Automated external defibrillators

INTRODUCTION — 

Sudden cardiac arrest (SCA) refers to the sudden cessation of cardiac activity with hemodynamic collapse and is most often due to sustained ventricular fibrillation (VF) or pulseless ventricular tachycardia (VT). SCA is a major public health challenge, accounting for approximately 5 to 15 percent of total mortality in industrialized nations [1-3]. (See "Overview of sudden cardiac arrest and sudden cardiac death".)

Although survival from SCA remains generally poor, there is evidence from contemporary population-based registries that outcomes following out-of-hospital and in-hospital cardiac arrest have improved compared with historical experiences. However, there is substantial disparity across systems and hence opportunity to improve outcomes. Based on contemporary estimates of out-of-hospital cardiac arrest, approximately 10 percent of emergency medical services-treated patients in any cardiac rhythm and 30 percent of patients whose initial rhythm is VF survive to be discharged from the hospital [2-5]. Based on registry data from the United States and Great Britain, contemporary survival rates for in-hospital arrest are estimated at 20 percent for all rhythms and nearly 50 percent for patients with an initial rhythm of VF [6]. (See "Prognosis and outcomes following sudden cardiac arrest in adults".)

Although several interventions can improve the likelihood of VF resuscitation, the single most important is early delivery of an external electric shock to reset the cardiac rhythm and restore spontaneous circulation [7,8]. Early defibrillation is consistently associated with a greater likelihood of survival, which decreases by approximately 5 to 10 percent with each additional minute from collapse to defibrillation [9]. The potential benefit of early defibrillation is best illustrated by the outcomes following defibrillation at casinos; 74 percent with witnessed VF survived when a shock was delivered within three minutes from collapse [10]. A 2017 review of observational studies reported early defibrillation with automated external defibrillators (AEDs) is associated with an approximate doubling of survival when an AED was applied by lay first responders (survival of 53 percent) compared with professional personnel dispatched by emergency medical dispatch centers (survival of 29 percent) [11]. Collectively, immediate access and application of an AED enables early defibrillation for shockable rhythms, thereby offering the best chance of functional survival following SCA.

This topic will review the development, use, allocation, and efficacy of AEDs. Other aspects of electrical cardioversion and defibrillation are discussed separately, as are basic and advanced cardiovascular life support. (See "Basic principles and technique of external electrical cardioversion and defibrillation" and "Cardioversion for specific arrhythmias" and "Adult basic life support (BLS) for health care providers" and "Advanced cardiac life support (ACLS) in adults" and "Supportive data for advanced cardiac life support in adults with sudden cardiac arrest".)

AED TRAINING and OPERATION

AED training — AEDs are designed to be straightforward to operate, and multiple studies have demonstrated that laypersons can operate them safely and effectively [12-15]. Nevertheless, they can be challenging to use, especially for the layperson [16-18].

The best approach for training laypersons to achieve and maintain AED operational proficiency is not well established, although face-to-face, video, and web-based training approaches have demonstrated merit [19,20]. A study of older laypersons previously trained in AED use showed that emergency dispatchers were successfully able to assist rescuers to use AEDs via telephone instructions [12].

Online and in-person classes are available (https://www.redcross.org/take-a-class/aed) to assist persons with AEDs or those who are considering purchase of an AED.

AED operation — AEDs utilize two self-adhesive electrode pads placed directly on the bare chest to detect the cardiac rhythm and to deliver shocks when indicated. Patient and background motion can impact diagnostic accuracy, which is why rescuers are typically instructed to pause cardiopulmonary resuscitation (CPR) during rhythm analysis. International standards require that AEDs have a sensitivity of >90 percent for detecting ventricular fibrillation (VF; at least 0.2 mV in amplitude) and an overall specificity of >95 percent, a level of discrimination that compares favorably with manual field interpretation [21,22]. In a 2015 study comparing four commercially available AEDs, all of the devices correctly identified VF greater than 95 percent of the time; however, there was a wide range of diagnostic accuracy for correctly identifying ventricular tachycardia (VT) and supraventricular tachycardia (SVT) [23]. In a 2018 study of seven different AED models that were tested in an airplane simulator under different levels of background motion and turbulence, five of the devices correctly identified all rhythms (sinus rhythm, asystole, and VF at five different amplitudes) at all levels of turbulence [24].

Most AEDs deliver between 120 and 360 Joules, with the output depending upon several factors including the number of shocks previously administered, the impedance of the chest wall, and whether a monophasic or biphasic waveform is used. Some AEDs are automatically adjusted to deliver less electricity (intended for children) when pediatric pads are attached. (See "Basic principles and technique of external electrical cardioversion and defibrillation", section on 'Monophasic versus biphasic waveforms'.)

