Redefining cardiovascular performance during resuscitation: ventricular stroke work, power, and the pressure-volume diagram.

Abstract

(1) To compare left ventricular stroke work index (SW) and left ventricular power output (LVP), hemodynamic variables that encompass blood pressure as well as blood flow, with the purely flow-derived hemodynamic and oxygen transport variables as markers of perfusion and outcome in critically injured patients during resuscitation. (2) To use the ventricular pressure-volume diagram to define characteristic hemodynamic patterns in the determinants of SW and LVP that are associated with survival. This was a cohort study at a university Level I trauma center during the course of 1 year. A consecutive series of patients was monitored with a volumetric pulmonary artery catheter during the initial 48 hours of resuscitation. Heart rate, SW, LVP, cardiac index, and oxygen delivery and consumption during resuscitation were compared using multivariate logistic regression analysis with regard to the ability to clear lactate in less than 24 hours and survival. Receiver operating characteristic curves were constructed to determine threshold values for SW and LVP. Ventricular pressure-volume diagrams were used to describe characteristic patterns in the determinants of SW and LVP in survivors and nonsurvivors. Preload was expressed as left ventricular end-diastolic volume index, afterload as aortic input impedance (Ea), and contractility as ventricular end-systolic elastance (Ees). The ratio of Ea/Ees (RATIO) was used as a measure of ventricular-arterial coupling, which describes the efficacy of energy transfer from the heart to the vascular system. One hundred eleven patients (87 survivors, 24 nonsurvivors) met study criteria. Survivors had a significantly higher SW (4,510 +/- 1,070 vs. 3,440 +/- 980 mm Hg x mL x m(-2); p < 0.0001) and LVP (370 +/- 94 vs. 270 +/- 81 mm Hg x L x min(-2) x m(-2); p < 0.0001) than nonsurvivors. Heart rate, SW, and LVP were the only studied variables that were significantly related to lactate clearance and survival by logistic regression. Threshold values determined by the receiver operating characteristic curves were 4,000 mm Hg x mL x m(-2) for SW and 320 mm Hg x L x min(-1) x m(-2) for LVP. Survivors had better ventricular-arterial coupling than nonsurvivors, indicated by a lower RATIO (0.32 +/- 0.22 vs. 0.54 +/- 0.38; p = 0.003). This lower RATIO was attributable to lower levels of Ea (2.7 +/- 0.7 vs. 3.4 +/- 0.8 mm Hg x mL(-1) x m(-2); p = 0.0003) and a trend toward higher levels of Ees (13 +/- 11 vs. 9.9 +/- 7.3 mm Hg x mL(-1) x m(-2); p = 0.12). Thermodynamic perfusion variables that encompass both pressure and flow, such as SW and LVP, are more closely related to perfusion and outcome than the purely flow-derived variables. The higher SW and LVP in survivors is related to better ventricular-arterial coupling, and therefore more efficient cardiac function. Cutoff values for LVP of 320 mm Hg x L x min(-1) x m(-2) and for SW of 4,000 mm Hg x mL x m(-2) may be useful thresholds for evaluating hemodynamic performance during resuscitation.