Abstract. ª 2006 by the Society for Academic Emergency Medicine ISSN

Transcription

1 How Well Are Cardiopulmonary Resuscitation and Automated External Defibrillator Skills Retained over Time? Results from the Public Access Defibrillation (PAD) Trial Barbara Riegel, DNSc, RN, Sarah D. Nafziger, MD, Mary Ann McBurnie, PhD, Judy Powell, BSN, Robert Ledingham, MS, Ruchir Sehra, MD, LynnMarie Mango, BS, Mark C. Henry, MD, and the PAD Trial Investigators Abstract Background: The current standard for cardiopulmonary resuscitation (CPR) and automated external defibrillator (AED) retraining for laypersons is a four-hour course every two years. Others have documented substantial skill deterioration during this time period. Objectives: To evaluate 1) the retention of core CPR and AED skills among volunteer laypersons and 2) the time required to retrain laypersons to proficiency as a function of time since initial training. Methods: This was an observational follow-up study evaluating CPR and AED skill retention and testing/ retraining time up through 17 months after initial training. The study took place at 1,260 facilities recruited by 24 North American clinical research centers, and included 6,182 volunteer laypersons participating in the Public Access Defibrillation (PAD) Trial. Training to proficiency in either CPR only (N = 2,426) or CPR+AED (N = 3,756) was followed by testing/retraining provided three to 17 months later. Retraining was done in brief, one-on-one, individualized, interactive sessions. The outcome studied was instructors global assessments of performance of CPR and AED skill adequacy, i.e., whether CPR actions would likely result in perfusion (yes/no) and whether AED actions would result in a shock through the heart (yes/no). Results: For global CPR performance, 79%, 73%, and 71% of volunteers tested for the first time since initial training three to five, six to 11, and 12 to 17 months after initial training, respectively, were judged by their instructors as having adequate performance (p < 0.001, chi-square for linear trend). For global AED performance, 91%, 86%, and 84% of volunteers, respectively, were judged as having adequate performance (p < 0.001). The mean ( standard deviation) times required to test and retrain volunteers to proficiency were 5.7 (4.0) minutes for CPR skills and 7.7 (4.6) minutes for CPR+AED skills. Conclusions: Among PAD Trial volunteer laypersons participating in a simulated resuscitation, the proportions of volunteers judged by instructors to have adequate CPR and AED skills demonstrated small declines associated with longer intervals between initial training and subsequent testing. However, based on instructors judgment, large majorities of volunteers still retained both CPR and AED core skills through 17 months after initial training. Furthermore, individual testing and retraining for CPR and AED skills were usually accomplished in less than 10 minutes per volunteer. Additional research is essential to identify training and evaluation techniques that predict adequate CPR and AED skill performance of laypersons when applied to an actual cardiac arrest. From the University of Pennsylvania (BR), Philadelphia, PA; University of Alabama at Birmingham (SDN), Birmingham, AL; Clinical Trial Center, University of Washington (MAM, JP, RL), Seattle, WA; Center for Health Research, Kaiser Permanente Northwest (MAM), Portland, OR; Loma Linda University (RS), Loma Linda, CA; Wayne State University (LM), Detroit, MI; and Stony Brook University (MCH), Stony Brook, NY. The PAD Trial investigators and coordinators are listed in: Birnbaum A, McBurnie MA, Powell J, et al., for the PAD Investigators. Modeling instructor preferences for CPR and AED competence estimation. Resuscitation. 2005; 64: Received September 11, 2005; revision received September 11, 2005; accepted October 11, Supported by contract N01-HC from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD. Additional support was provided by the American Heart Association, Dallas, TX; Medtronic, Incorporated, Minneapolis, MN; Guidant Foundation, Indianapolis, IN; Cardiac Science/SurVivaLink, Incorporated, Minneapolis, MN; Medtronic Physio-Control Corporation, Redmond, WA; Philips Medical Systems, Heartstream, Seattle, WA; and Laerdal Medical Corporation, Wappingers Falls, NY. Address for correspondence and reprints: Barbara Riegel, DNSc, RN, School of Nursing, University of Pennsylvania, 420 Guardian Drive, Philadelphia, PA Fax: ; briegel@nursing.upenn.edu. 254 ISSN PII ISSN ª 2006 by the Society for Academic Emergency Medicine doi: /j.aem

2 ACAD EMERG MED March 2006, Vol. 13, No ACADEMIC EMERGENCY MEDICINE 2006; 13: ª 2006 by the Society for Academic Emergency Medicine Keywords: cardiopulmonary resuscitation, automated external defibrillators, skill retention, publicaccess defibrillation, cardiac arrest, bystander CPR, laypersons, training The true incidence of out-of-hospital cardiac arrest (OOH-CA) is unknown; however, it is estimated that at least 340,000 people die suddenly of cardiovascular disease each year outside of the hospital. 1 Early defibrillation is the most effective treatment for patients whose initial arrest rhythm is ventricular fibrillation 2 ; however, even bystander-initiated cardiopulmonary resuscitation (CPR) deemed to be of poor or questionable quality is associated with improved survival and neurologic outcome after cardiac arrest One strategy for increasing survival from OOH-CA aims to increase the number of laypersons trained to perform CPR and defibrillation. Laypersons, unlike security guards, fire personnel, or other public safety personnel, are very unlikely to witness an OOH-CA during their lifetimes. Therefore, CPR and automated external defibrillator (AED) training for these laypersons must focus on essential skills that are easy to remember and perform if they are unexpectedly confronted with an actual OOH-CA. Over the past two decades, CPR (and, more recently, AED) training programs have been implemented throughout the world. Estimates of bystander CPR rates for OOH-CA episodes vary; however, despite intensified training efforts, these estimates generally remain unacceptably low One barrier to performing CPR could be failure to remember necessary skills. Diverse methods have been used for CPR training and assessment of skill retention Studies in simulated environments show that CPR skills are difficult to master for both professionals and laypeople, and skill retention declines significantly over time Although not as extensively studied, AED process skills appear to be easier to learn and retain than CPR skills Current American Heart Association (AHA) CPR and AED training courses certify laypersons for two years, but the optimal retraining frequency required to maintain these skills has not been rigorously established. 