AEDs typically provide audio prompts that direct rescuers to stand clear of the victim during rhythm analysis and to press a button to deliver the shock. AEDs are programmed to subsequently reanalyze the electrocardiogram (ECG) rhythm typically every two minutes. During the intervening time period, the AED prompts the rescuer to check for signs of life and, if needed, perform CPR. If the patient has an implantable cardioverter-defibrillator (ICD) that is delivering shocks, the ICD should be allowed to complete its treatment cycle (typically 30 to 60 seconds) before the AED is attached.

Pad placement — The 2010 Advanced Cardiac Life Support (ACLS) guidelines make the following recommendations regarding placement of AED pads [25]:

●AED pads should be placed in either the anterolateral or anteroposterior position (Panchal).

●Pad sizes of at least 8 cm in diameter are recommended for adults and should be placed at least 2.5 cm (1 inch) away from any implantable devices.

●AED pads should not be placed directly on top of a transdermal medication patch since it can interfere with therapy and also cause skin burns. The medication patch should be removed, and the skin should be wiped clean.

●Jewelry, clothing, and chest hair can potentially interfere with optimal pad adhesion and may need to be removed. Chest hair can be rapidly removed by an adhesive AED pad or by shaving the chest in the area where the pad will be placed.

Other features — Additional features present in some AED models include the ability to continuously record the arrest rhythm ECG, derive measures of CPR performance such as chest compressions or ventilations, and record the voices of rescuers involved in the event. The combination of these features enables case review that may be used for quality assurance or research. Such data indicate that CPR often does not meet guideline standards and is frequently interrupted [26-29].

Newer AED models can incorporate more dynamic prompts or real-time feedback to guide rescuer CPR actions. Although these real-time prompts can improve CPR performance, it is not certain if these features will improve survival [30]. Advances in signal processing can reduce or even eliminate the CPR pause for rhythm analysis [31], providing a potential improvement in care given that excess CPR interruption is associated with lower likelihood of survival [32].

AED ALLOCATION STRATEGY — 

AEDs are effective only in those patients who present with ventricular fibrillation (VF) or pulseless ventricular tachycardia (VT); pulseless electrical activity and asystole are not effectively treated by defibrillation. Although the proportion of VF/VT arrests is declining relative to nonshockable rhythms [33], VF/VT arrests still account for tens of thousands of deaths each year in the United States, and the improved resuscitation provided by AEDs can have a significant impact on public health.

AEDs enable people who are not trained in rhythm interpretation to provide life-saving therapy, which vastly increases the pool of potential rescuers who can provide early defibrillation. However, their allocation and use require programmatic support, training, and maintenance, all of which can contribute to overall cost. Thus, an important consideration regarding these devices is efficient distribution of AEDs throughout communities.

Strategies for AED allocation include providing them to traditional emergency medical services (EMS) and to nonmedical emergency responders (eg, police officers and firefighters), as well as placing them in public locations, in hospitals, and in homes of individuals. Community response strategies have incorporated AED registries and geospatial technology (smartphone applications) to alert laypersons of nearby suspected SCA and the location of available AEDs [34-37]. There is also a preliminary experience using an unmanned aerial drone to deliver an AED to the location of suspected SCA [38]. All these strategies are designed to reduce the interval from SCA to initial defibrillation given the powerful relationship between successful resuscitation and shorter interval from collapse to initial AED shock.

Emergency medical services AED programs — The initial large-scale implementation of AEDs was by emergency medical services (EMS) in the 1980s and 1990s. This strategy provided the potential for earlier defibrillation by allocating AEDs to EMS first responders (emergency medical technicians), many of whom were not typically trained in rhythm interpretation.

Some EMS AED programs were associated with improvements in survival, but others were not [39-43]. Meta-analyses found that EMS AED programs resulted in a significant 9 percent increase in survival [44,45]. However, weaknesses in individual study design limit the strength of these conclusions.

In some reports, survival from sudden cardiac arrest (SCA) due to VF did not improve despite reductions in the time to defibrillation observed with the adoption of the AED [46,47]. One explanation for this discrepancy is that the resuscitation algorithms that were originally used for AED rhythm analysis and processing required considerable interruptions in cardiopulmonary resuscitation (CPR) compared with treatment with a manual defibrillator [26-29] and that the increase in "hands-off" time reduced the chances of successful resuscitation [48-51]. Subsequent improvements in AED algorithms have reduced CPR interruption with corresponding improvement in survival [47]. (See "Advanced cardiac life support (ACLS) in adults", section on 'Excellent basic life support and its importance'.)

