Automatic External Defibrillation

Updated: Nov 10, 2018
  • Author: Joseph J Bocka, MD; Chief Editor: Barry E Brenner, MD, PhD, FACEP  more...
  • Print


Kouwenhouven showed that electrical shocks applied to dogs within 30 seconds of an induced ventricular fibrillation (VF) could produce a 98% rate of resuscitation; however, those shocked after 2 minutes of ventricular fibrillation had only a 27% resuscitation rate. [1] This gave rise to the goal of early defibrillation for ventricular fibrillation and pulseless ventricular tachycardia (VT). The use of automatic external defibrillators (AEDs) can aid in reaching this objective.

Survival rates for individuals with ventricular fibrillation treated by AEDs have been reported between 0% and 31%. Comparatively, the survival rates for performing basic cardiopulmonary resuscitation (CPR) alone are reported between 0% and 6%. Theoretically, even more lives could be saved if targeted members of the general public could obtain early access to and have training in the use of AEDs and CPR. Unfortunately, only about 10-15% of cardiac arrests occur in a public place and even fewer are witnessed. [2]

For patient education information, see Automated External Defibrillators (AED) and Cardiopulmonary Resuscitation (CPR), as well as the Public Health Center and Heart Center.




In 1775, Abildgaard described a series of experiments in which he made hens lifeless with electrical impulses applied through the body. He could not restore a pulse, however, unless shocks were delivered across the chest. In 1849, Ludwig and Hoffa first described what Abildgaard had induced. They originated and defined the term fibrillation of the ventricles. In 1900, Prevost and Batelli conducted research on ventricular fibrillation (VF) in dogs. They found that weak alternating current (AC) or direct current (DC) shocks produced ventricular fibrillation, while much stronger current was needed to defibrillate.

Wiggers and Wegria expanded on the work of Prevost and Battelli, describing a vulnerable period of the cardiac cycle utilized to induce ventricular fibrillation. They also reported that the current delivered was the key to successfully performing what they termed "countershocking" of ventricular fibrillation.


Development of practical defibrillators began in the 1920s with funding from Consolidated Edison of New York in response to an increasing number of electric shock accidents and deaths. In 1947, Beck et al performed the first successful human defibrillation using specially designed internal cardiac paddles. [3] He used 2 110-volt, 1.5-amp AC current shocks to resuscitate a 14-year-old boy who had become pulseless during elective chest surgery. In 1956, Zoll et al performed the first successful human external defibrillation using a 15-amp AC current that produced 710 volts applied across the chest for 0.15 seconds. [4] In 1961, Alexander, Kleiger, and Lown first described the use of AC current for terminating ventricular tachycardia (VT). [5] Work by Lown et al in the early 1960s demonstrated the superiority and safety of DC over AC for defibrillation.

Ambulance-transported Belfast physicians first performed successful prehospital defibrillation in 1966. Defibrillation by emergency medical technicians (EMTs) without the presence of physicians was first performed in Portland, Oregon, in 1969 and was reported in 1972.

Automatic defibrillators

In the early 1970s, Dr Arch Diack, Dr W. Stanley Welborn, and Robert Rullman [6] developed several prototype AEDs that were tested in the Portland area. They later formed the Cardiac Resuscitator Corporation to market their device.

Prehospital trials began in Brighton, England, in 1980 using the Heart Aid. The device weighed 28 pounds and used an oral/epigastric and a precordial electrode to record ECG tracings and deliver electrical shocks. It was also capable of transcutaneously pacing the heart. In 1982, the US Food and Drug Administration (FDA) gave approval for EMT-defibrillation (EMT-D) clinical trials. Early US investigations of manual EMT-D were carried out in Washington, Iowa, Minnesota, and Tennessee. 

In the early 1990s, successful training and use of AEDs by police officers and other first responders was reported. 

