Patterns of external chest compression.

We studied the performance of external chest compression by 20 fourth year medical students on 2 study days, separated by 5-7 weeks, 4-8 months after they had been initially trained in cardiopulmonary resuscitation (CPR). Our hypotheses were (1) that a given individual would perform external chest compressions in the same manner each time CPR was performed and (2) that the pattern of performance of chest compressions would be determined, in part, by the anthropometric characteristics of the rescuer. A Laerdal Skillmeter Resusci-Anne CPR manikin chest compression transducer was interfaced with an analogue-to-digital conversion board in a Macintosh LC III computer. Each subject on each study day performed four cycles of 15 compressions and two ventilations twice, once on a table (which raised the surface of the manikin sternum to 95 cm) and once on the floor. For each individual, on each attempt, we calculated the depth and rate of compressions, duty cycle, peak compression velocity, time to peak compression velocity and time to peak compression depth. In addition, we calculated the regression slope of compression depth versus compression number for each cycle of 15 chest compressions and over four cycles of compressions. Statistically significant correlations were observed between the first and second study days in each of the variables of chest compression measured, indicating that the performance of chest compressions was constant over time for a given individual. We observed that the depth of compression, duty cycle, time to peak compression, time to peak velocity and regression slope of depth of compressions versus compression number were significantly related to the height and weight of the rescuer.