asian Native American Other Education Grades 1-8 Some high school Graduated high school Some college College degree Postgraduate
Percentage
19.6 14.5 21.6 20.6 23.7 24.4 75.6 71.5 24.0 1.8 1.1 1.0 .6 6.0 13.5 34.2 25.7 11.9 8.7
Statistical Analysis Study forms were constructed using Teleform (Verity, Sunnyvale, CA) and scanned into a database (Access, Microsoft Corp., Redmond, WA); appropriate queries were resolved. Groups were evaluated at the baseline session and at follow-up using parametric statistical comparisons to determine if there were differences in key study variables across groups. All statistical analysis was done using software (SPSS Inc., Chicago, IL). Univariate logistic regression analyses were performed on the following dependent variables: (1) could the participant
remember without prompting at follow-up two or more major points of the program; (2) could the participant remember with prompting at follow-up that an IP was a risk for stroke; and (3) had the participant taken his or her pulse since the program. The independent variables for these analyses were the participant demographic variables, stated health problems, knowledge of pulse characteristics, and group session characteristics. Each variable was entered into a separate analysis. Forward stepwise logistic regression was then used to build a multivariate model for each dependent variable starting with variables with a P ⱕ .05 in the univariate analysis.
Results
Figure 1. Breakdown of total participants according to age, consent, and follow-up.
Some 6203 attended the group educational sessions. Of these, 4322 were older than 50 years and signed informed consent (Figure 1). There were significantly greater numbers of female than male participants reflected in the sex distribution at the principal community places such as senior centers and faith sites. The age, sex, race, and education distribution of participants responding at the educational sessions is shown in Table 1. Table 2 shows the participants at each of the types of community groups. At the beginning of the sessions, participants were questioned as to whether or not they had AF. Of participants, 8% responded as having AF, 41% reported that
PULSE SELF-EXAMINATION
211
Table 2. Number of consenting participants by community group Group type
N
Percentage
Senior center Faith site Senior residence Service organization Not indicated Work site Total
1517 1027 642 569 347 220 4322
35.1 23.8 14.9 13.1 8.0 5.1
they did not have it, and 51% did not respond. Before the session, only 61% of participants had ever taken their pulse, some 37% had never taken their pulse, and the remainder were either not sure or did not answer. Of those who did take their pulse, 26% reported that at some time they had noticed an irregularity. With the group instructional paradigm, 73.2% of participants could find their own pulse, whereas 16.9% either could not or were not sure, with 9.9% not responding. Of those who could find their pulse, 72.1% thought it was regular, 11% believed that it was either irregular or very irregular, 7.9% could not characterize the rhythm of the pulse, and 9% did not respond.
Awareness Outcomes The mean time from conduct of the session to telephone follow-up was 54 days. Of those who agreed to follow-up, 6.8% did not remember attending the program. Follow-up data excludes those who did not remember the program. On follow-up, for the unprompted awareness outcome assessing how many key points were remembered, 5.8% could remember none, 58.8% remembered 1, 31.2% remembered 2, 3.8% remembered 3, and
0.4% remembered all 4 key points. In a univariate analysis of the main unprompted awareness measure of whether the participant remembered two or more points of the program, the variables that correlated significantly were: (1) education; (2) whether the participant could take his own pulse; (3) whether the participant could identify the rhythm of his pulse; (4) race; (5) type of group; (6) city; and (7) history of AF (Table 3). The variables that were simultaneously significant in a multivariate model were: (1) race (P ⬍ .004); (2) city (P ⬍ .01); (3) group type (P ⬍ .02); and (4) whether the participant knew the rhythm of his pulse (P ⬍ .04). For the prompted key awareness point that an IP can lead to a stroke, 89.1% of participants responded correctly, whereas 10.9% did not. In the multivariate analysis the significant variables predicting incorrect awareness (P ⬍ .05) were: (1) age ⬎ 65 years; and (2) city. Participants also retained the knowledge that an IP should be evaluated by a health care professional. At follow-up, 83% said that they should seek medical evaluation if they had an IP, but that it was not an emergency. Some 14.4% recognized that they should seek medical advice, but incorrectly thought that a call to 911 was the most appropriate action.
