Adverse events associated with lay emergency response programs: The public access defibrillation trial experience

Resuscitation (2006) 70, 59—65

CLINICAL PAPER

Adverse events associated with lay emergency response programs: The public access defibrillation trial experience Mary Ann Peberdy a,∗, Lois Van Ottingham b, William J. Groh c, Jerris Hedges d, Thomas E. Terndrup e, Ronald G. Pirrallo f, N. Clay Mann g, Ruchir Sehra h , and the PAD Investigators1 a

Virginia Commonwealth University Health System, Box 908204, Richmond, VA 23298, USA Clinical Trial Center, University of Washington, Seattle, WA, USA c Indiana University School of Medicine, Indianapolis, IN, USA d Oregon Health and Science University, Portland, OR, USA e University of Alabama, Birmingham, AB, USA f Medical College of Wisconsin, Milwauke, WI, USA g University of Utah, School of Medicine, Salt Lake City, UT, USA h Loma Linda Medical Center, Loma Linda, CA, USA b

Received 25 July 2005 ; received in revised form 27 October 2005; accepted 27 October 2005 KEYWORDS Cardiac arrest; Automated external defibrillator; Defibrillation; Cardiopulmonary Resuscitation; Emergency treatment

∗ 1

Summary The adverse event (AE) profile of lay volunteer CPR and public access defibrillation (PAD) programs is unknown. We undertook to investigate the frequency, severity, and type of AE’s occurring in widespread PAD implementation. Design: A randomized-controlled clinical trial. Setting: One thousand two hundred and sixty public and residential facilities in the US and Canada. Participants: On-site, volunteer, lay personnel trained in CPR only compared to CPR plus automated external defibrillators (AEDs). Intervention: Persons experiencing possible cardiac arrest receiving lay volunteer first response with CPR + AED compared with CPR alone. Main outcome measure: An AE is defined as an event of significance that caused, or had the potential to cause, harm to a patient or volunteer, or a criminal act. AE data were collected prospectively.

Corresponding author. Tel.: +1 804 828 4571; fax: +1 804 828 7710. E-mail address: [email protected] (M.A. Peberdy). The PAD Trial investigators and coordinators are listed in Ref. 15.

0300-9572/$ — see front matter © 2005 Published by Elsevier Ireland Ltd. doi:10.1016/j.resuscitation.2005.10.030

60

M.A. Peberdy et al. Results: Twenty thousand three hundred and ninety six lay volunteers were trained in either CPR or CPR + AED. One thousand seven hundred and sixteen AEDs were placed in units randomized to the AED arm. There were 26,389 exposure months. Only 36 AE’s were reported. There were two patient-related AEs: both patients experienced rib fractures. There were seven volunteer-related AE’s: one had a muscle pull, four experienced significant emotional distress and two reported pressure by their employee to participate. There were 27 AED-related AEs: 17 episodes of theft involving 20 devices, three involved AEDs that were placed in locations inaccessible to the volunteer, four AEDs had mechanical problems not affecting patient safety, and three devices were improperly maintained by the facility. There were no inappropriate shocks and no failures to shock when indicated (95% upper bound for probability of inappropriate shock or failure to shock = 0.0012). Conclusions: AED use following widespread training of lay-persons in CPR and AED is generally safe for the volunteer and the patient. Lay volunteers may report significant, usually transient, emotional stress following response to a potential cardiac arrest. Within the context of this prospective, randomized multi-center study, AEDs have an exceptionally high safety profile when used by trained lay responders. © 2005 Published by Elsevier Ireland Ltd.

