Background and aim of the study: Patients with heart valve surgery may have a periprocedural mortality extending up to one year after surgery. The study aim was to determine independent predictors for in-hospital and long-term mortality after heart valve surgery.
Methods: A total of 1,376 consecutive patients who underwent isolated or combined heart valve surgery at a single institution was studied. Multivariate logistic regression analysis was used to determine independent predictors for in-hospital mortality. Long-term survival data (mean follow up 5.6 years) were obtained from the National Death Index. Multivariate Cox regression analysis was used to determine independent predictors for long-term mortality. All available preoperative, intraoperative and postoperative risk factors were included in these analyses.
Results: The mean EuroSCORE was 6.2 +/- 3.7. There were 86 (6.3%) in-hospital and 550 (40.0%) late deaths. Eleven independent predictors were determined for in-hospital mortality, and 13 for long-term mortality. There were six common independent predictors (preoperative dialysis, total bypass time, intraoperative stroke, postoperative sepsis and/or endocarditis, renal and respiratory failure). Unique independent predictors for in-hospital mortality included intra-aortic balloon pump, preoperative endocarditis, intravenous use of nitroglycerine, bleeding requiring reoperation and gastrointestinal complications. The model for in-hospital mortality showed acceptable calibration (Lemeshow-Hosmer, p = 0.629) and excellent discriminatory ability (C statistic 0.88). Unique independent predictors for long-term mortality included age, ejection fraction, stroke prior to surgery, hemodynamic instability, chronic obstructive pulmonary disease and deep sternal wound infection.
Conclusion: Independent predictors were determined for early and long-term mortality after heart valve surgery. The prevention of postoperative complications may be a key element for increased early and long-term survival in these patients.