Luis Gruberg, MD, FACC


December 30, 2003

Editorial Collaboration

Medscape &

Presenter: Joseph P. Omato, MD, Virginia Commonwealth University (Richmond)

Approximately 460,000 US deaths occur each year as a result of out-of-hospital cardiac arrest (OOH-CA), making sudden cardiac death the leading cause of death in the United States. An estimated 95% of OOH-CA victims die -- mostly due to inaccessibility and delay of time to defibrillation. It is estimated that 20% of OOH-CA cases occur in public settings, while 80% occur in the home (and usually are unwitnessed).

Time is of the essence for these victims; for every minute that elapses without defibrillation, the chances of survival decrease by about 10%. Unfortunately, the average response time for emergency medical services (EMS) in a typical community is 6 minutes. There is obviously an increased need to provide a means to respond more promptly. Typically, OOH-CA in the public setting is witnessed by a bystander. Current initiatives are under way to provide laypersons and bystanders with the tools they need to initiate resuscitative therapy until an EMS unit can arrive.

Previous studies have shown that automated external defibrillators (AEDs) placed in public settings where laypersons and innocent bystanders (with no prior exposure, training, or understanding of AEDs) can safely and effectively use the devices to restore organized rhythm in a victim experiencing an OOH-CA. Although cardiopulmonary resuscitation (CPR) alone has shown benefit in these patients, its use is limited by the fact that training is required to administer the therapy. In addition, if a patient is suffering from a ventricular arrhythmia, defibrillation therapy may be needed. To date, no randomized trial has been conducted to test whether nonmedical, true laypersons who do not have a formal duty to treat patients in medical emergencies can save more lives when they are trained and equipped to use AEDs in addition to calling 911 and performing CPR.


Sponsored by the National Heart, Lung, and Blood Institute, with support from the American Heart Association, the Public Access Defibrillation (PAD) trial was initiated to evaluate whether the addition of AEDs to a CPR-based community volunteer response system increases survival in victims of OOH-CA.

Study Design

PAD was a prospective, randomized, controlled clinical trial that compared 2 volunteer-based OOH-CA response systems involving laypersons. Community units predefined as high risk for OOH-CA were randomized to train layperson volunteers to either call 911 and perform CPR until EMS arrived or call 911, perform CPR, and use an AED. The study's primary prespecified outcome measure was the actual number of OOH-CA survivors to hospital discharge.

The study protocol had 3 components:

  • Identified and randomized community units with a predefined increased risk of OOH-CA;

  • Identified and trained volunteers to respond to OOH-CA using CPR only or CPR + AED; and

  • Tracked all emergency medical episodes in each of the study units and identified cases meeting pre-established criteria for analysis as probable or definite OOH-CA.

An adjudication committee blinded to the intervention was used to distinguish presumed vs definite cardiac arrests. A presumed cardiac arrest was defined as the delivery of > 2 ventilations or > 5 chest compressions, the delivery of any defibrillation shock, and death from any cause. Definite OOH-CA was defined as ventricular fibrillation/ventricular tachycardia and asystole identified by EMS or determined by an AED, or if the victim was found unresponsive, without pulse or respiration, and died after CPR was administered.

Community Unit and Volunteer Criteria

Community units defined as high-risk and eligible for the study had to have the equivalent of ≥ 250 persons aged ≥ 50 on site ≥ 16 hrs/day OR a history of >/1 witnessed OOH-CA within the past 2 years. Other eligibility criteria included:

  • An identifiable pool of eligible, on-site volunteers;

  • Ability to deliver an AED to the victim in ≤ 3 minutes; and

  • EMS time to defibrillation ≤ 15 min for ≥ 90% of calls.

Units that had an existing or previous PAD program, or those with on-site medical personnel responding in ≤ 3 min of notification in > 10% of cases, were excluded from the study.

Volunteers for the study had to be laypersons who have no duty to treat in medical emergencies (ie, MDs, nurses, emergency medical technicians, firefighters, and police were excluded). There was initial training for data collection, and volunteers underwent the 2-hour HeartSaver ABC CPR course or equivalent for CPR-only units or the 4-hour HeartSaver AED course or equivalent for CPR + AED units. In addition, 1-2 retraining sessions were conducted at various planned intervals.


A total of 993 community units (representing 1260 facilities) were randomized at 21 US sites and at 3 Canadian sites. Approximately 20,000 lay volunteers underwent training. The data collection period averaged 21.5 months.

The majority of OOH-CA cases (85%) occurred in public places (the majority in recreation, shopping, and entertainment facilities), and 15% occurred in private residences (apartment buildings or gated communities). OOH-CA occurred in 103 patients treated by a CPR-only community unit and in 129 patients in the CPR + AED group. Patient clinical characteristics were well-matched between the 2 groups (Table 1).

Table 1. Baseline Clinical Characteristics of OOH-CA Patients
Characteristic CPR Only
(n = 103)
(n = 129)
Age (yrs) 72 70
Male (%) 65 70
Caucasian (%) 67 79
Indoor location (%) 81 77
Nonsedentary (%) 66 66
Witnessed (%) 68 76

The study's primary endpoint, the number of survivors of OOH-CA through hospital discharge, was significantly higher in the CPR + AED compared with CPR-only community units (Table 2).

Dr. Ornato expressed disappointment at the low survival rates in residential units. Although residential units accounted for 15% of all community study units, they only accounted for < 5% of survivors, and outcome was not influenced by the presence or absence of an AED. There was only 1 survival to hospital discharge in each group when OOH-CA occurred in a residential unit (Table 2).

Table 2. Public Access Defibrillation Trial: Primary Endpoint
  CPR Only
(n = 103)
(n = 129)
Number of survivors to discharge 15 (14.6%) 29 (22.5) .042
Residential units 1 1 NS
Nonresidential units 14 28 NS

There were no clinically important adverse device effects, no patient received an inappropriate defibrillation shock, and no device failed to deliver a defibrillation shock when indicated.


Investigators drew the following conclusions from their study results:

  • Trained laypersons can use AEDs safely to provide early defibrillation

  • Survival rate almost doubles when AEDs are added to CPR-trained, volunteer, OOH-CA response systems in public facilities with predefined increased risk

  • The survival rate in multi-unit residential facilities is very low.


Roger D. White, MD, Mayo Clinic (Rochester, Minnesota), described PAD as "a truly landmark study with very important results that we have been awaiting for a long time." He noted, however, that further research is necessary to determine the basic rhythm encountered before defibrillation in patients who were unresponsive to AED shock. Knowing this will have important implications in our understanding of what benefits can be derived from deployment of AEDs in public places. Echoing Dr. Ornato's sentiments regarding defibrillation in residential units, Dr. White stated that not enough information is known about home defibrillation and identified the large, ongoing, multicenter Home AED Trial (HAT) as a means to answer some outstanding questions.

Editorial Commentary

This important study was relatively straightforward in that it found a definite benefit to implementing AEDs in communities with adequate volunteer training and deploying AEDs in places considered to be at high risk for OOH-CA. One issue that remains unresolved by the present study concerns the impact of AEDs in residential units and in private homes, where the majority of these events occur. Further studies will help us understand the true impact of this simple tool in these settings.