Behavioral Outcomes On follow-up, 70.3% of participants stated that they had checked their pulse since the program, while 29.7% had not. In the univariate analysis of this primary behavior outcome measure, significant variables were: (1) education; (2) whether the participant could take his own pulse; (3) hypertension before the study; (4) number of health problems; (5) group size; (6) type of group; and (7) city (Table 3). The multivariate analysis, with the dependent variable being whether the participant had taken his or her pulse since the program, depended on whether the
Table 3. Correlation between participant characteristics and awareness and behavior variables at follow-up Participant remembered 2⫹ main points (awareness)
Participant took own pulse since program (behavioral)
Characteristic
P value
P value
Age Age ⬎65 y Race Education City Type of group Group size Participant can take own pulse Participant identified rhythm of pulse History of atrial fibrillation History of hypertension
.3519 .7638 ⬍ .0001 .0005 ⬍ .0001 ⬍ .0001 .4328 .0001 .0069 .0329 .6883
.0858 .5615 .4966 .0122 ⬍ .0001 .0006 .0061 ⬍ .0001 .0564 .3395 .0032
F.E. MUNSCHAUER ET AL.
212
participant could take his pulse, the group size, and the city (P ⬍ .001 for all). A total of 174 participants who had either identified an IP at the time of the session or discovered an IP since the session were available at follow-up. Of these, 62.1% took no action (Table 4). As a result of the sessions, a total of 10 participants (0.5%) were identified and diagnosed by their physicians as having AF. An additional 18 (1%) were found to have a benign arrhythmia, with three (0.2%) reporting an arrhythmia that required some investigation. As a result of seeking medical attention, an additional 11 (0.6%) were found to have an unrelated medical condition requiring either investigation or therapy.
Discussion This clinical demonstration project establishes that community-derived groups biased toward older demographics can be taught that an IP may be a surrogate indicator for a major risk factor for stroke, AF. The community education sessions were effective in achieving a sustained improvement in awareness. At the end of the training sessions, 93% of the participants had identified an IP as a potential risk factor for stroke; on follow-up at a mean of 54 days, 89% had retained the message. Multivariate analysis demonstrated that the educational level of participants significantly correlated with the ability to retain this message. Further, the program achieved a sustained behavioral change in monitoring pulse rhythm. Before the session, only 61% of participants had ever taken their pulse. After the group educational session, 73% of participants could take their pulse. At follow-up, 70.3% of participants reported that they had taken their pulse in the interim based on the instructional paradigm emphasized during the session. During the course of the program, 174 participants discovered an IP. By the follow-up telephone call, an additional 9% believed that their pulse was irregular. A limitation of this study is that only 43% of consenting participants could be reached for follow-up. This could potentially bias the results pertaining to durability of awareness and behavioral change. Whether the IP de-
Table 4. Action taken by participants on discovery of an irregular pulse Action
N
Percentage
Did nothing Physician knew Called physician Visited physician Total
108 11 16 39 174
62.1 6.3 9.2 22.4
tected during, or subsequent to, the sessions represented AF or another cause of an IP, such as frequent premature atrial or ventricular contractions or even exaggerated sinus arrhythmia, was beyond the scope of this study. Therefore, this trial does not establish the sensitivity or specificity of pulse self-examination in determining AF electrocardiographically. The paradigm was simply to consult a physician if the participant found an IP. The concerns that this recommendation would result in unnecessary medical evaluations for benign conditions proved to be unjustified. On telephone follow-up, only 22.4% of all participants with an IP ever sought medical evaluation. We, therefore, think that the program was less successful at motivating participants to seek medical attention than it was at producing a durable change in awareness and pulse-monitoring behavior. This clinical demonstration project represents the first attempt to educate community groups of an older demographic that an IP may be a surrogate indicator for AF, a major risk factor for stroke. Sudlow and coworkers performed a community-based trial where nurses screened for AF by palpation of the radial pulse in 916 participants older than 65 years.11 If an IP was detected, the diagnosis of AF was confirmed by electrocardiography. In