Police AED programs — Police officers can sometimes respond to SCA victims more quickly than EMS providers [52]. A program of providing AEDs to police officers and training them in their use was initially introduced in the late 1980s in Rochester, Minnesota [53]. Police often arrived at SCAs due to ventricular fibrillation (VF) prior to EMS and defibrillated patients an average of 5.5 minutes following collapse [53]. Ten of 14 patients survived to hospital discharge. In a later series of 193 patients, survival from witnessed VF to hospital discharge was 46 percent. Most were neurologically intact [54]. In contrast, survival from SCA not caused by VF was only 5 percent.

Police AED programs in Pittsburgh, Pennsylvania, Miami-Dade County, Florida, King County, Washington, Portland, Oregon, and Zurich, Switzerland, have also been associated with greater survival through earlier CPR and defibrillation [55-59]. However, other police AED programs have not shown survival benefits compared with standard EMS, particularly when the police often have not arrived before EMS [60]. Thus, one of the keys to a successful police AED program is committed medical and police leaders who can motivate police about their potential lifesaving role and in turn respond quickly. A meta-analysis of police AED programs, published in 2013, described the potential benefits and challenges of implementation [61].

Public access defibrillation programs — Surveillance studies have identified particular locations where SCA occurs with high frequency, including public transit facilities, shopping malls, public sports venues, industrial sites, golf courses, casinos, dialysis centers, airports, and fitness centers [62-68]. AEDs strategically located in such places can be used by laypersons to deliver defibrillation prior to EMS arrival, a concept referred to as "public access defibrillation" (PAD). The benefits of PAD on survival rates and neurologic outcomes after SCA are illustrated by the following findings [10,69-73]:

●The Public Access Defibrillation (PAD) trial, a prospective multi-community randomized trial, evaluated survival to discharge in 526 patients with SCA. Following SCA, survival to discharge rates significantly increased in high-risk public sites where CPR-trained lay responders were equipped with AEDs compared with sites where lay responders were CPR-trained, but did not have access to AEDs (23.4 versus 14.0 percent) [73]. Furthermore, PAD programs implemented at high-risk sites offer reasonable health benefit for the cost, ranging from $35,000 to $57,000 per quality adjusted life-year, which is comparable to other widely-accepted medical interventions such as bone marrow transplant ($52,000 per quality adjusted life-year) and heart transplant ($59,000 per quality adjusted life-year) [74-78].

●A large, multi-community cohort study evaluated the outcome of over 13,000 patients with an out-of-hospital SCA. Survival to hospital discharge was markedly greater in patients who received an AED-delivered shock from a non-traditional responder (most often layperson) compared with patients receiving bystander CPR alone or those presenting with VF whose initial shock was delivered by EMS (38 versus 9 versus 22 percent, respectively) [72].

●Survival with intact neurologic function is higher in patients with SCA who receive treatment with an AED available at the site of the arrest. In a study of 2833 consecutive patients with out-of-hospital SCA, neurologically intact survival was significantly higher among those treated with an on-site AED in addition to basic life support (BLS; 50 versus 14 percent with BLS alone, adjusted odds ratio [OR] 2.72, 95% CI 1.77-4.18) [79].

●Nationwide dissemination of AEDs in public places in Japan from 2005 through 2013 was associated with an increase in the proportion of shocks for witnessed VF arrest administered by laypersons with AEDs from 1.1 to 16.5 percent [80]. As public access defibrillation increased, mean time to shock was reduced (from 3.7 to 2.2 minutes), with a significant improvement in one month survival with favorable neurologic function (38.5 percent compared with 18.2 percent for those who did not receive public access defibrillation; adjusted OR 2.0, 95% CI 1.8-2.2) [80,81]. A 2018 report from Japan also noted increased survival and improved neurological outcomes among school-aged patients receiving public access defibrillation [82].

●Improvements in time to defibrillation and survival with favorable neurologic function have also been reported from a 2006 to 2012 nationwide study of AED use in the Netherlands [83].

The survival benefit observed in these programs has led to advocacy for lower-risk sites to implement PAD programs. As an example, guidelines have been established for school-based AED programs which have increased substantially over time, partly because of legislation [84-86]. PAD programs are also mandated in many federal jurisdictions [87]. Use at lower-risk sites provides the opportunity to increase the number of SCA survivors but with a lower cost-effectiveness. At present, PAD program AEDs are involved in only a small fraction of all out-of-hospital SCAs so that strategies that increase their use are a promising strategy to improve survival [36,71,88]. The emergency dispatcher may be able to increase use of PAD AEDs in certain circumstances where there are multiple bystanders [89].

