American Heart Association Report on the Public Access Defibrillation Conference December 8-10, 1994
- History of the Automatic External Defibrillation Task Force
- Current State of Emergency Medical Services
- Optimal Design and Application
- Field Experience With the AED
- Training Guidelines
- Public Policy and Legal Considerations
- Conference Participant Recommendations
- Info & Metrics
During the past 20 years, morbidity and mortality rates for nearly all types of cardiovascular disease have declined. Progress in these areas is in stark contrast to that for sudden cardiac death, which continues unabated at a rate of approximately 1000 times per day in the United States, with little decline in incidence or improved outcome. Clearly, the problem of sudden cardiac death is best approached through prevention, but horizons in that area seem no more promising and in some respects less promising and substantially more costly than 2 decades ago.
The means necessary for successful resuscitation of a patient in cardiac arrest were known by the early 1960s. Externally performed cardiopulmonary resuscitation (CPR) could maintain an “oxygen plateau” and delay permanent brain damage long enough to allow external defibrillation using direct current (DC). The possibility of long-term survival was increasingly recognized, as early anecdotal experiences accumulated into published series.1 2 3
Given the hindsight of 3 decades, the obstacles to be overcome before significant progress could be made in out-of-hospital resuscitation were formidable. First, cardiac arrest was perceived as an event that typically occurred in the hospital. In-hospital cardiac arrests are now recognized to represent only a small proportion of sudden deaths based in the community. Second, the CPR technique was known to only a limited number of hospital-based physicians. CPR is no longer restricted to hospitals or physicians; it is routinely taught to the lay public. Third, only line-powered, bulky, and awkward defibrillators were available. The first out-of-hospital defibrillation device weighed 110 lb. Contemporary external defibrillators are available that weigh less than 10 lb.
History of the Automatic External Defibrillation Task Force
The present report details progress made in achieving the goal of facilitating out-of-hospital resuscitation and specifies those areas in which further headway is needed. This effort began in 1990 with an American Heart Association (AHA) task force on the future of CPR, under the leadership of Dr Leonard Cobb.4 That task force considered the possibility of implementing an enhanced defibrillation strategy and challenged the medical manufacturing industry to produce automatic external defibrillators (AEDs) that were small, lightweight, easy to use, durable, maintenance free, and voice prompted. The task force called for an inexpensive, mass-marketed device that potentially could be implemented at a variety of sites within the community. By 1993, it was clear that the medical manufacturing industry would respond quickly to the AHA’s challenge. Thus, the AHA Board of Directors appointed the Automatic External Defibrillation Task Force, with responsibility for planning a national conference on public access defibrillation; evaluating the research needed to establish the feasibility of a safe, reliable, and effective AED for broad community use; and determining the role of the AHA in the effort to bring defibrillation more rapidly to any individual facing sudden death.
The conference, titled “Public Access Defibrillation: A New Strategy to Prevent Sudden Death,” was convened December 8 through 10, 1994, in Washington, DC. This conference attracted some 300 participants, with extensive representation from leaders within the scientific, industrial, medical, nursing, public health, and engineering communities; US federal government agencies (eg, the Food and Drug Administration [FDA] and the National Heart, Lung, and Blood Institute [NHLBI]); and the lay and professional press. The remainder of this report outlines the content of the sessions and the interactive workshops that constituted the formal program and the recommendations of the conference participants.
Current State of Emergency Medical Services
In cardiac arrest, virtually all survivors are experiencing ventricular fibrillation (VF) when paramedics or emergency medical technicians first arrive.5 6 7 8 9 10 11 Use of AEDs in out-of-hospital cardiac arrests has been encouraging. For many patients, sudden death is the first manifestation of underlying cardiac disease. Two strategies are being considered to combat cardiac arrest. In the first, AEDs are progressively integrated into the existing public safety system: emergency medical technicians (EMTs), police, firefighters, and other first responders who may have a duty to respond to a sudden death event are trained in the use of the AED. The second strategy involves the lay public as potential first responders in the event of cardiac arrest.
When placed in a public safety vehicle, the AED can be brought to the patient by a trained rescuer. Time to defibrillation is the major predictor of outcome.5 6 7 8 9 10 11 Therefore, making AEDs available to first responders should significantly improve survival rates. Clinical trials of AED use by EMTs have shown with few exceptions that this technology is safe and saves lives. However, estimates indicate that only 30% of emergency medical service (EMS) systems in the United States have implemented AED programs by equipping conventional ambulances with defibrillators.12 However, EMS systems are changing. Basic EMTs who previously performed only CPR are rapidly being replaced by “EMT-Ds”: EMTs who are equipped with and trained to use an AED. Efforts are under way to improve other aspects of the EMS system as well, such as expansion of the universal emergency telephone number (911) to uncovered areas of the United States. Studies of non-EMT first responders (eg, firefighters) also demonstrate trends toward improved survival rates. Given these early successes, further studies in training and implementation of AEDs by the general population on a first-responder protocol are indicated to assess their efficacy. Likely sites for placement of AEDs to be used by the public include office buildings, shopping malls, hotels, retirement communities, public buildings, apartment buildings, airports, and cruise ships. AEDs should be placed in sites where sudden death might be expected to occur and where there are no persons who can be identified as having a duty to act.