Introduction Sudden, unexpected cardiac arrest is a leading cause of death and disability. Studies estimate nearly a quarter to a half a million episodes of out-of-hospital cardiac arrest occur annually in the US.1—3 Survival to hospital discharge is less than 5% in most cities, and has been reported as low as 1—2% in large metropolitan areas, despite ongoing scientific advances in resuscitation practice.4,5 Early defibrillation is the single most important intervention for improving survival from adult OOH-CA.6 The concept of early defibrillation by medical and, more recently, lay first-responders using automated external defibrillators (AEDs) is almost a decade old.7 Many communities have integrated first responder early defibrillation programs into their emergency medical services (EMS) systems, fire and police first response, and have targeted large venue public places such as airports and casinos using trained flight attendants and security guards.8—12 The success of these programs in improving time to defibrillation, return of spontaneous circulation, and overall survival has prompted more widespread implementation of early defibrillation by trained lay responders in a variety of public settings. Weisfeldt and Becker13 described three phases of resuscitation after cardiac arrest. The electrical phase, when defibrillation is most beneficial, only lasts for approximately 4 min. It is impractical for a typical EMS system to be activated quickly enough to reach the victim’s side and provide defibrillation to the vast majority of patients in this timeframe. Public access defibrillation (PAD) offers a way for lay responders to provide early defibrilla-

tion using AEDs while awaiting traditional EMS personnel. Many high volume public settings, such as casinos and airports, have demonstrated improved survival when non-medical employees are trained in CPR and AED use.9,11,12 Because of the small size of these earlier case series, safety data related to the use of AEDs by trained lay responders are limited. The purpose of this study was to determine the frequency, type, and severity of adverse events in patients, volunteers, and AED devices occurring in the large, prospective, multi-center public access defibrillation trial.

Methods The PAD trial was a prospective, multi-center, randomized clinical trial which demonstrated that volunteer, non-medical responders can improve survival from OOH-CA by using AEDs in addition to performing CPR and calling 911. Twenty-four research centers in the United States and Canada recruited 1260 individual public and residential facilities, which were combined to form a total of 993 randomized ‘‘community units’’ meeting specified cardiac arrest risk criteria. On-site lay volunteers without a contractual duty to act in a medical emergency participated as first responders. Facilities were randomized to have volunteer responders trained in CPR alone (CPR-only) or CPR with AED (CPR + AED). The sites randomized to the AED intervention arm received an adequate number of AEDs to respond to all areas in the facility within 3 min. All participating facilities received the standard emergency medical services (EMS)

Adverse events associated with lay emergency response programs care typically provided in their community. The study protocol was approved by the Institutional Review Board of the University of Washington as well as the Boards at each participating site. Detailed methods and results of the PAD trial have been published previously.14,15 All lay volunteers were trained in CPR or CPR + AED using American Heart Association (AHA), American Red Cross, Canadian Heart and Stroke Foundation or similar programs that closely followed the AHA guidelines. All volunteers were required to demonstrate proficiency in assessing unresponsiveness, accessing 911, providing ventilation with chest rise, correct hand placement, and adequate chest compression depth. Volunteers trained to operate the AED were also required to demonstrate proficiency in baring the chest, placing and attaching the defibrillation pads correctly, clearing themselves and others, and delivering a shock safely within 90 s. An emergency episode was defined as any of the following: EMS dispatch to the facility for unconsciousness, any activation of the volunteer system for a presumed cardiac arrest, attempted CPR, any shock delivered within unit boundaries, and any death within unit boundaries. An adverse event was defined prospectively as any event that caused, or had the potential to cause, harm to a patient or volunteer, or the commission of a criminal act. Collection of possible adverse events occurred by several systematic mechanisms throughout the trial. Volunteers were instructed during their initial training to report any adverse event, problem, or unusual occurrence to the study coordinator or investigator. Coordinators and investigators kept in close contact with volunteers by regular phone contact, faxes, emails and personal site visits to elicit information on system-wide issues or problems. Participating facilities were required to check AEDs monthly and report their findings to the study coordinator by either fax or verbal communication. Volunteer responders were debriefed after each episode in order to gather data relevant to potential adverse effects. Finally, patient EMS and hospital records were reviewed carefully for any possible adverse events. If unclear whether an incident qualified as an AE, the coordinator or investigator communicated with the PAD Coordinating Center for assistance in making the determination. An AE subcommittee, consisting of study investigators and coordinators also reviewed incidents for determination of AE’s. All AE data were collected as part of the IRB approved study design and were also reviewed by the study’s data safety monitoring board.

