BackgroundThe mechanisms that underlie cardiopulmonary resuscitation (CPR) in humans remain controversial and difficult to study. This report describes a new human model to evaluate CPR during the first 1 to 2 minutes after the onset of ventricular fibrillation (VF). With this model, standard CPR was compared with active compression-decompression (ACD) CPR, a method that uses a handheld suction device to actively compress and actively decompress the chest.Methods and ResultsDuring routine inductions of VF as part of a transvenous lead cardioverter/defibrillator implantation procedure, CPR was performed in 21 patients if the first defibrillation shock failed and until a successful rescue shock was delivered. Compressions during CPR were performed according to American Heart Association guidelines. For ACD CPR, decompression was performed with up to −30 lbs. Radial arterial and right atrial pressures were measured in all patients. Esophageal pressures, intratracheal pressures, or minute ventilation was measured in the last 13 patients. Application of both CPR techniques increased arterial and right atrial pressures. The mean coronary perfusion pressure was increased throughout the entire CPR cycle with ACD CPR (compression, 21.5±9.0 mm Hg; decompression, 21.9±8.7 mm Hg) compared with standard CPR (compression, 17.9±8.2 mm Hg; decompression, 18.5 ±6.9 mm Hg;P< .02 andP< .02, respectively). Ventilation per compression-decompression cycle was 97.3±65.6 mL with standard CPR and 168.4±68.6 mL with ACD CPR (n=7,P< .001). Negative inspiratory pressure was −0.8±4.8 mm Hg with standard CPR and −11.4±6.3 mm Hg with ACD CPR (n=6,P< .04).ConclusionsPatients undergoing multiple inductions of VF during cardioverter/defibrillator implantation with transvenous leads provide a well-controlled and reproducible model to study the mechanisms of CPR. Using this model, ACD CPR significantly increased arterial blood pressure, coronary perfusion pressure, minute ventilation, and negative inspiratory pressure compared with standard CPR.