Public access

In the 1990s, AED use by lay personnel was approved by the FDA, and Good Samaritan legislation soon followed. AED training was included in the American Red Cross basic CPR course beginning in March of 1999. In November 2002, the Phillips HeartStart AED was approved for home use with a prescription. New York State became the first state to mandate AEDs in schools in May 2003. The Federal Aviation Administration (FAA) mandated in April 2004 that all large passenger-carrying US airlines carry and have personnel trained in the use of AEDs.

In September 2004, the FDA began to allow home AED sales without a prescription. However, routine AED use in high-risk homes is controversial. On the other hand, its use in high-risk populated areas has resulted in numerous lives saved in casinos, schools, stadiums, and airports.


Machine Mechanics

Early models of AEDs required inserting an oral/epigastric electrode and placing a second electrode on the chest. Current AEDs require the placement of pads at the right sternal border and at the cardiac apex. These electrodes serve to both monitor and defibrillate. AEDs also can inform the user when lead contact is poor, when the machine is preparing to defibrillate, when to check for a pulse, when a nonshockable rhythm is present, or when motion is detected.

Rhythm analysis

Early AEDs were designed to respond primarily to a heart rate greater than 150 electrical complexes per minute and an electrocardiographic wave (QRS) amplitude greater than 0.15 mm. Presently, the ECG rhythm is analyzed via a combination of several methods. In addition to rate and amplitude criteria, the QRS is analyzed as to its slope, morphology, power spectrum density, and time away from the isoelectric baseline for preset levels defined as abnormal. Checks are made in 2- to 4-second intervals. In general, if abnormal complexes are detected for more than double the frequency of any other QRS for 3 consecutive checks, the AED will be primed to deliver a shock.

Fine ventricular fibrillation (VF) presents the greatest detection challenge. A trade-off exists between setting the amplitude criterion low enough to detect fine fibrillation, yet high enough to avoid shocking asystole or artifact. The sensitivity of detecting VF by AEDs has been reported as 76-96%. Specificity (correctly identifying non-VF rhythms) is reported to be nearly 100%.

Biphasic versus monophasic

Monophasic defibrillation delivers a charge in only one direction. Biphasic defibrillation delivers a charge in one direction for half of the shock and in the electrically opposite direction for the second half.

Dog studies have shown less conduction block and fewer ST segment changes after biphasic shock delivery than after monophasic delivery. Human studies have shown monophasic delivery to be equivalent to biphasic shocks for electrophysiologic study-induced (EPS-induced) and prehospital VF and VT. [7] Also, studies have shown that a biphasic waveform of 115 J is equivalent to a monophasic wave of about 200 J. Because of the decreased energy needed, most internal cardioversion defibrillators now use biphasic waveforms. Most manufacturers are converting to biphasic AEDs, as the lower amount of energy used can result in both longer battery life and a shorter time to full charge.

While there may be a theoretical clinical advantage to using biphasic defibrillation, most patient studies and reports have shown equivalency and not superiority of one form when using equivalent dosages (biphasic dosages are lower).


Studies with implantable defibrillators have shown a difference in outcome related to electrode polarity. The only prospective study with AEDs, which was performed by Weaver in 1993, showed no difference in survival related to polarity while using a monophasic unit. [8]


Classic AED Studies

Seattle (Suburban EMT AED better than manual defibrillation)

In 1987, Cummins et al reported a controlled study comparing the effectiveness of EMTs with automatic external defibrillators (AEDs) and EMTs with manual defibrillators in treating 147 patients in ventricular fibrillation (VF) in suburban Seattle, Washington. No statistically significant differences in rates of admission (54% AED; 50% manual) or survival to discharge (30% AED; 23% manual) were noted. [9]

In 1988, Weaver et al reported on the use of AEDs primarily by non-EMT first responders (3.3-min response time) followed by paramedics (8.8-min response), compared with basic EMTs (3.4-min response) followed by paramedics (5.1-min response). A prototype AED was used and modified halfway through the study. The combined results of 504 patients in VF showed no difference in admission rates (59% AED; 53% EMT) but a higher rate of survival to discharge in the AED group (30% versus 19%). [10]