61

Adverse events were collected under three main categories: patient, volunteer, and AED device related. Patient related adverse event categories include rib fractures, liver laceration, pneumothorax, head trauma, neck and back trauma, and resuscitation attempts despite the patient having a do not attempt resuscitation (DNAR) status. Postmortem examinations were not required. Surviving patients were not evaluated for the presence of trauma by study staff. All emergency department and hospital records were reviewed for potential patient AEs but, if not noted in the records, an AE would remain unknown. These findings were noted to be present only if documented at some point in the patient record. Patient related adverse events were collected for both CPR-only trained lay responders and CPR + AED trained lay responders to determine if such events were associated more frequently with resuscitation performed with AEDs. Volunteer related adverse event categories included the occurrence of any physical injury or emotional trauma. Our protocol required the clinical centers to conduct a volunteer debriefing after each event. The volunteer was asked to rate his or her stress level on a scale of 0—5 immediately after the event. They received follow-up if the score suggested moderate to severe stress defined as scale 4—5. If the volunteer experienced loss of sleep, physical symptoms, or difficulty with daily activities due to the event that did not resolve within a brief period, the site provided or referred the volunteer for psychological follow-up. Data were also collected on whether or not the volunteer responders could feel the AED shock being delivered to the patient. AED device-related adverse event categories include any device malfunction, including shocks delivered to a non-arrest patient, AED use by a nontrained person, a criminal or inappropriate use of the device, whether or not retrieval of the device caused a substantial delay in accessing EMS, and AED theft. An independent committee reviewed all AED recordings for rhythm determination and appropriateness of shocks. Ninety-five percent confidence intervals were computed for adverse event rates based on exposure years. For the case where no adverse events were observed a 95% upper bound was computed.

Results A total of 20,396 lay volunteer responders were trained in CPR alone or CPR plus the use of an AED. One thousand seven hundred and sixteen AEDs were

62 placed in public and multi-family residential facilities. There were 26,389 months of public availability evaluated for adverse events during the duration of the trial. Of the 3952 emergency episodes captured during the trial, 649 were presumed cardiac arrests. Data on adverse events (AEs) were collected from 20 July 2000 through 30 September 2003. Thirty-six AEs were confirmed.

Patient-related AEs There were two patient-related AEs (95% CI per exposure-year = [0.0, 0.0022]) consisting of two patients with rib fractures discovered on autopsy after both volunteer and EMS CPR. No patient with a known DNAR order had resuscitation started inappropriately. No patient-related adverse event was related to the use of an AED.

Volunteer related AEs There were seven (95% CI per exposure-year = [0.0008, 0.0055]) volunteer related adverse events consisting of one volunteer who developed a muscle pull after responding to an emergency, four volunteers with increased emotional stress levels requiring intervention, and two volunteers who felt pressured by their employer to participate as a first responder. The majority of rescuers with initially elevated stress levels returned to normal by the following day but four required further follow up or interventions. One volunteer was tearful, nauseated, and had difficulty sleeping after performing CPR on a victim that was well known to them. This volunteer was upset at seeing the patient cyanotic and was uncomfortable with being labelled a hero, however had complete resolution of their symptoms in 2 days and continued to participate as a volunteer first responder. Another volunteer responded to a traumatic injury to a child. This volunteer had difficulty in school after the event, was referred to a counsellor and was subsequently reported to be ‘‘all right’’ by a family member after counseling. A third volunteer experienced nausea and vomiting for 20 min after the episode and rated their stress level as a four out of five. The stress resolved by the following day. Finally, a fourth volunteer was involved in two episodes, and knew both of the victims. This volunteer reported a stress level was a ‘‘five’’ after the second event and additional follow-up with study personnel was refused. The volunteer left their job at the unit, reportedly due to a number of reasons not solely related to the arrests. No volunteer was harmed by the use of an AED. The AED was retrieved from its storage site in 690

M.A. Peberdy et al. episodes and placed on 128 patients. A shock was delivered to 58 patients. The AED was never operated by an untrained person.