Research is needed to demonstrate the effectiveness of the use of AED technology by the general public before widespread distribution is sought. In cities where CPR training is widespread and EMS response is rapid, the survival rate was increased from 9% to 30% when AEDs were made available to first responders.6 In cities in which EMS response times are prolonged because of traffic congestion and high-rise buildings and in which bystander CPR is infrequent, public use of AEDs may be the only approach currently available to improve long-term survival outcome from its current rate of 1% to 2%.7
Once research has demonstrated effectiveness, the collaborative efforts of the FDA, the American Heart Association, and state regulatory agencies will be required to clear legal and regulatory obstacles through modification of existing laws and policies. The general public will need some training; how much is not clear. Safety issues require resolution as well. The technology must be developed to such a level that it would be extremely difficult to inadvertently or intentionally misuse an AED.
Optimal Design and Application
AED technology has produced safe, effective devices that can be used by a wide range of trained persons to convert ventricular tachyarrhythmias (VTs) and VFs to perfusing rhythms. From an engineering standpoint, these devices will continue to evolve as better waveforms, algorithms, and basic design modifications are introduced. Special applications of the devices, such as use in children, need to be considered.
A great deal of research has been done recently on waveforms used in automated implantable cardioverter defibrillators (AICDs). Damped sinusoidal and monophasic truncated exponential waveforms currently are used in AEDs. New investigational defibrillator waveforms, such as biphasic truncated exponential pulses or biphasic sinusoidal pulses, may terminate VF and VT effectively at lower energies.13 14 15
The continued search for better waveforms may be an important aspect of AED development.16 17 Waveforms used in the future must be at least as effective as waveforms now in use, but other parameters also should be considered. For example, if a new waveform reduces the cost of a device, makes the device easier to use, or has an engineering advantage, it may be a significant advance even if its ability to terminate VT and VF is no greater than that of standard waveforms.
There are two approaches to the search for better defibrillation waveforms: the empirical approach (testing the extremes of knowledge and technology) and a search for the basic mechanism of defibrillation at the cellular membrane level. Both approaches are worthy, but the latter may be preferable because it could lead to device designs that attack the basic mechanisms of fibrillation.
Many approaches exist for developing an effective algorithm to detect VF in digitized signals acquired by AEDs.16 The problem requires development of a numeric scale to separate shockable from nonshockable rhythms. The analysis strategy must capture essential differences in the electrical pattern or the rate of shockable and nonshockable rhythms. The engineering challenge is to devise descriptors that can be used to separate these two populations. Experience to date indicates that no single descriptor is adequate.16
Algorithms generally are either fully automatic (the AED is charged and the shock is delivered automatically) or semiautomatic or shock advisory (the operator must press buttons to analyze and fire the unit). Rhythm analysis is the same for both types of algorithms. The simplest AED algorithm is blind defibrillation. In the out-of-hospital setting, most pulseless rhythms are VF, pulseless electrical activity, or asystole. Because the outcomes of cardiac arrest due to pulseless electrical activity or asystole are dismal and VF can masquerade as asystole, it can be argued that it is reasonable from a public health standpoint to use defibrillation in every instance that appears to be clinical cardiac arrest. The principal drawbacks are ethical (ie, primum non nocere) and medicolegal (ie, delivering electric countershocks to individuals who do not need them).
Fortunately, electrocardiographic (ECG) waveform analysis, a valid, noninvasive indicator of VF, has been used effectively since the early 1980s. Early devices shocked rhythms with more than 150 complexes per minute and a peak amplitude greater than 0.15 mV. Subsequent algorithms use rate criteria (160 to 180 complexes per minute), morphological criteria (isoelectric content or probability density waveform), frequency of narrow QRS-like complexes with a high rise rate of more than 20 mV/s after low-pass filtering, and variability in the beat-to-beat interval.
Multidimensional analysis combines several independent descriptors to help separate shockable from nonshockable rhythms. The special challenge for engineers is to design improved AEDs that can accurately assess the rhythm in a short sampling time in a heterogeneous patient population and that perform better than or as well as trained paramedics.
The actual algorithms used in marketed devices are trade secrets. However, a common strategy is to establish a series of gates or hurdles through which the algorithm must pass. The rhythm being analyzed must pass through all the gates before the device can make a decision to recommend (or deliver) a shock. If the rhythm fails to make it through any of the gates, the device will not advise (or deliver) a shock, and it will begin a new analysis. There are many creative ways to address this challenge.