their study, the sensitivity of AF as detected by radial pulse examination by nurses was more than 90% with a specificity of 71%. In a previous study, we demonstrated that nurses could instruct participants in both pulse selfexamination and pulse examination in disease state model. Further, participants were able to determine whether the pulse of a disease state model was regular or irregular with a sensitivity and specificity both more than 70%. Thus, members of the community at large can be taught to examine the pulse for the presence of an irregular rhythm. Such paradigms fulfill the sensitivity and specificity criteria required for effective screening evaluations. The current trial results extend the pulse selfexamination paradigm by validating that group instruction is effective in achieving both a sustained awareness that an IP is a risk factor for stroke and a behavioral change in pulse self-examination can be learned through group instruction. We believe that this clinical demonstration project effectively increases the awareness of AF as a risk factor for stroke. At baseline, 66% of participants responded that they agreed an IP may represent AF, which can lead to a stroke. By follow-up this percentage increased to nearly 90%. Unlike other risk factors such as hypertension, smoking, diabetes, and obesity, the public is not aware that an IP may indicate a major risk factor for stroke. Because 20% of all ischemic stroke is associated with AF, programs to increase awareness that an IP may indicate AF is an important public health educational objective. Whether the technique of pulse self-examination taught in this study improves diagnosis and appropriate therapy was not an objective of this trial. However, extension of the pulse self-examination par-
PULSE SELF-EXAMINATION
adigm resulted in an increased awareness of AF and its relationship to stroke. Radial pulse self-examination may be analogous to breast self-examination. Whether breast self-examination results in increased early detection of malignant lesions is controversial.12 However, public campaigns emphasizing monthly breast self-examination have resulted in increased use of mammography, particularly in individuals at high risk.13 To date, little public health education has been directed to AF as a significant risk factor for stroke. Community-based instruction increasing the awareness of the link between an IP and a treatable risk factor for stroke might well result in improved detection of AF and should be further developed.
213
5.
6.
7.
8.
9.
References 1. Benjamin EJ, Wolf PA, D’Agostino RB, et al. Impact of atrial fibrillation on the risk of death: The Framingham heart study. Circulation 1998;10:946-952. 2. Kilander L, Andrén B, Nyman H, et al. Atrial fibrillation is an independent determinant of low cognitive function: A cross-sectional study in elderly men. Stroke 1998; 29:1816-1820 (suppl). 3. Wolf PA, Benjamin EJ, Belanger AJ, et al. Secular trends in the prevalence of atrial fibrillation: The Framingham study. Am Heart J 1996;131:790-795 (suppl). 4. Go AS, Hylek EM, Phillips KA, et al. Prevalence of diagnosed atrial fibrillation in adults. National implica-
10.
11.
12.
13.
tions for rhythm management and stroke prevention: The anticoagulation and risk factors in atrial fibrillation (ATRIA) study JAMA 2001;285:2370-2375 (suppl). Lin HJ, Wolf PA, Kelly-Hayes M, et al. Stroke severity in atrial fibrillation: The Framingham study. Stroke 1996; 27:1760-4 (suppl). Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation: Analysis of pooled data from five randomized controlled trials. Arch Intern Med 1994;154:1449-1457. Ruigomez A, Johansson S, Wallander MA, et al. Incidence of chronic atrial fibrillation in general practice and its treatment pattern. J Clin Epidemiol 2002;55:358-363 (suppl). Somerville S, Somerville J, Croft P, et al. Atrial fibrillation: A comparison of methods to identify cases in general practice. Br J Gen Pract 2000;50:727-729 (suppl). Munschauer FE, Priore RL, Hens M, et al. Thromboembolism prophylaxis in chronic atrial fibrillation. Stroke 1997;28:72-76. Munschauer FE, Hens M, Priore RL, et al. Screening for atrial fibrillation in the community: A multicenter validation trial. J Stroke Cerebrovasc Dis 1999;8:99-103 (suppl). Sudlow M, Rodgers H, Kenny RA, et al. Identification of patients with atrial fibrillation in general practice: A study of screening methods. BMJ 1998;317:327-328. Aubard Y, Genet D, Eyraud JL, et al. Impact of screening on breast cancer detection: Retrospective comparative study of two periods ten years apart. Eur J Gynaecol Oncol 2002;23:37-41 (suppl). Worden JK, Mickey RM, Flynn BS, et al. Development of a community breast screening promotion program using baseline data. Prev Med 1994;23:267-275 (suppl).