Device-related AEs There were 27 (95% CI per exposure year = [0.0076, 0.0169]) AED device-related AE’s. There were seventeen incidents of theft involving 20 devices. On three occasions, the AEDs were moved to locations not readily accessible to the first responder. There were four incidents of mechanical difficulty or battery failure that did not affect patient safety. One AED developed a failed circuit board, another displayed a service message, a third showed an electrode warning and the fourth was found to have a dead battery that was not indicated by the AED. Three AEDs were maintained inappropriately by the facility. Two were found to have the battery ajar and the third involved an episode where the lid to the AED box was closed over the electrode wires making it difficult to re-open the device. None of these episodes affected patient safety because another AED was available promptly in each setting. During this study, there were no inappropriate shocks and no device failed to shock when indicated (95% upper bound for probability of inappropriate shock or failure to shock = 0.0012).

Discussion Survival from OOH-CA is poor, yet when sudden death occurs, if defibrillation and CPR are available immediately (e.g., in cardiac rehabilitation or in the electrophysiology laboratory), survival has the potential to be greater than 90%.6 The opportunity for significant public benefit from increased survival is huge considering the large difference between what is theoretically possible and what actually occurs in our communities. The results of the PAD trial indicate that trained lay rescuers can use modern generation shock-advisory AEDs effectively and safely. There were no severe adverse events detected in a carefully monitored system of significant size in North American communities. A similar lack of problems has been noted in other published series. Valenzuela et al. trained security officers in CPR and AED use and significantly improved survival from cardiac arrest in numerous casinos without reporting any adverse events.12 Many communities have developed law enforcement agency AED programs and, similarly, have not reported any patient, device, or rescuer adverse events related to the provision of early

Adverse events associated with lay emergency response programs defibrillation.8,10,16—21 Two United States airline companies have published their experience with training flight attendants to use AEDs with an exceptional safety record.11,22 Caffrey et al.’s9 report of experience in the Chicago airport authority was the first publication of AED use by untrained first responders. Even though this program was designed to have only security personnel respond, in 19 of 21 episodes of cardiac arrest, Good Samaritan lay responders used the AED without harm to themselves or the victim. Despite over 20 years of experience with lay responder AED use, the literature reporting adverse events is scarce. In the United States, the Food and Drug Administration (FDA) has the responsibility for assuring the safety and efficacy of all regulated marketed medical products. The FDA developed the Safety Information and Adverse Event Reporting Program (MedWatch) to provide health care professionals and consumers a mechanism to report serious problems that are associated with use of a drug or device. Although MedWatch provides vital safety information to the FDA, the database has a limited perspective related to devices such as AEDs. It collects only data specifically reported to the system and links the device with the outcome. Death and poor neurological outcomes in survivors are common outcomes from cardiac arrest, regardless of up to date treatment. This association can be interpreted erroneously as a cause and effect relationship, when in fact the AED may have nothing to do with the poor outcome. Data from the PAD trial provides a different perspective on AED safety than the FDA MedWatch program. The FDA’s medical device reporting system has no denominator. Thus, everything reported to the FDA must be considered to be a potential device problem without any balancing information on the extent to which the device contributed to an adverse patient outcome. The PAD trial experience provides an objective assessment of the true frequency of AED related adverse events as determined by rigorous, prospective surveillance techniques. Device problems in this trial were all relatively minor, never affecting the safety of the patient or lay responder. Rhythm recognition tests performed in the laboratory and in clinical trials in the mid 1980’s validated the accuracy of AEDs. Most device errors were due to a lack of arrhythmia recognition. These devices re-analyzed the rhythm every several seconds so the lack of appropriate detection on one analysis typically did not result in a markedly delayed shock delivery. Stults et al.23 reported two cases in 122 in which coarse VF was consistently not recognized as VF by the AED and resulted in

