The effect of time since training on house officers' retention of cardiopulmonary resuscitation skills.

Currently, the American Heart Association (AHA) recommends that physicians be certified in cardiopulmonary resuscitation (CPR) every two years. This study was undertaken to determine the effects of time since training on retention of CPR skills of physicians and to identify at what point performance deteriorates to a level requiring retraining with supervised practice. The physicians' performance of CPR one year or less after training was compared with that of more than one year after training. Thirty-three medical residents who had been taught CPR by the same instructor were tested without warning for one-person CPR on a recording mannikin. Performance was evaluated according to AHA Heartsaver criteria. The data were analyzed by organizing all CPR steps or behavioral objectives into three categories: assessment, skills (which included ventilation and compression), and sequence (which included calls for assistance). The data suggest that the knowledge of CPR sequence remains stable and that assessment improves while skill performance deteriorates after one year. This apparent contradiction in overall CPR performance may relate to the effect of experience. Assessment may improve because of involvement in actual resuscitations in the hospital. Deterioration of skills may reflect the fact that senior residents do not actually perform CPR, but become team leaders and thereby lose their skills, or that poor performance is not corrected in actual "code" situations. If a two-year certification standard is maintained, CPR skill testing at least every 12 months should be considered. If skills have deteriorated, hands-on-practice should be undertaken at that time.