Mechanisms of blood flow during pneumatic vest cardiopulmonary resuscitation.

Mechanisms of blood flow during cardiopulmonary resuscitation (CPR) were studied in a canine model with implanted mitral and aortic flow probes and by use of cineangiography. Intrathoracic pressure (ITP) fluctuations were induced by a circumferential pneumatic vest, with and without simultaneous ventilation, and by use of positive-pressure ventilation alone. Vascular volume and compression rate were altered with each CPR mode. Antegrade mitral flow was interpreted as left ventricular (LV) inflow, and antegrade aortic flow was interpreted as LV outflow. The pneumatic vest was expected to elevate ITP uniformly and thus produce simultaneous LV inflow and LV outflow throughout compression. This pattern, the passive conduit of "thoracic pump" physiology, was unequivocally demonstrated only during ITP elevation with positive-pressure ventilation alone at slow rates. During vest CPR, LV outflow started promptly with the onset of compression, whereas LV inflow was delayed. At compression rates of 50 times/min and normal vascular filling pressures, the delay was sufficiently long that all LV filling occurred with release of compression. This is the pattern that would be expected with direct LV compression or "cardiac pump" physiology. During the early part of the compression phase, catheter tip transducer LV and left atrial pressure measurements demonstrated gradients necessitating mitral valve closure, while cineangiography showed dye droplets moving from the large pulmonary veins retrograde to the small pulmonary veins. When the compression rate was reduced and/or when intravascular pressures were raised with volume infusion, LV inflow was observed at some point during the compressive phase. Thus, under these conditions, features of both thoracic pump and cardiac pump physiology occurred within the same compression. Our findings are not explained by the conventional conceptions of either thoracic pump or cardiac compression CPR mechanisms alone.