Abstract
To evaluate the impact of adding first-responder defibrillation by fire-fighters to an existing advanced life-support emergency medical services system. Nonrandomized, controlled clinical trial with periodic crossover. Memphis, Tenn, a city of 610,337 people, which is served by a fire department-based emergency medical services system. All city ambulances provide advanced life support. Adult victims of out-of-hospital cardiac arrest due to heart disease. Twenty of 40 participating engine companies were equipped with an automated external defibrillator and ordered to apply it immediately in all cases of cardiac arrest. The other 20 companies were ordered to start cardiopulmonary resuscitation (CPR) immediately and wait for paramedics to arrive. Every 75 days, group roles were reversed. Care otherwise proceeded according to 1986 American Heart Association guidelines. Return of spontaneous circulation in the field, survival to hospital admission, survival to hospital discharge, and neurological status at discharge. During the 39-month study interval, 879 patients were treated by a project engine company. Four hundred thirty-one (49%) of these were found in ventricular fibrillation. Bystander CPR was started in only 12% of cases. Overall, firefighters reached the scene a mean of 2.5 minutes faster than simultaneously dispatched paramedics. Although our automated external defibrillators proved to be reliable and efficacious for terminating ventricular fibrillation and pulseless ventricular tachycardia, patients treated by an automated external defibrillator-equipped engine company were no more likely than CPR-treated controls to be resuscitated (32% vs 34%, respectively), to survive to hospital admission (31% vs 29%), or to survive to hospital discharge (14% vs 10%). Neurological outcomes were also similar in the two treatment groups. In a fast-response, urban emergency medical services system served by paramedics, the impact of adding first-responder defibrillation appears to be small. Early defibrillation alone cannot overcome low community rates of bystander CPR. Careful attention to every link in the "chain of survival" is needed to achieve optimal rates of survival after cardiac arrest.