Decision support in surgical management of ischemic cardiomyopathy.

Objectives: The surgical approach to ischemic cardiomyopathy maximizing survival remains a dilemma, with decisions complicated by secondary mitral regurgitation, ventricular remodeling, and heart failure. As a component of decision support, we sought to develop prediction models for comparing survival after coronary artery bypass grafting alone, coronary artery bypass grafting plus mitral valve anuloplasty, coronary artery bypass grafting plus surgical ventricular restoration, and listing for cardiac transplantation.

Methods: From 1997 to 2007, 1468 patients with ischemic cardiomyopathy (ejection fraction <30%) underwent coronary artery bypass grafting alone (n = 386), coronary artery bypass grafting plus mitral valve anuloplasty (n = 212), coronary artery bypass grafting plus surgical ventricular restoration (n = 360), or listing for cardiac transplantation (n = 510). Mean follow-up was 3.8 +/- 2.8 years, with 5577 patient-years of data available for analysis. Risk factors were identified for early and late mortality by using 80% training and 20% validation sets. Outcomes were calculated for each applicable strategy to identify which maximized predicted 5-year survival. Models were programmed as a strategic decision-support tool.

Results: One-, 5-, and 9-year survival were as follows, respectively: coronary artery bypass grafting, 92%, 72%, and 53%; coronary artery bypass grafting plus mitral valve anuloplasty, 88%, 57%, and 34%; coronary artery bypass grafting plus surgical ventricular restoration, 94%, 76%, and 55%; and listing for cardiac transplantation, 79%, 66%, and 54%. Risk factors included older age, higher New York Heart Association class, lower ejection fraction, longer interval from myocardial infarction to operation, and numerous comorbidities. Predicted and observed survivals in validation groups were similar (P > .1). Patient-specific simultaneous solutions of applicable models revealed therapy potentially providing maximum survival benefit. Coronary artery bypass grafting alone and listing for cardiac transplantation often maximized 5-year survival; only 15% of patients undergoing coronary artery bypass grafting plus mitral valve anuloplasty were predicted to fare best with this therapy.

Conclusion: Validated prediction models can aid surgeons in recommending personalized treatment plans that maximize short- and long-term survival for ischemic cardiomyopathy.