Before 1987, healthcare providers were generally better than AEDs at distinguishing shockable from nonshockable rhythms, but now most newer AEDs outperform trained providers. A recent study18 showed that the AEDs now in use are 90% sensitive for VF and 90% to 95% specific for other rhythms. There is little evidence to suggest that accurate detection of VF can be improved by using additional biosensors. The original HeartAid device used an oropharyngeal biosensor to measure impedance and detect respiration. In later models, these biosensors were eliminated because they reduced the sensitivity of the algorithm for detecting VF. Several proposed biosensors can detect non-VF rhythms, but, in general, specificity for non-VF rhythms has not been a problem. Signal conditioning is necessary in automatic VF detection devices because it helps reduce noise and baseline drift. Algorithms for detection of VF are more than adequate for widespread field use of AEDs. Refinement of algorithms can, will, and should occur during the course of product development and postmarketing surveillance.
Reliability testing of AEDs is facilitated by the use of automatic detection algorithms, which vary depending on the setting. For example, an AICD is subject to little environmental noise, whereas the AED algorithm is subject to a variety of complicating noise sources. An extensive tape library of treatable VT and VF rhythms is essential for development of effective algorithms. After development, a different tape library of treatable VT and VF rhythms is used for testing and verification to determine sensitivity. Further testing using real-noise recordings challenges that algorithm and improves specificity. Common sources of noise include that induced by motion, electrode manipulation noise, 60-Hz electromechanical failure, and baseline shift secondary to exercise. Probability density function enhances the reliability of such algorithms.
The most frequently reported problems in the FDA Medical Device Reporting Database for AEDs are nonclearable message prompts, device shutdown (inoperable), and inappropriate analysis.19 These problems are usually the result of improper application of electrodes; device maintenance; or rhythms, such as low-amplitude VFs that are undetected by existing algorithms. Corrective alternatives include minimizing high impedance during electrode application, electrode redesign, automatic internal-complement maintenance testing, and development of new or modified AED algorithms.
AED Design Modifications
Needs drive the design of new devices. The designer makes observations and uses old designs and assumptions to translate needs into design. Original needs in the 1960s centered on effective treatment of VF, so initial designs quickly evolved into the DC defibrillator. The underlying assumptions of these devices were formed in the hospital: It was assumed that these defibrillators would be used frequently; that they would be complicated and expensive; and that the requirements for skills, maintenance, and training would be demanding.
Since that time, in-hospital devices have been adapted for use in the field by modifying the hardware and software. We now realize the need to reduce the time to defibrillation if we are to improve survival rates; having more defibrillators available can decrease response time.
If widely distributed, public access AEDs would be used infrequently. For infrequent use, the design may change. A public access AED must be simple; easy to use; small, light, and rugged, with low maintenance requirements; and inexpensive to buy and maintain. Devices must self-test, indicate readiness, deter misuse and misapplication, and provide for data archiving and retrieval. No screen or printer is necessary. The device could be designed for single-patient use; however, a design for multipatient use is preferred. Multiple-use devices could be handled like that of laser printer toner cartridges: replacement cartridges are shipped with prepaid return-shipping labels for reconditioning, recycling, or proper disposal. This would also provide manufacturers with an opportunity to retrieve data and track use and performance of their devices. The medical community must be active in determining the optimal characteristics of AEDs. Standards for AED manufacturers have been developed by the Association for the Advancement of Medical Instrumentation, but the reliability and effective use of AEDs manufactured to this standard should be monitored under field conditions.
Recent work on optimal defibrillation techniques indicates that inappropriate electrode placement is among the most frequently encountered errors in AED use. Improper placement results in low interelectrode impedance. Proximity of electrodes causes the defibrillatory discharge to be conducted across the surface, rendering it ineffective.20 With properly placed electrodes, transcardiac current flow is maximized. Multiple high-energy shocks are associated with increased frequency of postcardioversion atrioventricular block and evidence of cardiac damage.21 22 Current-based defibrillation with an optimal range of 30 to 40 A may be more effective than an energy-based system, especially in patients with high transthoracic impedance. New waveforms may reduce these energy and current requirements.13 14
Manufacturing costs and sales prices are unlikely to impede rapid implementation of AEDs for EMT first responders, but use of AEDs in ambulances that provide basic life support will constitute only a moderate market. Significant price sensitivity is a barrier if AEDs are to be used more generally, like fire extinguishers, by bystanders. A very large market would be realized if AEDs were to become generally available to the public. However, medical device manufacturers lack experience in high-volume, low-cost production. Production and distribution costs can be reduced by achieving high volume and simplifying units intended for public use, but the extensive testing, recordkeeping, and medical device reports required by regulatory agencies add to the cost per unit. Implementation of these measures could reduce the price per unit by $1000 to $1500, but price-performance tradeoffs are inevitable.
