Predictors of perioperative morbidity and mortality in repeat valve replacement: a seven-year experience

Forty-nine consecutive repeat cardiac valve replacements in 46 patients were reviewed to define in-hospital morbidity, mortality, and determinants of risk. The overall operative mortality rate was 10.2%; it was 9% for the first reoperation and 25% for the second (four patients). The mortality rate was 12% at the mitral position (25 patients), 11% at the aortic position (18 patients), and 0% for repeat double valve replacement (four patients). Univariate and multivariate stepwise logistic regression models show that age over 70 years and the use of an intraaortic balloon pump preoperatively predict operative mortality, and age over 70, preoperative care in the coronary care unit, endocarditis, and the presence of an intraaortic balloon pump are determinants of perioperative morbidity in patients undergoing repeat valve replacement. Based on these data, close follow-up of patients with initial valve replacement is encouraged so that reoperation to correct prosthetic malfunction or periprosthetic leak can be done before hemodynamic collapse occurs. Furthermore, since age was the most significant factor affecting mortality (P < .01) and since most of the deaths (80%) in this study occurred in patients with torn bioprosthetic leaflets, the use of bioprostheses for initial valve replacement in patients under 70 years of age is discouraged.