63

several minutes delay of defibrillation. Ornato et al.24 reported an inappropriate automated shock of a patient in sinus rhythm. In this instance, the EMS responders did not follow the device’s instructions and analyzed the rhythm in a moving vehicle. The motion artifact was misinterpreted as a shockable rhythm and two shocks were delivered resulting in ventricular tachycardia. The device then appropriately detected VT and restored sinus rhythm with a third shock. Reports such as these were all early devices that were unable to detect motion artifact and, in some cases, were fully automatic, not giving the responder control of shock delivery. Newer devices with advanced technology have nearly eliminated these types of adverse events. All AEDs used in the PAD trial performed selfchecks that were helpful in the early identification of potential problems. Since some of the device-related adverse events were secondary to improper maintenance or location selection, human factors clearly play a role in the ultimate success of these programs. Education on proper maintenance and accessibility is crucial for developing lay first responder AED programs. The most common adverse event identified in the PAD trial was AED theft. Interestingly, the majority of stolen devices were removed from locked locations as opposed to those kept unlocked and visible, suggesting that restricting access of AEDs to the general public may not be an important factor in preventing AED theft. The only patient related problems were related primarily to CPR. No patient was harmed by an AED during the trial. DNAR status is becoming increasingly frequent in the pre-hospital setting. A theoretical concern for PAD programs is that resuscitation may be started on those who do not want it. Assisted and independent living facilities for the elderly not only have an exceptionally high-risk population but also have a significant percentage of residents who do not want to be resuscitated should they suffer a cardiac arrest. The PAD trial demonstrated that it is possible to implement widespread lay responder early defibrillation programs in public and residential locations without unwanted resuscitations. Psychological issues represented the most significant adverse events in the volunteer responder group. A small number of volunteers developed stress levels severe and sustained enough to require intervention. Although they represent a very small fraction of all the volunteers who participated in the trial, the concept of psychological stress on the part of the first responder has not been reported previously. The PAD trial, unlike other reported case series of lay first responder

64 AED programs, specifically evaluated volunteer stress using a prospective surveillance system. The psychological stress seen in the PAD trial was indigenous to the entire resuscitation response and was not specific to AED use. PAD programs should consider having procedures in place to identify and deal with responder stress levels based on these findings. Employers also need to be sensitive to an employee’s willingness to participate in such a program. Volunteers who feel pressured to participate by their employer may interpret this erroneously as a threat to their employment. The most obvious limitation of this trial was that it included only trained responders in a limited number of public and residential venues. Thus, it is not appropriate to extrapolate these low adverse event rates to PAD programs that might use untrained lay responders. Other limitations include the fact that adverse events were self-reported and, despite significant efforts to collect a comprehensive list of adversities, some data may have been missed. Volunteers and clinical coordinators may have been apprehensive or unwilling to report AEs, but they were encouraged and regularly questioned about reporting. There was variability in the amount of contact each investigator and coordinator had with her/his respective facilities. This may have affected either the occurrence or reporting of adverse events. Despite these limitations, AEDs appear to be safe devices in the hands of trained laypersons using them in settings similar to those used in the PAD trial. The Food and Drug Administration made a landmark decision in September 2004 in allowing a single manufacturer’s AED to be available for purchase by the public without a prescription. This was a crucial step in removing a potentially large barrier to the public’s access to AEDs. The results of this study are particularly relevant, given the increased availability of AEDs and the decrease in mandated physician supervision over PAD programs and lay responders that will result from ‘‘over the counter’’ availability of devices.

Conclusion Our data confirms the safety of AEDs in a culturally diverse, appropriately instructed, supervised lay responder population throughout widespread geographical areas in a variety of public and multifacility residential settings. Further research is needed to determine whether safety is similar for untrained lay responders and those not participating in supervised programs.

M.A. Peberdy et al.

Acknowledgements Funding support: supported by contract #N01— HC—95177 from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD. Additional support by: American Heart Association, Dallas, TX; Medtronic, Incorporated, Minneapolis, MN; Guidant Foundation, Indianapolis, IN; Cardiac Science/Survivalink, Incorporated, Minneapolis, MN; Medtronic PhysioControl Corporation, Redmond, WA; Philips Medical Systems, Heartstream, Seattle, WA; and Laerdal Medical Corporation, Wappingers Falls, NY.