Although sufficient funds have been committed since the beginning of the AHA initiative to make real progress in device development, the venture capital community has yet to perceive AED market opportunities to be sufficiently attractive to justify major investments. A reasonably large market may exist for AEDs, but potential investors perceive the market demand for AEDs as limited. To date, only 20 000 AEDs have been sold, despite years of marketing by vendors.
From 1987 to 1993, venture capital of approximately $7.4 billion was invested in 885 major healthcare companies. In contrast, this source funded only two start-up defibrillator companies, for $10 million total. Investments have also been limited from major producers in AED research and development. The few small companies now funding AED research and development together have spent less than $25 million, or less than the amount needed to fund the annual costs of testing a single medical device.
Notwithstanding these limited infusions of capital funds, the technology is available to produce a lightweight, safe, reliable, and effective AED in the price range of $2500 (compared with the current price range of $4000 to $10 000). Significant challenges to the manufacturers of AEDs include reducing the costs of regulation and postmarket testing (20% to 30% of the sales price) and of distribution and marketing (20% to 30% of the sales price). These costs contribute heavily to the total per-unit cost of AEDs; the cost of the device components appears to be only 20% of the sales price. Given the relatively high cost of regulation, marketing, and distribution, major price reductions below $2000 per unit are unlikely over the next several years, barring unexpected technological breakthroughs.
Special Considerations for Use in Children
Pediatric (age less than 21 years) sudden death is one tenth as common as in adults, occurring in only 1 to 2 per 100 000 children per year. Current AHA pediatric recommendations are based on the assumption that ventricular arrhythmias are a rare cause of cardiac arrest in children, but this assumption may be incorrect. Data from the Mayo Clinic reveal the following causes of sudden death from out-of-hospital events in children: idiopathic hypertrophic subaortic stenosis 47%, anomalous coronary artery 13%, dilated cardiomyopathy 4%, ruptured aorta 6%, myocarditis 6%, and unknown 24%. Except in the case of ruptured aorta, the remaining deaths are usually the result of VT. A review of published studies reporting initial cardiac rhythms during arrest in children indicates that the majority (40% to 90%) have asystole or pulseless electrical activity when initially evaluated. However, VF or VT is found in up to 23%. More important, data suggest that the survival rate is significantly higher in children with initial VT or VF than in those with other arrhythmias.23 This finding suggests that AEDs should be used in children, but there appears to be a need for additional training of AED users and modification of the recommendations for use of AEDs in children.
Transthoracic impedance has been found to have little relation to body size, but pad size is an important determinant of transthoracic impedance. Transthoracic impedance is high when small, self-adhesive pads are used and decreases dramatically as pad diameter increases. Therefore, large pads are probably better than small pads; large pads can be used for children weighing more than 10 kg, the approximate weight of a 1-year-old. Only infants, who have a low incidence of VF, appear to need small pads.24 Electrodes probably should be labeled “small” and “regular” rather than “pediatric” and “adult.” More flexibility is needed in the recommendation for energy dosing if AEDs are to be used in children. Ideally, settings of 20, 50, and 100 J are needed on AEDs to be used on children. If necessary, omission of the 20-J setting probably could be justified. Gutgesell et al25 found that 2 J/kg was 91% successful in defibrillation of pediatric patients with VF.
There may be problems with VF detection algorithms by AED in children. For example, the upper rate limit of 150 to 180 overlaps the rate range of sinus tachycardia in young children. The real question is whether differences in the frequency and amplitude of the signal would alter detection of VT or VF.
Field Experience With the AED
To date, AED use outside the hospital has been limited primarily to defibrillation by trained EMTs, as was discussed earlier (see “Current State of Emergency Medical Services”). Most out-of-hospital cardiac arrests occur at home, and most are witnessed by family members who are highly motivated to assist their loved ones. These facts provide the rationale for placement of AEDs in the homes of persons at risk of cardiac arrest and training of family members in the use of AEDs. Candidates for AED use are patients known to be at high risk for sudden death, including those with a reduced left ventricular ejection fraction, complex ventricular ectopy and nonsustained VT, late potentials on signal-averaged ECG, and diminished heart-rate variability after myocardial infarction. These persons have a high probability of experiencing an arrhythmic event or sudden death within 1 year after identification, and they are good candidates for surveillance with an AED. However, patients at highest risk are candidates for monitoring by AICD or antiarrhythmic therapy. AED surveillance may be needed for patients treated with antiarrhythmic therapy or continued observation after syncope.
