BACKGROUND: The mechanisms of blood flow during closed-chest cardiopulmonary resuscitation (CPR) in humans have been debated since the technique was first described in 1960. Two competing models, the cardiac pump theory and the thoracic pump theory, have been proposed, and some investigators have used mitral valve position during the downstroke of chest compression to distinguish between them. Previous studies using either transthoracic or transesophageal echocardiography have yielded conflicting results, and there have been few, if any, hemodynamic or echocardiographic studies on pulmonary venous flow (PVF) during CPR. METHODS AND RESULTS: In this study, transesophageal two-dimensional and pulsed Doppler echocardiography were used to study mitral valve position and flow, together with PVF, in 20 adult patients undergoing manual CPR. In the 17 patients who could be analyzed, the mitral valve closed in 5 patients (group 1) during chest compression but stayed open or opened further in the remaining 12 patients (group 2). Peak forward mitral flow occurred during the release phase in group 1 but during the compression phase in group 2. During chest compression, PVF occurred in the forward direction (from the pulmonary vein to the left atrium) in 8 of the group 2 patients (group 2a) and in the backward direction (from the left atrium to the pulmonary vein) in all group 1 patients and the remaining 4 patients in group 2 (group 2b). The downtime (time from collapse to CPR) was significantly shorter (P < .05) for those in group 1 (7.0 +/- 4.4 minutes) than in groups 2a (19.8 +/- 7.7 minutes) and 2b (17.8 +/- 6.8 minutes). CONCLUSIONS: Transesophageal echocardiography performed during manual CPR in humans disclosed three different patterns of mitral valve position and PVF during chest compression. The presence of an opened mitral valve with forward mitral flow and backward pulmonary venous flow during chest compression in a small number of subjects underscores this heterogeneity in blood flow and suggests the possible existence of a "left atrium pump" in addition to the currently known "left ventricle pump" and "chest pump" mechanisms.
BACKGROUND: The mechanisms of blood flow during closed-chest cardiopulmonary resuscitation (CPR) in humans have been debated since the technique was first described in 1960. Two competing models, the cardiac pump theory and the thoracic pump theory, have been proposed, and some investigators have used mitral valve position during the downstroke of chest compression to distinguish between them. Previous studies using either transthoracic or transesophageal echocardiography have yielded conflicting results, and there have been few, if any, hemodynamic or echocardiographic studies on pulmonary venous flow (PVF) during CPR. METHODS AND RESULTS: In this study, transesophageal two-dimensional and pulsed Doppler echocardiography were used to study mitral valve position and flow, together with PVF, in 20 adult patients undergoing manual CPR. In the 17 patients who could be analyzed, the mitral valve closed in 5 patients (group 1) during chest compression but stayed open or opened further in the remaining 12 patients (group 2). Peak forward mitral flow occurred during the release phase in group 1 but during the compression phase in group 2. During chest compression, PVF occurred in the forward direction (from the pulmonary vein to the left atrium) in 8 of the group 2 patients (group 2a) and in the backward direction (from the left atrium to the pulmonary vein) in all group 1 patients and the remaining 4 patients in group 2 (group 2b). The downtime (time from collapse to CPR) was significantly shorter (P < .05) for those in group 1 (7.0 +/- 4.4 minutes) than in groups 2a (19.8 +/- 7.7 minutes) and 2b (17.8 +/- 6.8 minutes). CONCLUSIONS: Transesophageal echocardiography performed during manual CPR in humans disclosed three different patterns of mitral valve position and PVF during chest compression. The presence of an opened mitral valve with forward mitral flow and backward pulmonary venous flow during chest compression in a small number of subjects underscores this heterogeneity in blood flow and suggests the possible existence of a "left atrium pump" in addition to the currently known "left ventricle pump" and "chest pump" mechanisms.