38 The purpose of this study was to evaluate CPR and AED skill retention in laypersons after initial training. METHODS Study Design The methodology of the PAD Trial has been previously described. 39 Briefly, the PAD Trial was a prospective, multicenter, randomized, controlled community trial involving 24 North American research centers. The trial began in 2000 and ended in Institutional review board approval for the PAD Trial was obtained at each research center participating in the trial. Written informed consent was obtained from volunteers prior to training. Study Setting and Population Centers that were randomized comprised a total of 993 community units, which were recruited based on their estimated risk of OOH-CA. Facilities such as shopping malls, fitness centers, office complexes, and apartment buildings could be combined to meet this criterion; thus, a total of 1,260 individual facilities participated in the trial. Most units (84%) were workplace facilities; 16% comprised multiunit residential facilities. Volunteers were recruited from each facility and received CPR-only or CPR+AED training according to the community unit s randomization assignment. The trial demonstrated that adding AEDs to a CPR-trained layperson volunteer response system doubled the number of survivors of cardiac arrest. 40 A secondary objective of the PAD Trial, which is the focus of this report, was to evaluate CPR and AED skill retention in volunteers over time. Volunteers were enrolled if they had no advanced medical training or a primary duty to respond to medical emergencies. All volunteers were trained to recognize a cardiac arrest, to call 9-1-1, and to provide CPR until emergency medical services (EMS) arrival. In facilities randomized to CPR+AED, volunteers received additional AED training and access to on-site AEDs. Study Protocol Instructors volunteered or were hired to provide the CPR and CPR+AED training and retraining at the facilities. The same instructors taught both types of classes and were required to be experienced trainers with at least basic life support (BLS) instructor certification. Approximately 56 instructors conducted the retraining at the various facilities (mean 2.4, range 1 8). Some trained and retrained the same units throughout the study, while others rotated through various units. The instructors were unaffiliated with the units and unknown to the volunteer laypersons. The study coordinators at the research centers carefully trained the instructors in the testing and retraining protocol. No demographic or experience data were collected on these study personnel, because they were not considered study participants. The PAD training guidelines closely followed the AHA HeartSaver AED course, which was used by most centers. Courses were required to meet the following criteria: class length of three to four hours, depending on whether AED training was included; student-to-instructor ratio of not more than 6:1 (preferably 4:1); no more than 12 students per class; case-based training scenarios; at least 20 minutes of skill practice per trainee (preferably 30 minutes); and not more than 45 minutes of lecture and demonstration. Use of a skill video was recommended. CPR training methods were the same in both groups, with CPR skills generally taught first in the AED+CPR course, followed by instruction in AED skills. The PAD training guidelines did not require a written evaluation or proficiency in a pulse check. All centers taught ventilations using a pocket mask or a face shield. Barrier devices were distributed to facilities and packaged with AEDs.

3 256 Riegel et al. CPR/AED SKILL RETENTION In order to evaluate skill retention prior to retraining, each volunteer completed a skill test conducted on a one-to-one basis with an instructor. The volunteer was asked to demonstrate CPR (with or without AED) skills on a manikin, without prompting, while the instructor completed a skill checklist. The test consisted of one to five cycles of CPR and one to four cycles of AED shocks, depending on how well the volunteer performed during the first cycles. Those who successfully completed the skill test were not given additional training. When mistakes or inconsistencies were identified, the instructor retrained the volunteer after the test to correct skills as needed in order to reestablish proficiency. Instructors used a stopwatch to document the total amount of time (in minutes) needed to test and retrain each volunteer. In some cases, retraining was performed in a group fashion after all volunteers completed the skill test. In this situation, total testing/retraining time was not calculable on an individual basis. Five core CPR skills and five core AED skills were agreed upon by the full PAD Trial steering committee prior to the start of the trial. Proficiency in these skills was evaluated using a checklist with predefined criteria for each skill (Table 1). In addition, instructors were asked for an overall global assessment, an a priori identified measure of performance adequacy. The global assessment reflects the judgment of proficient and