References 1. Centers for Disease Control and Prevention (CDC). StateSpecific Mortality from Sudden Cardiac Death, United States, 1999, MMWR, 2002; 51(6):123—6. 2. Chugh SS, Jui J, Gunson K, et al. Current burden of sudden cardiac death: multiple source surveillance versus retrospective death certificate-based review in a large U.S. community. J Am Coll Cardiol Sep 15 2004;44(6):1268—75. 3. Rea TD, Eisenberg MS, Sinibaldi G, White RD. Incidence of EMS-treated out-of-hospital cardiac arrest in the United States. Resuscitation Oct 2004;63(1):17—24. 4. Lombardi G, Gallagher J, Gennis P. Outcome of out of hospital cardiac arrest in New York City. The Pre Hospital Arrest Survival Evaluation (PHASE) Study. JAMA 1994;271(9):678—83. 5. 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(4):355—61. 6. 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(5):1832—47. 7. Weisfeldt ML, Kerber RE, McGoldrick RP, et al. Public access defibrillation: A statement for healthcare professionals from the American Heart Association Task Force on Automatic External Defibrillation. Circulation 1995;92:2763. 8. White RD, Hankins DG, Bugliosi TF. Seven years’ experience with early defibrillation by police and paramedics in an emergency medical services system. Resuscitation Dec 1998;39(3):145—51. 9. Caffrey SL, Willoughby PJ, Pepe PE, Becker LB. Public use of automated external defibrillators. N Engl J Med 17 Oct 2002;347(16):1242—7. 10. Mosesso Jr VN, Newman MM, Ornato JP, et al. Law enforcement agency defibrillation (LEA-D). Proceedings of the National Center for Early Defibrillation Police AED Issues Forum. Prehosp Emerg Care, Jul to 6 Sep, 3 2002:273— 82. 11. Page RL, Joglar JA, Kowal RC, et al. Use of automated external defibrillators by a US airline. N Engl J Med 26 Oct 2000;343(17):1210—6. 12. Valenzuela TD, Roe DJ, Nichol G, Clark LL, Spaite DW, Hardman RG. Outcomes of rapid defibrillation by security officers after cardiac arrest in casinos. N Engl J Med 26 Oct 2000;343(17):1206—9.

Adverse events associated with lay emergency response programs 13. Weisfeldt ML, Becker LB. Resuscitation after cardiac arrest: a 3-phase time-sensitive model. JAMA 2002;288(23):3035—8. 14. Ornato JP, McBurnie MA, Nichol G, et al. The Public Access Defibrillation (PAD) trial: study design and rationale. Resuscitation Feb 2003;56(2):135—47. 15. The PAD Trial Investigators. Public-access defibrillation and survival after out-of-hospital cardiac arrest. N Engl J Med 2004;351(7):637—46. 16. Myerburg RJ, Fenster J, Velez M, et al. Impact of communitywide police car deployment of automated external defibrillators on survival from out-of-hospital cardiac arrest. Circulation 27 Aug 2002;106(9):1058—64. 17. Groh WJ, Lowe MR, Overgaard AD, Neal JM, Fishburn WC, Zipes DP. Attitudes of law enforcement officers regarding automated external defibrillators. Acad Emerg Med Jul 2002;9(7):751—3. 18. Ross P, Nolan J, Hill E, Dawson J, Whimster F, Skinner D. The use of AEDs by police officers in the City of London. Automated external defibrillators. Resuscitation Aug 2001;50(2):141—6. 19. Groh WJ, Newman MM, Beal PE, Fineberg NS, Zipes DP. Limited response to cardiac arrest by police equipped with

20.

21.

22. 23.

24.

65

automated external defibrillators: lack of survival benefit in suburban and rural Indiana, the police as responder automated defibrillation evaluation (PARADE). Acad Emerg Med Apr 2001;8(4):324—30. Davis EA, McCrorry J, Mosesso Jr VN. Institution of a police automated external defibrillation program: concepts and practice. Prehosp Emerg Care Jan to Mar 1999;3(1):60— 5. Mosesso Jr VN, Davis EA, Auble TE, Paris PM, Yealy DM. Use of automated external defibrillators by police officers for treatment of out-of-hospital cardiac arrest. Ann Emerg Med Aug 1998;32(2):200—7. O’Rourke MF, Donaldson E, Geddes JS. An airline cardiac arrest program. Circulation 1997;96(9):2849—53. Stults KR, Brown DD, Kerber RE. Efficacy of an automated external defibrillator in the management of out-of-hospital cardiac arrest: validation of the diagnostic algorithm and initial clinical experience in a rural environment. Circulation 1986;73(4):701—9. Ornato JP, Shipley J, Powell RG, Racht EM. Inappropriate electrical countershocks by an automated external defibrillator. Ann Emerg Med 1992;21(10):1278—82.