The major focus of this analysis was to assess the predictive value of right-sided hemodynamic measurements independent of systolic LV function and the severity of coronary atherosclerosis. For this purpose, LVEF was taken as a measure of LV systolic function and Cox regression models were constructed to estimate the relative contribution of pulmonary hemodynamics to the prediction of death. In the univariate Cox regression model, the expected predictors of survival were confirmed, namely male sex, LVEDP, LVEF, and the number of vessels stenosed. In addition, however, PAMP was highly related to survival, being associated with the largest x2 value of any of the angiographic variables (Table 3). In the stepwise regression model, three variables—PAMP the number of stenosed vessels, and the LVEF—maintained statistical significance (Table 3). Among these variables, PAMP was associated with the greatest predictive power, entering the model first and achieving the largest standardized J3-coefficient. Cheap Diskus Advair
Additional analyses were performed using pulmonary wedge pressure, pulmonary vascular resistance, and right atrial pressure. Although these variables were predictive of a fatal event in the univariate model (X2 to enter = 12.4 for pulmonary wedge pressure, 9.7 for pulmonary vascular resistance, and 10.9 for right atrial pressure), only right atrial pressure was retained in the stepwise multivariate model, including LVEF and number of stenosed vessels. Thus, despite the high degree of intercorrelation, PAMP was a more powerful predictor of mortality than right atrial or pulmonary wedge pressure.
Table 3—Cox Stepwise Regression Analysis of Determinants of Survival after Cardiac Catheterization (N —1371)
|A. Univariate Model||VariablePAMP
Ejection fraction No. of vesselst
|X! to enter 47.4 39.3 28.8||p value.000.000.000|
|B. Multivariate Model|
|No. of vesselst||19.1||-3.2||.000||9.8|