experienced trainers with regard to whether the CPR performed would have been adequate to produce perfusion (yes/no) and whether the AED application would have delivered a shock approximately through the heart (yes/no). Global measures have been evaluated by others 41 and found to better reflect the process that experienced CPR instructors use in real-life training situations. Thus, an instructor could judge overall performance as adequate, even if one or more individual CPR skills were not completed successfully. Both individual skills and global assessments are reported here. Measures The primary CPR and AED skill retention outcome measures were the instructors overall global assessments of whether the set of skills was performed adequately (yes/ no). Secondary outcome measures were individual skill adequacy and the number of minutes required for testing and retraining. These measures were used to evaluate the relationships of individual skill performance and the amount of time required for testing and retraining with the months elapsed since initial training. Data Analysis Analyses excluded volunteers who reported participation in any additional skill refresher or retraining courses (or in a real or mock OOH-CA) between initial training and scheduled retraining, volunteers who were tested and retrained before three months or after 17 months, and volunteers for whom the testing protocol was not followed (Figure 1). Deviations in testing protocol included CPR-only trained volunteers tested in CPR+AED and group testing. In addition, five centers with artificially high pass rates (e.g., 100% proficiency in all volunteers on all skills) were identified early in the trial; these early data were excluded from analyses because Table 1 Performance Skill Checklist Developed through Consensus of the Public Access Defibrillation (PAD) Trial Steering Committee Skill CPR skills Assess responsiveness Access Description The volunteer must have physical contact with the manikin and vocalize loud enough to awaken the victim, if possible. The volunteer must pretend to call or send someone to call Provide adequate The volunteer must provide ventilations (use of adequate ventilations to the barrier device optional) manikin, using the head tilt/chin lift maneuver necessary to open the airway, sufficient to cause the chest to rise. Apply proper hand placement Provide adequate compression depth AED skills Bare chest for electrode pad placement Place electrode pads correctly Clear self Verbally clear area Deliver shock within 90 seconds of AED arrival The volunteer must demonstrate the proper hand position over the sternum. The volunteer must depress the chest of the manikin approximately 1 1 / 2 2 inches. The reviewers may use the manikin click as an indication of appropriate depth. The volunteer must remove all clothing over the chest of the manikin prior to applying AED pads in order to successfully accomplish this action. The volunteer must remove the protective backing and affix the AED pads to the manikin s bare chest. The volunteer must make an attempt to secure the AED pads to the contour of the manikin s chest. One pad is placed on the right upper chest to the right of the sternum, and the second pad is placed on the lower left chest, covering the anterior axillary line. The volunteer must remain clear of the manikin, manikin clothing, cables, and AED from the time the AED begins analyzing. The volunteer must also be alert to potential contact by others in the situation. The volunteer must call all clear and clear others prior to pushing the shock button. Timing starts as soon the AED arrives at the side of the manikin. CPR = cardiopulmonary resuscitation; AED = automated external defibrillator. of their potential to bias results in a positive direction. Corrective actions were taken at these centers and the issue was emphasized during the subsequent steering committee meeting. Thereafter, 100% pass rates were

4 ACAD EMERG MED March 2006, Vol. 13, No function. Models were adjusted for center and for characteristics previously reported to be associated with the instructors global assessments. 42 For the CPR model, these included age, gender, minority status, marital status, previous experience in an emergency, and previous CPR training (those receiving interim retraining during the study were excluded from this analysis); and for the AED model, these included age, college education, and English as the native language. An interaction term for training group (AED+CPR vs. CPR only) and retraining interval was included to evaluate whether the addition of AED training influenced performance of CPR skills. Because of the large sample size, a p-value of 0.01 was considered to be statistically significant for the primary outcome measures. The p-values that compare volunteer characteristics across training interval groups are reported descriptively, but should not be strictly interpreted because of the number of statistical tests performed. All analyses were performed using Stata (StataCorp LP, College Station, TX). RESULTS Figure 1. Volunteers included in the analysis. not observed for any center or any trainer, and all data collected subsequently were included in the analyses. Volunteers retrained in groups were excluded from analyses involving the time required to retrain to proficiency, but they were retained in analyses of skill performance. Volunteers were grouped into three retraining intervals for analysis: those retrained three to five, six to 11, and 12 to 17 months after initial training (groups 1, 2, and 3, respectively). These intervals were chosen based on what might be considered convenient retraining intervals, i.e., less than six months, six to 12 months, and more than 12 months. Frequencies, means, and error bar plots of skill retention were generated for volunteer characteristics and outcome measures for each group. Group comparisons were made using analysis of variance or chi-square tests for