Published reports of use of defibrillation devices in the home (at-home defibrillation) are limited, but successful defibrillation and survival have been reported.26 27 Problems include physician reluctance to recommend or prescribe at-home AED use. Even if an AED is available, successful outcome depends on several sometimes fortuitous events. The arrest must be witnessed, the device must be immediately available and used rapidly and correctly, the rhythm must be VF rather than asystole or pulseless electrical activity, and the postshock rhythm must perfuse the patient.27 Other impediments include the prohibitive cost of the defibrillation device, feelings of apprehension and guilt among prospective family rescuers regarding its use, and documented emotional sequelae among patients when loved ones are trained in its use.28 Despite these problems, at-home defibrillation remains an intriguing and potentially important measure against cardiac arrest.
Other field experience with AEDs involves public defibrillation by trained EMTs. Reports from large public gatherings (eg, the Kingdome, the University of Washington football stadium, World’s Fairs, and other expositions in the Seattle, Wash, area) demonstrate that cardiac arrest is relatively uncommon in such environments (L.A. Cobb, M.K. Copass, D.M. Sharp, unpublished data, 1994). Only approximately 1 in 2 million attendees, or less than 1 event per the equivalent of 1000 patient-years, is reported. The survival rate is relatively high (67% for VF arrests, 56% overall): approximately 1 life saved per 3.7 million attendees (L.A. Cobb, M.K. Copass, D.M. Sharp, unpublished data, 1994), which is attributable to the fact that the events are almost always witnessed and well-equipped paramedics are present.
Very recently, impressive data were obtained from Rochester, Minn, where remarkable survival rates were demonstrated when police vehicles were equipped with automatic defibrillators. Twenty-one of 44 persons with out-of-hospital cardiac arrest were long-time survivors. Of these 44 victims, 14 had initial defibrillation by the police force. Of those 14, 10 survived to be discharged. The overall survival rate of 21 of 44 victims can be compared with the survival rate in New York, where 26 of 2329 victims survived.29
The social context in which learning, practice, and use of resuscitation occurs appears to influence outcomes. Dracup and colleagues28 demonstrated that high-risk cardiac patients’ psychological adjustment at 3 and 6 months is influenced by the support provided to their families during CPR classes. These data suggest the contagious nature of emotions, since families, not patients, were taught CPR. When encouraged to review CPR through a mailed packet, the vast majority of family members did not do so. Formal retraining at 1 year also was deferred by most family members. Despite this, four of five persons who performed CPR during the follow-up period used the skill successfully, suggesting that retraining may not be essential if skills are used within a relatively short period. A survey of adults in southern California by Thompson (unpublished data, 1994) in preparation for the Public Access Defibrillation Conference demonstrated far more support for training family members of cardiac patients in the use of AEDs (68%) than for training the public (24%). A major concern of respondents was training and retraining the public in the use of AEDs.
Training requirements depend on whether the objective is familiarity with the concept or ease in use of the device. The public could easily be familiarized with the concept through existing media (eg, television, billboards, and newspapers). Ease of use could be minimally addressed through a training manual and an 800 telephone number for customer use. Instruction in the operation of the AED by a manufacturer’s representative could provide a higher level of training than that achieved by reading a manual. However, operation of the device does not constitute training in medical indications and proper clinical use. The latter would require a formal course, designed in accordance with clear guidelines. Lay responders may need to be trained to perform defibrillation differently than it is done in the hospital, and the medical community must accept that difference if the concept is to succeed.
Issues in training first responders are similar to those described for the public. Familiarity with the concept can easily be accomplished in 1 hour or less by integrating AED training into existing materials. Ease in use of the AED would be better accomplished with a longer teaching module. However, the current basic-life-support teaching module requires 4 to 6 hours, and many experienced trainers think it is too long.
The use of technological innovations, such as cellular telephones, on-line interactive television, and home computers that link to medical information resources, may be useful in training. These technologies can provide preparatory education on the use of AEDs, tutorial guidance at the time of resuscitation occurrence, access to the victim’s medical records, and direct medical assistance. Such technological advances create the possibility of wide distribution of health messages designed for specific symptoms or needs. In many cases, interactive television could be used to bring the doctor to the patient. Also, in the future, the 911 telephone number may be available as a television channel. Use of technology to activate EMS systems would require safeguards to minimize false alarms and thus avoid overburdening the system with unnecessary runs. This technology would have to be so reliable that calling 911 via telephone would be unnecessary. An alternative to a 911 television channel is access to a leased medical alert system that would forward the 911 call when the AED is powered on. Equipment updates and maintenance could be incorporated into such a system.
