Factors influencing postoperative survival in aortic regurgitation. Analysis by Cox regression model

From 1965 to 1981, 114 patients underwent aortic valve replacement for severe aortic regurgitation in our institution. Catheterization was performed preoperatively in 103 patients. Follow-up was possible in 98% of the survivors. Long-term survival was significantly different between patients in preoperative Functional Class I-II and those in Class III-IV (p less than 0.03); those with a preoperative cardiothoracic ratio less than 0.64 versus greater than or equal to 0.64 (p less than 0.001); and those with a preoperative ejection fraction greater than 0.50 versus less than or equal to 0.50 (p less than 0.03). A multifactorial analysis was used to identify the dominant preoperative prognostic variables affecting survival. Three of the 13 parameters examined simultaneously were found to independently influence survival rates: cardiothoracic ratio (p = 0.001), strain pattern on the electrocardiogram (p = 0.072), and left ventricular end-systolic pressure (p = 0.127). After stratification of the population into two groups according to preoperative functional class, the predictive variables were cardiothoracic ratio (p = 0.014), strain pattern (p = 0.050), and acute/chronic form of aortic regurgitation (p = 0.034). This statistical analysis enabled us to derive a mathematical equation for predicting an individual patient's probability of survival. We found a close fit between the survival rate predicted by the mathematical model and the observed survival rate.