Smartphone applications for notification of cardiac arrest — Smartphone applications now exist that allow a person close to a suspected cardiac arrest to be notified of the arrest by an emergency dispatch center. These apps (PulsePoint, HeartRunner, GoodSAM as examples) are downloaded on volunteers' phones, and when a suspected cardiac arrest occurs nearby, volunteers (typically within a quarter mile of the event) are alerted on their smartphone with the location pinpointed on a map. The alerting application can be linked to an AED registry such that the activation can direct volunteer responders to the suspected cardiac arrest location and/or the location of a nearby accessible AED. The strategy of crowdsourcing volunteer response through smartphone applications to both public and residential settings for suspected cardiac arrest appears safe and can increase early CPR, defibrillation, and survival in select communities and population density circumstances [34,35,90-94]. The optimal activation strategy is uncertain with regard to radius of alert or the priority of direct patient response versus AED retrieval and may depend upon the number of eligible volunteer responders, the distribution of accessible AEDs, and the community's EMS response characteristics [95].

AED via drone delivery — A provocative approach to early AED availability considers deployment of AED via unmanned arial vehicles (ie, drones). Simulation evaluations suggest that AEDs can arrive sooner via drone delivery in a relevant proportion of SCA, especially when locations have slower conventional emergency response [96-98]. Preliminary experiences have demonstrated operational feasibility and safety [38]. These drone-delivered AEDs can arrive within meters of the patient, sometimes before EMS [99]. Whether such a strategy can improve outcomes requires more investigation. One potential novel challenge of this strategy is how such a program operates in compliance with airspace and flight regulations from regional and national oversight organizations.

AEDs for use in private homes — Since approximately three-quarters of SCAs occur in private homes, one strategy to reduce mortality is to distribute AEDs for use in the home, based on models from other home safety strategies such as smoke detectors and fire extinguishers [100].

Home use of AEDs was investigated in a randomized trial of 7001 patients with previous anterior wall myocardial infarct ion who were not candidates for an implantable cardioverter-defibrillator [101]. The median age was 62 years, and the median left ventricular ejection fraction was 45 percent. The designated rescuers were predominantly female (83 percent) and their median age was 58 years.

Patients were randomly assigned to AED use followed by calling EMS and performing CPR or to the control response of calling EMS and performing CPR. Access to a home AED did not improve survival as compared with conventional resuscitation (6.4 versus 6.5 percent, hazard ratio 0.97, 95% CI 0.81-1.17).

Several factors may have contributed to the lack of benefit in this trial:

●The incidence of sudden cardiac arrest (2.3 percent) and overall mortality were lower than predicted.

●One-half of the tachyarrhythmia arrests that took place at home were witnessed (58 of 117), and an AED was used in only 32 patients.

●Spouses and companions in the control group received training in resuscitation, with frequent reminders.

In determining whether AEDs are appropriate for home use, cost and the role of implantable cardioverter-defibrillators in individuals at high risk of SCA must be considered.

In-hospital AED allocation — Delayed defibrillation is common during in-hospital arrest even though medical personnel are often trained in advanced cardiac life support and ECG rhythm interpretation and are capable of implementing manual defibrillation.

The frequency of delayed in-hospital defibrillation (defined as greater than two minutes from the time of recognition of arrest) and its adverse effect on survival were illustrated in a study utilizing 6789 patient records from the National Registry of Cardiopulmonary Resuscitation [102]. Delayed defibrillation for ventricular fibrillation or pulseless ventricular tachycardia was observed in 30 percent of SCAs and was associated with a significantly lower probability of survival to discharge compared with survival when defibrillation was performed within two minutes (22 percent versus 39 percent).

The possibility that AED could improve survival of in-hospital SCA was suggested by small studies showing that AEDs allocated to specific clinical and non-clinical areas of the hospital allowed for more rapid defibrillation [103,104].

Subsequently, the outcomes of in-hospital SCA were analyzed using data obtained from 253 US and Canadian hospitals as part of the National Registry of Cardiopulmonary Resuscitation. Amongst a cohort of 11,695 hospitalized patients who suffered SCA between 2000 and 2008, 39 percent of cardiac arrests were treated using an AED [105].