linear trend, as appropriate for continuous and categorical measures. Logistic generalized estimating equations (GEE) models, incorporating an exchangeable correlation structure to model correlation within facilities, were used to obtain the (adjusted) mean predicted probabilities of performing adequate CPR and AED skills for each retraining interval. Retraining interval was modeled as ln (number of months); natural log transformation was used to meet the assumption of linearity for the logit Of 19,320 volunteers who received initial training prior to January 1, 2003, 11,079 were retrained. Of these, 4,144 were excluded from this analysis because testing and retraining occurred less than three months or more than 17 months apart (38%), they participated in interim training or in a mock or real OOH-CA episode (33%), or the testing protocol was not followed as described above. Reasons why the other 8,241 were not retrained included study attrition (18%), having volunteered at a facility where retraining was not offered (12%), and not being an active study participant when retraining occurred (8%). Active volunteers were defined as those still actively participating in the trial (had not dropped out, moved, etc.). A large percentage (62% of the 8,241 who were not retrained) of active volunteers did not attend retraining sessions offered at their facility (Table 2). In addition, retraining status was unknown for 753 presumably active participants. Thus, a total of 6,182 volunteers (32% of volunteers initially trained or 53% of volunteers offered retraining) were included in the final analysis. Volunteers included in the analysis were slightly older than those excluded (41 14 vs years, p < 0.001) but otherwise had similar baseline characteristics (gender, educational level, and history of prior CPR training) (Table 2). Volunteers who did not attend scheduled testing and retraining but were presumably active (n = 5,105) were slightly younger than those who attended retraining (39 15, p < 0.001). Volunteers tested in CPR only tended to be younger than those tested in CPR+AED ( years vs years, respectively, p < 0.001) and to have less education (30% high school or less vs. 26%, p = 0.006). There were no differences between the CPRonly and CPR+AED groups for gender (52% male vs. 54% male), prior CPR training (54% vs. 55%), and the time from initial training to testing and retraining ( vs months). There were 2,839 (group 1), 2,549 (group 2), and 794 (group 3) volunteers tested and retrained three to five,

5 258 Riegel et al. CPR/AED SKILL RETENTION Table 2 Characteristics of the Volunteers by Retraining Status* Age of volunteers mean SD Volunteers with No Retraining Excluded from Analysis Not Not Dropped/ Did Not <3 or >17 Interim Protocol Offered Eligible Moved Attend Total Months Training Problem Analyzed (n = 995) (n = 623) (n = 1,518) (n = 5,105) (n = 8,241) (n = 1,577) (n = 1,381) (n = 1,186) (n = 6,182) yr yr yr yr yr yr yr yr yr p- valuey Male gender 50.1% 52.8% 55.7% 55.3% 54.5% 57.6% 52.0% 55.0% 53.4% Minority 18.7% 23.7% 25.5% 22.5% 22.7% 19.8% 16.0% 21.4% 19.6% Married 55.4% 46.8% 41.1% 51.2% 49.5% 60.0% 57.6% 61.4% 59.4% <.001 Prior 40.9% 43.8% 46.2% 47.0% 45.9% 51.0% 61.1% 48.9% 49.0% emergency experience Prior advanced 7.0% 10.5% 11.6% 10.5% 10.2% 11.6% 21.1% 13.6% 8.2% <0.001 training High school 34.0% 33.1% 29.6% 26.1% 28.3% 26.8% 23.4% 30.9% 27.4% education or less Previous CPR 44.8% 53.1% 57.6% 52.3% 52.4% 59.9% 73.1% 50.9% 54.7% training Native English 95.4% 90.8% 92.4% 93.6% 93.4% 94.4% 96.4% 97.0% 93.6% speaker Passed initial training without extra help 97.4% 98.9% 97.0% 98.4% 98.1% 99.2% 98.8% 99.5% 98.4% CPR = cardiopulmonary resuscitation. * Excludes 753 volunteers (397 CPR only, 356 CPR + automated external defibrillator) with retraining status unavailable. y p-values are for comparison of analyzed group vs. the excluded group (total). <0.001 six to 11, and 12 to 17 months, respectively, after initial training. Volunteers tended to be slightly older in the later groups and to have less prior CPR training. Linear declines over time were detected in individual CPR and AED skill performance and in the instructors global assessments. Despite these declines, however, even in the latest retraining group, a large majority (70%) of volunteers was judged to have adequate CPR skills. Overall, higher proportions of volunteers were judged to be adequate in AED skills than in CPR skills; however, relative differences over time were similar to those for CPR. Among the 1,504 volunteers judged unable to perform adequate CPR, more than 50% failed the individual ventilation, hand placement, and/or compression depth checklist skills. For those judged to have adequate CPR skills, the failure rate was lower than 20% on all individual skills. Among volunteers judged to have inadequate AED skills, 90% failed correct electrode placement. The ln (months since initial training) term was significantly associated with the global CPR assessment (p = 0.001). These associations did not differ significantly between the AED+CPR and CPR-only groups (p = for interaction term). The retraining interval term was not statistically significant in the global AED assessment model (p = 0.162). Mean predicted probabilities associated with adequate CPR and AED skills for each of the three retraining period groups are shown in Table 3. Results are similar to those based on the (unadjusted) global assessment proportions. Table 3 Mean Predicted Probabilities of Performing Adequate CPR and AED Skills 3 5 Months 6 11 Months Months p-value* All volunteers Number (total = 6,182) 2,839 2, Mean predicted probability, adequate CPR % 76.1% 70.4% <0.001 SEM 0.4% 0.5% 0.9% AED+CPR group Number (total = 3,756) 1,717 1, Mean predicted probability, adequate AED % 87.0% 87.1% <0.001 SEM 0.3% 0.4% 0.5% CPR = cardiopulmonary resuscitation; AED = automated external defibrillation; SEM = standard error of the mean. * Analysis of variance.