One limitation to teaching with highly technical media is that the lack of interpersonal interaction may inhibit learning. The quality of social interaction during a crisis has a potent effect on the bystander’s ability to perform a complex psychomotor skill while under emotional duress. Clawson30 argues that emergency medical dispatchers who respond according to protocol (ie, by reasserting control and issuing repetitive instructions) achieve higher compliance with telephone instructions than do emergency medical dispatchers functioning without protocols. Unfortunately, only 15% of emergency medical dispatchers nationwide are formally trained; the percentage of those trained effectively is significantly lower. Bystanders have been shown to perform CPR effectively in response to telephone instruction. The need for debriefing—talking through the event with another person—after a crisis repeatedly has been shown to be essential for later resolution and adjustment.
The cost-effectiveness of AEDs is contingent on consideration of the perspective and goals of the decision-maker, the cost and effectiveness of the improvements, and their interrelation, discounting, and sensitivity analyses. Cost is measured in dollars, whereas effectiveness may be measured in lives saved or years of life saved, with or without adjustment for quality of life. Although total cost and effectiveness are pertinent, comparison with other programs is facilitated by calculating the incremental cost-effectiveness ratio, defined as the quotient of incremental costs divided by incremental effectiveness.
For a decision-maker in an EMS system at the local government level, relevant costs may be limited to the additional budgetary costs of instituting and operating an AED program. For a decision-maker with the broader perspective of the healthcare system or society, the additional relevant direct and indirect costs may include the cost of wages for first responders or EMS workers, the hospitalization costs for those who have spontaneous circulation restored and are admitted to the hospital, the cost of cardiac care received or averted (irrespective of whether such costs are directly related to the intervention), the cost of noncardiac diseases affected by changes in cardiac prognosis, and the cost of low economic productivity.
For decision-makers in an EMS system or local government, the relevant effectiveness measure may be lives saved at hospital discharge. For a decision-maker within the healthcare system or society, effectiveness may be measured in life-years or quality-adjusted life-years.
The best available relevant information pertains to the incremental cost-effectiveness of AEDs as part of an EMS system (T. Valenzuela, S. Cretin, M.P. Larsen, M. Hlatky, unpublished data, 1994). The incremental cost-effectiveness of an EMS system with first-responder access to AED use versus an identical EMS system without this capability in the treatment of witnessed VF is $9000 per life saved with performance of bystander CPR or $15 000 without bystander CPR. With longer response times, the incremental cost-effectiveness of first-responder AED use is substantially less attractive. These cost-effectiveness ratios were not sensitive to changes in population density.
From a societal perspective, the estimated incremental cost-effectiveness ratio for upgrading a one-tier EMS system to include defibrillation with basic life support or that of implementing first-responder AED on a two-tier EMS system is comparable to those for other existing healthcare programs (G. Nichol, A. Laupacis, I. Stiell, K. O’Rourke, A. Anis, H. Bolley, A. Detsky, unpublished data, 1994). Thus, the addition of AEDs to an existing EMS system appears to be economically attractive. It is plausible that the institution of public access defibrillation may be associated with similar cost-effectiveness ratios. Additional research is necessary to determine the incremental costs and effectiveness of instituting a public access defibrillation program.
Public Policy and Legal Considerations
A key challenge to widespread availability of AEDs lies with the appropriate legal requirements for FDA approvals, postmarket tracking, and medical device reporting. The FDA has considered tentative criteria for the achievement of 510(k) FDA approval, including relatively small premarket clinical trials to document AED efficacy in controlled settings. Such clinical trials might involve as few as 25 patients but must document the efficacy of the device using reliable evaluative data. Additional postmarket evaluations in the EMS setting would be required for new AED devices, covering perhaps 150 consecutive events at preselected EMS sites, again using reliable transcribed data supporting device safety, performance characteristics, and efficacy. The FDA seeks the development of a standardized, stronger medical device–reporting database to track AED efficacy and safety. While the FDA has pledged to work with industry representatives, FDA regulatory measures continue to be a major challenge to widespread availability of small, portable AEDs.
In addition, widespread public availability of AEDs will require close involvement with state and local EMS systems, which vary widely in their structure and relations to other agencies within each jurisdiction. These variations must be recognized in any effort to increase AED availability to the public. Training in the use of AEDs would need to be added to EMT curricula in many areas. The choice between placing AEDs in ambulances or with persons other than EMTs, including volunteers, should be made by each locality. A flexible approach seems to be a key to AED access among EMS providers.
None of the legal and regulatory obstacles to greater public use of AEDs appear to be insurmountable. The most significant obstacles are FDA regulatory requirements. The FDA approves medical devices only for labeled uses. Current FDA-approved labels state that defibrillators are intended for use by first responders or on prescription of a physician. Broader public use of AEDs by laypersons will be inhibited unless either practical mechanisms can be devised for use of AEDs by laypersons within the FDA-labeled use requirement or the use requirements are amended to permit use by laypersons. Many state regulations similarly limit use of defibrillators, suggesting the need for revision of state statutes that define the use of AEDs as the practice of medicine. Another FDA rule that needs to be amended is the requirement for informed consent of patients, who presumably would be unconscious.