●Patients with a shockable rhythm (ie, pulseless ventricular tachycardia or ventricular fibrillation) had similar survival to hospital discharge in the AED and non-AED group (38.4 versus 39.8 percent, adjusted relative risk [RR] 1.00, 95% CI 0.88-1.13).

●Patients without a shockable rhythm (ie, asystole or pulseless electrical activity) had a lower rate of survival to hospital discharge when an AED was employed (10.4 versus 15.4 percent in the non-AED group, adjusted RR 0.74, 95% CI 0.65-0.83).

These results suggest that, for patients with a shockable rhythm, AED use was not associated with a survival difference compared with manual external defibrillation in a hospital setting. Among nonshockable rhythms, AED use was associated with a lower survival. The explanation for lower survival may be the disproportionate excess of asystole in the AED group, some other unmeasured confounder, or the potential that AED application and use may truly delay or interrupt other beneficial therapies.

The optimal strategy of AED distribution and its ultimate benefit may depend upon a particular hospital's staffing, geography, and patient profile [106].

AEDs in medical and dental practices — Cardiac arrest in a medical or dental setting is an infrequent event. Data from King County, Washington, rank the likelihood of a cardiac arrest occurring, with dialysis centers having the highest risk (approximately one per year). The next highest risk locations are cardiology practices, urgent care centers, internal medicine, and family medicine. The lowest risk locations are dental settings. Despite the low risk in most practices, we believe virtually all medical practices should have an AED.

CHALLENGES AND OPPORTUNITIES — 

Although AEDs have saved many lives, they have several potential drawbacks.

●AEDs require the presence of a bystander to apply and operate.

●Only about 50 percent of sudden cardiac arrest (SCA) events are witnessed. Unwitnessed events are often present with nonshockable rhythms at first analysis, possibly because the rhythm deteriorates from shockable to a nonshockable (asystole) over time. Thus, effective defibrillation is often not relevant by the time the victim of an unwitnessed SCA is found.

●AEDs require interruptions in cardiopulmonary resuscitation (CPR) while they assess the cardiac rhythm. This analysis time is typically longer with AEDs than with manual defibrillators. Ongoing efforts are aimed at minimizing this time, and technical advances may eventually enable accurate rhythm interpretation even while CPR is ongoing [31,107]. (See "Advanced cardiac life support (ACLS) in adults", section on 'Excellent basic life support and its importance'.)

●The cost of AEDs can be an important obstacle to acquisition, distribution, and the potential health benefits.

Prognostic information on patients with SCA is available from continuous AED bio-signals, most commonly the ECG. Such information may be able to better align the patient's physiologic status with the timing, sequence, and dose of treatments compared with the current resuscitation protocols to improve SCA care and outcome. For example, the shape and pattern of the ventricular fibrillation waveform recorded in the ECG can predict the physiologic response to shock [108]. Likewise, ECG characteristics of an organized rhythm can predict the likelihood of spontaneous pulse versus pulseless electrical activity [109]. Increasingly, the ECG can be accurately assessed despite the artifact of CPR, supporting a strategy of patient-specific physiologic assessment that does not require CPR interruption [110]. The use of artificial intelligence to leverage these types of continuous bio-signals to improve treatment is a provocative but yet unproven strategy [111].

Wearable AEDs have been developed. Their evidence-based role has not been established, and their use is best determined through clinical assessment of risk and benefit for the individual patient. (See "Wearable cardioverter-defibrillator".)

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Basic and advanced cardiac life support in adults" and "Society guideline links: Cardiac arrest in adults".)

SUMMARY

●AED operation – Historically, definitive treatment of out-of-hospital sudden cardiac arrest (SCA) with early defibrillation was limited by the small number of qualified rescuers who could interpret cardiac rhythms. Since automated external defibrillators (AEDs) analyze cardiac rhythms and directly inform rescuers whether a shock is indicated, their advent has enabled lay rescuers and additional public safety personnel to provide early defibrillation. (See 'AED operation' above.)

●Challenges and opportunities – While AEDs can be highly effective, appropriate use typically requires interruptions in cardiopulmonary resuscitation (CPR) to assess the cardiac rhythm. The increase in "hands-off" time may reduce the chances of successful resuscitation. Advances in signal processing may ultimately reduce or even eliminate the need to interrupt CPR to accurately assess the ECG and patient status. (See 'Challenges and opportunities' above and "Advanced cardiac life support (ACLS) in adults", section on 'Excellent basic life support and its importance'.)

ACKNOWLEDGMENTS — 

The editorial staff at UpToDate acknowledges Laura Gold, PhD, and Mickey S Eisenberg, MD, PhD, who contributed to earlier versions of this topic review.