6 ACAD EMERG MED March 2006, Vol. 13, No Table 4 Characteristics of the Volunteers by Retraining Period 3 5 Months 6 11 Months Months p-value* All volunteers Number (total = 6,182) 2,839 2, Age at the time of initial training mean (SD) 41 (14) yr 42 (14) yr 43 (15) yr <0.001 Male gender 54% 52% 57% High school education or less 27% 27% 30% Previous CPR class (last five years) 57% 55% 46% <0.001 Passed initial training without extra help 99% 98% 98% Minority 21% 19% 18% Married 58% 61% 60% Prior emergency experience 51% 47% 46% Prior advanced training 10% 8% 6% Native English speaker 94% 93% 94% Pretest CPR adequate (by global assessment, all) 79% 73% 71% <0.001 Assess responsiveness 89% 87% 84% <0.001 Access % 80% 78% <0.001 Adequate ventilation/chest rise 82% 76% 70% <0.001 Correct hand placement 81% 78% 77% Adequate compression depth 87% 85% 83% AED+CPR group Number (total = 3,756) 1,717 1, Pretest AED adequate (by global assessment) 91% 86% 84% <0.001 Bare chest 96% 94% 89% <0.001 Correct electrode placement 87% 78% 78% <0.001 Clear self 85% 72% 79% <0.001 Verbally clear area 78% 68% 64% <0.001 Deliver shock within 90 seconds (n = 3,715) 81% 74% 70% <0.001 AED arrival after time zero* (n = 3,369) 84% 77% 72% AED arrival at time zero* (n = 227) 61% 35% 47% Deliver shock (no time limit) (n = 3,715) 97% 94% 91% <0.001 Time from AED arrival to shock (n = 3,522) 68 (29) sec 71 (35) sec 74 (41) sec mean ( SD) AED arrival after time zero (n = 3,238)y 66 (27) sec 68 (30) sec 72 (38) sec <0.001 AED arrival at time zero (n = 219)y 93 (38) sec 139 (70) sec 119 (73) sec <0.001 Time from AED arrival to shock, for shock 115 (27) sec 122 (37) sec 129 (48) sec delivered >90 sec (n = 669) mean (SD) [range] [91, 288] [91, 302] [91, 375] Test/retrain time, CPR only (n = 2,142) 5.6 (3.5) min 5.5 (4.0) min 7.3 (5.1) min <0.001 mean (SD) Test/retrain time, CPR+AED (n = 3,346) mean (SD) 6.8 (4.1) min 8.3 (5.0) min 9.1 (4.7) min <0.001 CPR = cardiopulmonary resuscitation; AED = automated external defibrillation. * Chi-square (categorical measures) or analysis of variance (continuous measures) test for trend. y During testing, most volunteers in the CPR+AED arm performed all CPR steps while the AED was brought to the manikin. However, these volunteers responded to the test with an AED by performing only the assessment and ventilation steps before applying the AED, resulting in artificially longer AED-arrival-to-shock times. Testing/retraining times tended to be longer for volunteers trained in later intervals, but the differences were only by about 1 to 2 minutes per volunteer, on average, for both the CPR-only and CPR+AED groups (Table 4). DISCUSSION The implementation of effective volunteer layperson response teams in public locations requires that a sufficient level of proficient volunteers be available to respond to emergencies, without incurring unsupportable costs (e.g., ongoing training program costs and volunteer time). The PAD Trial is the largest study of CPR and AED skill retention in lay volunteers to date. In this study, 79% of volunteer laypersons evaluated between three and five months after initial training were judged, in the opinion of the instructors, to be able to provide perfusing CPR, compared with 71% of those retrained between 12 and 17 months after training. A similar decrement was seen over time in the ability to provide a defibrillation shock (91% vs. 84%). If subsequent studies confirm that layperson responders can retain essential resuscitation skills for a year or more, the feasibility of implementing such programs on a broader scale is greatly enhanced. These results differ from most studies of resuscitation skill retention, which generally report that CPR skills deteriorate considerably over a short time period. Moser et al. studied CPR skill retention in 31 family members of cardiac patients. Of those tested at seven months, more than half were rated poor in initial assessment, chest compression, ventilation, and overall CPR. 22 Others

7 260 Riegel et al. CPR/AED SKILL RETENTION studying CPR skill retention in parents of infants at risk for cardiac arrest 23,24 and CPR skill retention in lay volunteers also have reported poor skill retention over time. Discrepancies between these results and those of other studies are likely related to a number of factors. First, the primary outcome measure for this study was the instructors global assessment of overall performance rather than the more traditional composite summary score based on the proportion of individual skills performed correctly. These traditional measures weigh all skill components equally for example, the skill of calling is considered as important as correct hand placement or compression depth. In addition, to be counted as correct, each skill must be performed in a specified order. Using this approach, it is likely that some lay responders who would perform well enough to provide perfusion and/or shock during a cardiac arrest would receive low scores on a skill test. The use of the global assessment was specified (a priori) as an attempt to overcome the unnecessary rigidity inherent in this traditional approach. Birnbaum et al. 41 evaluated the relationship between the PAD instructors global assessments and their individual skill assessments, and reported that volunteers who performed well on the ventilations, hand placement, and compression depth skills tended to be rated as having adequate CPR skills by instructors, and that the assessing responsiveness and calling skills had little influence on the global measure. For AED skills, pad placement was the primary factor influencing the instructors global assessments in the Birnbaum et al. study. While actions such as calling are important in an actual cardiac arrest situation, it is unclear how meaningful their performance (or lack thereof) is in an artificial testing environment. During testing of PAD volunteers, roughly 20% of volunteers failed to call 9-1-1; however, there were no reported instances of failure to call during the 3,413 actual emergency events. Furthermore, up to 36% of volunteers failed to clear the area verbally during testing, but there were no reported instances of volunteers or bystanders receiving a shock during the trial (approximately 115 shocks were delivered during the PAD Trial). Further evaluation of what comprises essential skills is needed. A second possible reason for the discrepancy between these results and those of others is differences in the populations studied. Volunteers in the PAD Trial agreed to be part of the emergency response team and to attend training and retraining sessions, and thus may have been more highly motivated to learn and retain emergency skills than other layperson populations studied. Tweed et al. reported excellent skill retention in motivated, mature layperson rescuers. 