Iowa (Rural AED better than manual defibrillation)

In 1986, Stults et al reported on a study comparing AEDs used by EMTs with manual defibrillators used by EMTs. The results of 88 patients in VF showed no significant difference in rates of admission (29% AED; 32% manual) or survival to discharge (17% AED; 13% manual). [11]

Minnesota (No rural AED survivors)

Bachman et al failed to confirm the results from Iowa and Seattle in rural northeast Minnesota. [12] They reported survival to discharge rates of 11% for paramedics, 5% for EMTs with manual defibrillators, and 2.5% for cardiac arrests handled by basic EMTs. Separate analysis for VF was not performed. They found no unwitnessed arrest survivors as earlier studies had, and results caused them to question the use of AEDs in rural areas.

In contrast, Vukov studied EMT defibrillation in rural southeast Minnesota in 1988. [13] In a report of 63 patients, EMTs with AEDs had significantly greater admission rates (30% versus 12%) and survival to discharge rates (17% versus 4%) than EMTs without AEDs.

Detroit, Chicago, and New York (Poor urban results)

In data that have been presented but not published, basic EMTs with AEDs treated 595 patients who were in cardiac arrest in Detroit. Only about 20% were found in VF. About 5% were admitted, and none survived to discharge. Possible factors contributing to the low initial VF and survival rates were an EMS response time of more than 10 minutes and an estimated 5-10 minute time from collapse until EMS was summoned.

Similar studies reported a 4% survival rate from VF in Chicago and a 5% survival rate in New York. Average response times were greater than 10-12 minutes.


Stults and Brown looked at the question of EMTs handling defibrillation without paramedic backup. [14] Of the 271 patients with VF who were shocked, 111 patients who initially converted to organized rhythms after defibrillation were analyzed. Of these, 19 (17%) refibrillated and 11 were reconverted by the EMT-Ds. Among the 111 who initially converted, admission rates were lower (53% versus 76%) but not statistically significantly lower (P > 0.05) for those who refibrillated. Survival to discharge rates were similar (37% refibrillation versus 35% nonrefibrillation).

Time to defibrillation

Early Seattle studies found a significant difference in time to defibrillation: 1.1 minutes for AEDs versus 2 minutes for EMTs with manual defibrillators. Bocka found that EMTs using fully automatic defibrillators in the field were on average 30 seconds faster than counterparts using semiautomatic devices. [15]


AED Precautions

Patients must be medically unstable or pulseless before AED is applied. They should be pulseless before assessing the rhythm. This becomes increasingly important as more laypersons are given the opportunity to use AEDs.

Most states have passed a liability waiver for bystanders who use an AED to assist an unconscious patient. The principal risk is that another person may be touching the patient when the shock is administered.

To prevent inaccurate analysis, CPR should be halted while the unit is analyzing the rhythm. Most new units have a motion or CPR detector.

The AED should be used only with caution in a moving vehicle. [16] If being used on a patient in transport, frequent stops for pulse and AED monitoring checks should be made. Most units are designed to warn when motion or poor contact is detected.

As with manual defibrillation, the chest does not need to be shaved prior to electrode placement but, if wet, it should be dried off. Water spots or nicks in the skin result in areas of decreased resistance and could lead to local burns as well as uneven and ineffective defibrillation. Look for nitroglycerin patches and remove them to prevent possible explosive risk.

Proper maintenance is important. However, it is estimated that less than 10% of AED users maintain their AED or replace the batteries according to the manufacturer's recommendations.

Repeated "check electrode" warnings when electrodes seem properly placed may be from fractured electrodes. Replace them and retry.

Caution should be used when applying AEDs in the presence of electromagnetic interference (EMI). One study showed a 2% false-positive rate, with shocks advised for sinus rhythm and 3.5% sensing motion artifact in the presence of EMI. [17]

Each AED should be programmed to the most current American Heart Association Defibrillation Guidelines. This should be checked, especially when purchasing an older or used AED. Arrangements should be made for reprogramming whenever changes are announced.