One potential liability exposure for manufacturers if AEDs become widely available to the public is the threat of product liability, breach of warranty, or negligent misrepresentation claims arising when an AED allegedly malfunctions or injures the victim or patient. FDA approval of a device will not insulate the manufacturer against such claims. Legislative limitations on product liability claims, already an element of several tort-reform and healthcare-reform initiatives, would be desirable. Of course, a defect-free product is the goal of device manufacturers.
Product liability is not a major problem for manufacturers, who are exposed to product liability as a matter of course in all domains. A larger issue is liability of sponsors, employers, and employees who provide defibrillation to victims of cardiac arrest. Insurers of public buildings, country clubs, churches, etc, have in some instances been unwilling to provide an umbrella for such liability. Bystanders are more likely to be protected by Good Samaritan laws in jurisdictions such as the state of California. Such laws should be extended to include all bystanders, including those using AEDs. The AHA, the American College of Cardiology, and the American Red Cross should join together to encourage passage of such legislation.
A potential disincentive to lay users of AEDs, particularly when the person receiving defibrillation is a stranger rather than a family member, is the threat of a personal injury claim. The person who is defibrillated (or his or her family or survivors) may file suit against the person who performed defibrillation (AED user). When one person comes to the aid of another, the assumed-duty doctrine requires that person to exercise reasonable care. Accordingly, AED users owe the victim of cardiac arrest a legal duty of reasonable care. The standard of care—whether liability results from gross negligence or ordinary negligence—will depend on the relation between the AED user and the victim (eg, was the victim a stranger or an employee, guest, or passenger). The argument that the victim was clinically dead when the AED was used may be persuasive in seeking dismissal of such a lawsuit or in assessment of damages but may not deter the filing of a lawsuit. Likewise, the presumption that most lay users would be spouses, family members, or friends may afford practical protection against a lawsuit, but it is not without precedent for a plaintiff to sue a family member or spouse to recover an insurance settlement. Legislative expansion of Good Samaritan–type statutes (which typically provide that someone acting in good faith in an emergency is liable only for gross negligence, not ordinary negligence, when gratuitously rendering aid) to cover AED users could remove a potential disincentive to lay use of AEDs.
Police, firefighters, and other public safety personnel generally will not benefit from Good Samaritan statutes because they have a preexisting duty to the general public to render aid. They may enjoy other forms of statutory or common law immunity, but such protection should be examined to determine whether use of an AED is within the intended scope of the protection.
Conference Participant Recommendations
1. Increase public awareness of the issues surrounding sudden death through public education.
2. Prepare specific guidelines for use of AEDs by first responders.
3. Collaborate with the Association for the Advancement of Medical Instrumentation (AAMI) and the Emergency Care Research Institute (ECRI) to develop criteria for assessing the safety and efficacy of AEDs.
4. Work with medical manufacturing industry to define markets, to identify barriers to widespread use, and to promote development of an AED unit that is easy to use, low maintenance, and low cost.
5. Participate actively in the device evaluation and regulatory process, including support for postmarket surveillance.
6. Promote legislation that would permit use of AEDs by the lay public. Involve the FDA from the beginning and throughout the dialogue to identify a stepped approach to overcoming legislative barriers (ie, in-home use by family members as a first step).
7. Form relationships with the National Organization of Police Chiefs and Fire Chiefs to facilitate implementation of AEDs by public services such as firefighters and police.
8. Initiate and obtain funding for collaborative research in medical, engineering, and epidemiological areas that is designed to test the assumption that AEDs will substantially improve outcomes.
9. Advocate the need for research by other agencies, such as the National Institutes of Health and the NHLBI.
10. Explore medical device industry funding of research.
11. Develop methods for evaluating AED training to ensure that participants who have completed a course successfully have an acceptable level of mastery and comfort in use of the device.
12. Prepare general guidelines in collaboration with other organizations, such as the American Red Cross, outlining current and future efforts and desired outcomes of the AED movement.
13. Explore special considerations in the use of AEDs with pediatric victims of sudden cardiac arrest.
1. Take into account the high failure rate of conventional out-of-hospital CPR, the time constraints that preclude meaningful clinical research, and the negative effect on patient outcomes of conventional informed-consent requirements.
2. Qualify the AED for 510(k) approval, with the understanding that there will be rigorous postmarket surveillance.
3. Assume a benevolent or “scoutmaster” role in approval of devices.
Initiate and fund large clinical trials aimed at evaluating AEDs in use by first responders and the lay public. Such trials should document the safety, efficacy, and cost-effectiveness of the device as well as the training methods required to achieve desirable outcomes.