43 Furthermore, more than half of the PAD Trial volunteers had received CPR training prior to study participation. Many previous studies have focused on skill retention among spouses or parents of individuals at risk of cardiac arrest. These populations are generally older than the PAD Trial population, and their performance may be adversely affected by emotional responses to circumstances requiring the resuscitation of a loved one. A third potential explanation for differing results could be that CPR course content and techniques of training and evaluation are constantly evolving. 44 Since 1998, the AHA HeartSaver AED course has deemphasized didactic lectures in favor of video-based learning and substantial amounts of hands-on practice time using case-based scenarios. In the PAD Trial, initial training methods closely followed the AHA HeartSaver AED course, allowing significant time for skill practice, while skills such as foreign-body airway obstruction, infant/ child resuscitation, pulse check, and a written test were not required. It is possible that minimizing content and complexity in resuscitation skill training leads to better skill retention. Clinical resuscitation standards are evolving away from the previously held notion that effective CPR requires high precision in the performance of complex tasks. Researchers have reported that even imperfect bystander CPR is associated with increased survival from cardiac arrest. 3 5 Not surprisingly, in the PAD Trial, among individual CPR skills, ventilations producing chest rise appeared to be the most difficult skill to retain. Hallstrom et al. 45,46 found no difference in survival between patients whose rescuers were instructed to deliver chest compressions alone and those whose rescuers were instructed to deliver standard CPR. Using a simulated CPR model, Woollard et al. 21 found that ventilations were generally ineffective and more compressions were delivered in a given time period when ventilation instructions were omitted. If ventilation training were deemphasized, further emphasis could be applied to teaching adequate chest compression technique. Perhaps a more simplified approach to CPR training and evaluation is warranted and could improve outcomes. Retention of AED skills was higher than retention of CPR skills in all groups, possibly because AED skills require less psychomotor coordination, and because the AED provides both audible and visual action prompts. Interestingly, among volunteers assessed as inadequate in overall AED skills, placing electrodes correctly, a skill not specifically addressed by verbal prompts, had the highest failure rate. The AED skills with the largest differences in adequate performance across retraining interval were the clearing actions and the ability to shock within 90 seconds of the AED arrival. Cummins et al. also reported an increase in time to deliver the first AED shock as the duration from initial training increased. 35 Nevertheless, in this study, the vast majority of volunteers functioned adequately, through 17 months after initial training. Another significant observation of this study was the short time required to retrain volunteers back to proficiency after they had inadequately performed part or all of the individual CPR/AED skills during testing. This finding has important implications for organizations evaluating the costs of implementing or enhancing an emergency response program. Short, one-on-one retraining sessions, in which training is tailored to specific needs, may prove to be at least as effective as, and more efficient than, requiring a four-hour refresher course on a routine basis. Based on the PAD Trial approach, approximately ten individuals could be comfortably evaluated and retrained by one instructor during a two-hour time period. Before this approach is widely adopted,

8 ACAD EMERG MED March 2006, Vol. 13, No standard performance assessment and retention criteria need to be developed and validated. LIMITATIONS There are several significant limitations to this study. First, the global assessment of performance, the primary outcome measure, is new and has not been previously validated. Because of the large numbers of instructors and volunteers in this study, we were unable to assess instructor interrater reliability. Furthermore, few instructors will have participated in an actual cardiac arrest, and their ability to determine whether a trainee has demonstrated the ability to provide perfusing CPR and/or an adequate shock is open to question. 27,47 Unfortunately, as is the case with most training studies, very little is known about how classroom content, instruction, and evaluation of performance in a simulated environment are associated with performance during an actual cardiac arrest. 