Several AED units have been recalled for various problems. The FDA posts summaries of information about the most serious recalls of medical devices, such as AEDs, on its List of Device Recalls Web page.

Concern has been raised about the safety of providers and AED assessment during CPR. However, a study Edelson et al concludes that continuing chest compression while the AED is charging is an underused technique in clinical practice, as it is associated with decreased hands-off time preceding defibrillation, with minimal risk to patients or rescuers. [18]


AED Selection Factors

Population density : Stapczynski et al concluded that areas with a population density of fewer than 100 persons per square mile received little benefit from AEDs. [19] A study from Washington identified 172 sites of higher incidence (of 71,000 sites). Similar local assessments may aid in efficient public AED location.

It is becoming increasingly common for AEDs to be placed in locations with significant public traffic or density (see Public access defibrillation below), such as stadiums and other event venues and airports. American Airlines was the first to carry AEDs on aircraft during long overwater flights in 1997, and subsequent FAA regulations mandated aircraft to carry AEDs and for flight personnel to be trained in their use (amendment to part 121 of Aviation Medical Assistance Act of 1998).

Response time: On average, response time should be less than 4-6 minutes for the patient to benefit from the use of AEDs.

Tiers: Tiered response systems should have compatible equipment. If an AED is to be used by a first responder, defibrillators or their electrode wiring system should be compatible with the transporting units that follow.

Monitor: If the unit also is used by paramedics for monitoring and not just by AED technicians for cardiac arrest, a monitor screen is needed. Printout capabilities are also desirable.

Event recording: A paper recording of the event may be sufficient for medical control; however, the need for a full set of data should influence equipment choice, since the data need to be downloaded. Because most of the data downloaded by AED units are not interchangeable, a setting in which data from various agencies needs to be collected and combined must use similar AEDs that can communicate; if they do not have this capability, data must be collected by hand.

Good Samaritan Immunity and Local Requirements : While most states provide excellent liability immunity for proper use and have minimal requirements for placement, one should check local laws before placement. State requirements can be found at the American Heart AED Legistation page.


Public Access Defibrillation

Public access defibrillation (PAD) has been shown to be an important part of successful chain of survival programs. [20, 21] Placement of AEDs has been most cost effective in select locations, including casinos, airports, stadiums, health clubs, universities, and senior centers. [22]

In the 1990s, AED use by lay personnel was approved by the FDA and Good Samaritan legislation soon followed. AED training was included in the American Red Cross basic CPR course beginning in March of 1999. In November 2002, the Phillips HeartStart AED was approved for home use with a prescription. The FAA mandated in April 2004 that all large passenger-carrying US airlines carry and have personnel trained in the use of AEDs.

In September 2004, the FDA began to allow home AED sales without a prescription. However, routine AED use in high-risk homes is controversial. On the other hand, its use in high-risk populated areas has resulted in numerous lives saved in casinos, schools, stadiums, and airports. In 2014, the California Supreme Court rule, in the case of Verdugo vs The Target Corporation (S207313), ruled unanimously that California law does not require retailers to have defibrillators available for medical emergencies. [23]


AED Manufacturers

AED manufacturers include the following:


The FDA and AEDs

The FDA issued a: Requirement for Premarket Approval for Automated External Defibrillator System, which became effective on February 3, 2015, which represents a tailored approach to help manufacturers improve the quality and reliability of AEDs. The order requires manufacturers of AEDs and accessories to submit premarket approval (PMA) applications that focus specifically on the critical requirements necessary to ensure AEDs are safe and reliable. These have not yet met final approval.the filing of premarket approval applications (PMA) for AED systems, which consist of an AED and those AED accessories necessary for the AED to detect and interpret an electrocardiogram and deliver an electrical shock (eg, pad electrodes, batteries, adapters, and hardware keys for pediatric use).