The rate of sudden cardiac death in any community is low. Time to defibrillation after onset of VF is critical. Currently, most out-of-hospital cardiac arrests are managed by EMS systems, but the availability of AEDs to these systems remains low across the United States. In environments such as large urban or rural communities, EMS personnel often cannot reach the victim in time to achieve meaningful resuscitation. In these settings, wide distribution to the lay public of simply designed AEDs appears to be a promising approach to improving survival rates from cardiac arrest.
Semiautomatic external defibrillators are safe and effective for terminating VF, but there is a need to develop enhanced, simple-to-operate, reliable, fully automatic external defibrillators for use by the lay public. Additional research is needed to enhance the technological development of AEDs, and this will require the cooperation and collaboration of the medical manufacturing industry, academia, government, venture capitalists, and the AHA.
Clinical trials of public access defibrillation are warranted to substantiate the safety, efficacy, and cost-effectiveness of this approach. The medical device manufacturing industry, the National Institutes of Health, the FDA, and the AHA should work together in these trials to prove that this innovative approach saves significantly more lives than current emergency cardiac care methods. Data gathered in well-designed, properly executed studies are needed to convince society, the medical community, and regulatory and funding agencies of the need for public access defibrillation, a promising new strategy to improve survival rates after cardiac arrest.
An AHA medical/scientific statement related to this topic is printed on page 2763 of this issue of Circulation.
This report was approved by the American Heart Association Board of Directors in June 1995.
Requests for reprints should be sent to the American Heart Association, Emergency Cardiac Care Department, 7272 Greenville Avenue, Dallas, TX 75231-4596.
- Copyright © 1995 by American Heart Association
Moss AJ, Osborne R, Baue A, Lees R, Jamison R, Spann J. Closed-chest cardiac massage in the treatment of ventricular fibrillation complicating acute myocardial infarction. N Engl J Med. 1962;267:629-631.
Thompson RG, Hallstrom AP, Cobb LA. Bystander-initiated cardiopulmonary resuscitation in the management of ventricular fibrillation. Ann Intern Med. 1979;90:737-740.
Cobb LA, Eliastam M, Kerber RE, Melker R, Moss AJ, Newell L, Paraskos JA, Weaver WD, Weil M, Weisfeldt ML. Report of the American Heart Association Task Force on the Future of Cardiopulmonary Resuscitation. Circulation. 1992;85:2346-2355.
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;73:701-709.
Rural Emergency Medical Services: Special Report. Washington, DC: US Congress, Office of Technology Assessment; 1989. Report No. OTA-H-445.
Bardy GH, Gliner BE, Kudenchuk PJ, Poole JE, Dolack GL, Jones GK, Anderson J, Troutman C, Johnson G. Truncated biphasic pulses for transthoracic defibrillation. Circulation. 1995;91:1768-1774.
Greene HL, DiMarco JP, Kudenchuk PJ, Scheinman MM, Tang AS, Reiter MJ, Echt DS, Chapman PD, Jazayeri MR, Chapman FW, et al. Comparison of monophasic and biphasic defibrillating pulse waveforms for transthoracic cardioversion: Biphasic Waveform Defibrillation Investigators. Am J Cardiol. 1995;75:1135-1139.
Walcott GP, Walcott KT, Knisley SB, Zhou X, Ideke RE. Mechanisms of defibrillation for monophasic and biphasic waveforms. PACE Pacing Clin Electrophysiol. 1994;17(pt 2):478-498.
Charbonnier FM. Algorithms for arrhythmia analysis in AEDs. In: Tacker WA Jr, ed. Defibrillation of the Heart: ICDs, AEDs and Manual. St Louis, Mo: Mosby/Yearbook Inc; 1994.
Troutman K. Historical summary: how CDRH came to focus on AEDs. Presented at the FDA Center for Devices and Radiological Health Meeting With AED Manufacturers; Washington, DC. September 23, 1994.
Caterine MR, Yoerger DM, Spencer KT, Miller SG, Kerber RE. Are paddle electrode position and gel technique important for transthoracic defibrillation? Circulation. 1994;(suppl I)90:I-287. Abstract.
Atkins DL, Sirna S, Kiieso R, Charbonnier F, Kerber RE. Pediatric defibrillation: importance of paddle size in determining transthoracic impedance. Pediatrics. 1988;82:914-918.
Gutgesell HP, Tacker WA, Geddes LA, Davis S, Lie JT, McNamara DG. Energy dose for ventricular defibrillation of children. Pediatrics. 1976;58:898-901.
Clawson JJ. Emergency Medical Dispatching: Rapid Identification and Treatment of Acute Myocardial Infarction. Bethesda, Md: National Institutes of Health; 1994. Publication NIH 94-3287.
- History of the Automatic External Defibrillation Task Force
- Current State of Emergency Medical Services
- Optimal Design and Application
- Field Experience With the AED
- Training Guidelines
- Public Policy and Legal Considerations
- Conference Participant Recommendations
- Info & Metrics