48 Another potential limitation is that the lengths of training differed in the two study arms because of the addition of AED content. Instructors also may have been biased toward assigning positive outcomes and shorter retraining times in order to have high pass rates and rapid retraining times. It is also possible that they measured their testing/retraining intervals inaccurately with the stopwatches. However, even if times were measured and/or recorded inaccurately by some, the total time required for the procedure would not have changed drastically. Only 53% of volunteers who were offered retraining participated (32% of all volunteers initially trained), and a selection bias could exist. Although measured characteristics did not differ greatly between volunteers who were trained and those who were not, volunteers who were retrained tended to be older, were more likely to be married, were less likely to be white, and were less likely to have had previous advanced training. It is possible that these volunteers were less confident in their skills than those who did not return for retraining. They also may have been more motivated, as discussed above. If participating volunteers were more highly motivated, they may have performed better than a typical trainee. Another limitation is the potential for confounding; volunteers tested and retrained in later groups were less likely to have had CPR training prior to the PAD Trial. Although a sizable portion of units was multiunit residential communities such as apartment complexes, volunteers were part of an organized response team; these findings are not relevant to the home environment where family members are trained to respond. Volunteer attrition over time may have caused additional unmeasured biases. We have attempted to control or adjust for potential biases where possible; however, the cumulative impact is unknown and the generalizability of the results is limited. Rather, these results should provide a backdrop for the design of future studies to: 1) evaluate validity and reliability of outcome measures focusing on the performance and retention of essential skills by laypersons; 2) evaluate the effectiveness of short, one-on-one retraining; and, most importantly, 3) relate classroom performance and evaluation to survival in the field. CONCLUSIONS Among PAD Trial volunteer laypersons participating in a simulated resuscitation, the proportions of volunteers judged by instructors to have adequate CPR and AED skills demonstrated small declines associated with longer intervals between initial training and subsequent testing. However, based on instructors judgment, large majorities of volunteers still retained both CPR and AED core skills up through 17 months after initial training. Furthermore, individual testing and retraining for CPR and AED skills were usually accomplished in less than 10 minutes per volunteer. Additional research is essential to identify training and evaluation techniques that predict adequate CPR and AED skills performance of laypersons during an actual arrest. References 1. American Heart Association. Heart Disease and Stroke Statistics 2004 Update. Dallas, TX: American Heart Association, Weaver WD, Cobb LA, Hallstrom AP, Fahrenbruch C, Copass MK, Ray R. Factors influencing survival after out-of-hospital cardiac arrest. J Am Coll Cardiol. 1986; 7: Cummins RO, Ornato JP, Thies WH, Pepe PE. Improving survival from sudden cardiac arrest: the chain of survival concept. A statement for health professionals from the Advanced Cardiac Life Support Subcommittee and the Emergency Cardiac Care Committee, American Heart Association. Circulation. 1991; 83: Van Hoeyweghen RJ, Bossaert LL, Mullie A, et al. Quality and efficiency of bystander CPR. Belgian Cerebral Resuscitation Study Group. Resuscitation. 1993; 26: Cobb LA, Hallstrom AP. Community-based cardiopulmonary resuscitation: what have we learned? Ann N Y Acad Sci. 1982; 382: Cummins RO, Eisenberg MS. Prehospital cardiopulmonary resuscitation. Is it effective? JAMA. 1985; 253: Lund I, Skulberg A. Cardiopulmonary resuscitation by lay people. Lancet. 1976; 2: Wik L, Steen PA, Bircher NG. Quality of bystander cardiopulmonary resuscitation influences outcome after prehospital cardiac arrest. Resuscitation. 1994; 28: Gallagher EJ, Lombardi G, Gennis P. Effectiveness of bystander cardiopulmonary resuscitation and survival following out-of-hospital cardiac arrest. JAMA. 1995; 274: Bossaert L, Van Hoeyweghen R. Bystander cardiopulmonary resuscitation (CPR) in out-of-hospital cardiac arrest. The Cerebral Resuscitation Study Group. Resuscitation. 1989; 17(suppl 17):S55 69; discussion S Nichol G, Detsky AS, Stiell IG, O Rourke K, Wells G, Laupacis A. Effectiveness of emergency medical services for victims of out-of-hospital cardiac arrest: a metaanalysis. Ann Emerg Med. 1996; 27:

9 262 Riegel et al. CPR/AED SKILL RETENTION 12. Naughton MJ, Luepker RV, Sprafka JM, McGovern PG, Burke GL. Community trends in CPR training and use: the Minnesota Heart Survey. Ann Emerg Med. 1992; 21: Valenzuela TD, Roe DJ, Cretin S, Spaite DW, Larsen MP. Estimating effectiveness of cardiac arrest interventions: a logistic regression survival model. Circulation. 1997; 96: Becker LB, Ostrander MP, Barrett J, Kondos GT. Outcome of CPR in a large metropolitan area where are the survivors? Ann Emerg Med. 1991; 20: Herlitz J, Bahr J, Fischer M, Kuisma M, Lexow K, Thorgeirsson G. Resuscitation in Europe: a tale of five European regions. Resuscitation. 1999; 41: Kuisma M, Jaara K. Unwitnessed out-of-hospital cardiac arrest: is resuscitation worthwhile? Ann Emerg Med. 1997; 30: Kern KB, Sanders AB, Raife J, et al. A study of chest compression rates during cardiopulmonary resuscitation in humans. The importance of rate-directed chest compressions. Arch Intern Med. 1992; 152: Todd KH, Braslow A, Brennan RT, et al. Randomized, controlled trial of video self-instruction versus traditional CPR training. Ann Emerg Med. 1998; 31: Chamberlain D, Smith A, Colquhoun M, Handley AJ, Kern KB, Woollard M. Randomised controlled trials of staged teaching for basic life support: 2. Comparison of CPR performance and skill retention using either staged instruction or conventional training. Resuscitation. 