Compression-Only CPR

Abstract

FOR MORE THAN A DECADE, CARDIOPULMONARY RESUScitation (CPR) using only chest compressions, forgoing ventilations by rescue breathing, has been described as an option in the out-of-hospital management of sudden cardiac arrest when bystanders are unwilling or unable to provide standard CPR that combines chest compressions with rescue breathing. The “Guidelines 2005” CPR standards promulgated by the American Heart Association (AHA) and the International Liaison Committee on Resuscitation (ILCOR) indicate that “the best method of CPR is compressions coordinated with ventilations.” However, the guidelines also note that laypersons who are unable or unwilling to provide rescue breaths should be encouraged to provide compressiononly CPR to individuals experiencing out-of-hospital cardiac arrest. This recommendation followed a 1997 AHA science statement that was the first document to propose compression-only CPR as a reasonable alternative to standard CPR by bystanders. Several fairly compelling theoretical arguments support the concept of compression-only CPR. First, it seems likely that compression-only CPR is easier to teach and remember than standard CPR, particularly for those who do not work in health care professions. Second, it has been postulated that bystanders may prefer to avoid mouth-to-mouth contact with persons who sustain cardiac arrest, particularly when those persons are strangers. However, in a study of actual bystanders of out-ofhospital cardiac arrest, very few (1.1%) mentioned reluctance to perform rescue breathing, and none reported fear of contracting infection. Third, and perhaps most compelling, an increasing body of evidence suggests that interruptions in chest compressions, even for such seemingly important interventions as providing artificial ventilation, are detrimental. Forward flow of blood ceases very soon after chest compressions are halted, and several compressions are needed to reestablish perfusion when compressions are resumed. The “push hard, push fast, don’t stop” mantra of current CPR teaching is designed to reinforce the need for minimal interruptions in chest compressions to maintain some degree of perfusion to the vital organs until more definitive therapy (such as defibrillation) can be delivered. Despite these theoretical advantages, most animal studies and human trials comparing standard CPR and compression-only CPR have reported similar outcomes for the 2 techniques; data showing a convincing survival advantage to the compression-only technique are lacking. Additionally, most of the existing human data are from observational studies, with few experimental trials in the published literature. Randomization is one of several significant barriers to performing these trials, particularly when examining CPR provided by bystanders who witness the collapse and provide immediate intervention while emergency medical services (EMS) personnel are summoned. Although it is possible to randomize the type of CPR provided by EMS personnel responding to an out-of-hospital cardiac arrest, or by a team responding to an in-hospital cardiac arrest, it is nearly impossible to randomize the type of CPR provided by a bystander at the scene of an out-of-hospital cardiac arrest. One elegant solution to this problem was used in a 2000 study by Hallstrom et al and in 2 recently published studies. Randomization was performed by 9-1-1 dispatchers for callers who were willing to perform CPR but did not know how to do so. In each study, the survival rate for patients who experienced out-of-hospital cardiac arrest and received dispatcher-assisted standard CPR was similar to that for those with cardiac arrest who received dispatcherassisted compression-only CPR, despite data showing that chest compressions are delivered sooner when the dispatcher omits the ventilation instructions, providing another theoretical benefit to compression-only CPR. However, in these 3 studies, no CPR was delivered from the time the patient collapsed until the time the caller recognized the emergency, reached 9-1-1, completed the questions in the dispatcher’s algorithm, was randomized by the dispatcher, and received CPR instructions. This no-perfusion time, typically longer than 2 minutes, may have contributed to the findings.