2001; 50: Ward P, Johnson LA, Mulligan NW, Ward MC, Jones DL. Improving cardiopulmonary resuscitation skills retention: effect of two checklists designed to prompt correct performance. Resuscitation. 1997; 34: Woollard M, Smith A, Whitfield R, et al. To blow or not to blow: a randomised controlled trial of compression-only and standard telephone CPR instructions in simulated cardiac arrest. Resuscitation. 2003; 59: Moser DK, Dracup K, Guzy PM, Taylor SE, Breu C. Cardiopulmonary resuscitation skills retention in family members of cardiac patients. Am J Emerg Med. 1990; 8: Wright S, Norton C, Kesten K. Retention of infant CPR instruction by parents. Pediatr Nurs. 1989; 15:37 41, Dracup K, Doering LV, Moser DK, Evangelista L. Retention and use of cardiopulmonary resuscitation skills in parents of infants at risk for cardiopulmonary arrest. Pediatr Nurs. 1998; 24:219 25; quiz Wilson E, Brooks B, Tweed WA. CPR skills retention of lay basic rescuers. Ann Emerg Med. 1983; 12: Chamberlain D, Smith A, Woollard M, et al. Trials of teaching methods in basic life support (3): comparison of simulated CPR performance after first training and at 6 months, with a note on the value of re-training. Resuscitation. 2002; 53: Brennan RT, Braslow A. Skill mastery in cardiopulmonary resuscitation training classes. Am J Emerg Med. 1995; 13: Gombeski WR Jr, Effron DM, Ramirez AG, Moore TJ. Impact on retention: comparison of two CPR training programs. Am J Public Health. 1982; 72: Nyman J, Sihvonen M. Cardiopulmonary resuscitation skills in nurses and nursing students. Resuscitation. 2000; 47: Gasco C, Avellanal M, Sanchez M. Cardiopulmonary resuscitation training for students of odontology: skills acquisition after two periods of learning. Resuscitation. 2000; 45: Berden HJ, Bierens JJ, Willems FF, Hendrick JM, Pijls NH, Knape JT. Resuscitation skills of lay public after recent training. Ann Emerg Med. 1994; 23: Morgan CL, Donnelly PD, Lester CA, Assar DH. Effectiveness of the BBC s 999 training roadshows on cardiopulmonary resuscitation: video performance of cohort of unforewarned participants at home six months afterwards. BMJ. 1996; 313: Donnelly P, Assar D, Lester C. A comparison of manikin CPR performance by lay persons trained in three variations of basic life support guidelines. Resuscitation. 2000; 45: Gundry JW, Comess KA, DeRook FA, Jorgenson D, Bardy GH. Comparison of naive sixth-grade children with trained professionals in the use of an automated external defibrillator. Circulation. 1999; 100: Cummins RO, Schubach JA, Litwin PE, Hearne TR. Training lay persons to use automatic external defibrillators: success of initial training and one-year retention of skills. Am J Emerg Med. 1989; 7: Meischke HW, Rea T, Eisenberg MS, Schaeffer SM, Kudenchuk P. Training seniors in the operation of an automated external defibrillator: a randomized trial comparing two training methods. Ann Emerg Med. 2001; 38: Meischke HW, Rea TD, Eisenberg MS, Rowe SM. Intentions to use an automated external defibrillator during a cardiac emergency among a group of seniors trained in its operation. Heart Lung. 2002; 31: Aufderheide TP, Stapleton ER, Hazinski MF (eds). Instructor s Manual, HeartSaver AED. Adult Cardiopulmonary Resuscitation and Automated External Defibrillation. Dallas, TX: American Heart Association, Ornato JP, McBurnie MA, Nichol G, et al. The Public Access Defibrillation (PAD) trial: study design and rationale. Resuscitation. 2003; 56: The Public Access Defibrillation Trial Investigators. Public-access defibrillation and survival after out-ofhospital cardiac arrest. N Engl J Med. 2004; 351: Birnbaum A, McBurnie MA, Powell J, et al., for the PAD Investigators. Modeling instructor preferences for CPR and AED competence estimation. Resuscitation. 2005; 64: Riegel B, Birnbaum A, Aufderheide TP, et al., and the PAD Investigators. Predictors of cardiopulmonary resuscitation and automated external defibrillator skill retention. Am Heart J. 2005; 150: Tweed WA, Wilson E, Isfeld B. Retention of cardiopulmonary resuscitation skills after initial overtraining. Crit Care Med. 1980; 8: Guidelines 2000 for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Part 4: the automated external defibrillator: key link in the chain of survival. The American Heart Association in

10 ACAD EMERG MED March 2006, Vol. 13, No Collaboration with the International Liaison Committee on Resuscitation. Circulation. 2000; 102(suppl 8): I Hallstrom A, Cobb L, Johnson E, Copass M. Cardiopulmonary resuscitation by chest compression alone or with mouth-to-mouth ventilation. N Engl J Med. 2000; 342: Hallstrom AP. Dispatcher-assisted phone cardiopulmonary resuscitation by chest compression alone or with mouth-to-mouth ventilation. Crit Care Med. 2000; 28(suppl 11):N Kaye W, Rallis SF, Mancini ME, et al. The problem of poor retention of cardiopulmonary resuscitation skills may lie with the instructor, not the learner or the curriculum. Resuscitation. 1991; 21: Aufderheide TP, Sigurdsson G, Pirrallo RG, et al. Hyperventilation-induced hypotension during cardiopulmonary resuscitation. Circulation. 2004; 109: d James E. Olson, PhD Department of Emergency Medicine Wright State University Dayton, Ohio (james.olson@wright.edu) Scholastic Sharecropping Positively influenced, we ve sown seeds from mentorship. We ve furrowed the academic pasture acquiring satisfaction And sense of identity with raising mosaic scholarships Of discovery, application, teaching and integration. Visible committees have granted a crop Full Professor In both hybrid clinical roles and tenure tracks. Longevity s exacted sufficiency for a few who ll endorse Altering the appellation to Emeritus in their almanac. Beware, those willing to redact a vitae aren t privy To the duration of tempered light that s been commanded! The Posthumous rank rests invisibly in the hands Of He, who promotes and shapes the landscape of eternity. Jonathan Singer, MD Department of Emergency Medicine Wright State University Dayton, Ohio (jonathan.singer@wright.edu) doi: /j.aem