Predictors of in-hospital mortality, 20-year survival and reflection on risk scoring in patients with impaired left ventricular function undergoing isolated coronary artery bypass grafting

Abstract

Abstract Background Impaired left ventricular ejection fraction (LVEF) is a risk factor for mortality in patients undergoing coronary artery bypass grafting (CABG). Limited data is available on survival after 5 years and sensitivity of preoperative risk scoring systems in this high-risk population. Purpose To investigate in-hospital mortality, 20-year survival, predictors of early and late mortality and sensitivity of available risk-scoring systems in a large population with moderate to poor baseline LVEF and undergoing isolated CABG. Methods Patients presenting with moderate to poor LVEF and undergoing isolated CABG between 1996 and 2015 were selected for the purpose of this study. LVEF was defined moderate if between 30–39% and poor if <30%. Data collection was prospective and retrospectively analysed. Early and long-term mortality was derived from the NHS National Mortality Tracking system. Exclusion criteria included reoperations and combined procedures only. Patients undergoing urgent and emergency CABG procedures were included. Key outcome measures included 30-day health outcome, long-term survival, impact of incomplete revascularization (IC) and off-pump coronary artery bypass (OPCAB) surgery, independent predictors of poor in-hospital outcome and long-term mortality and reliability of Euroscore in predicting mortality in this high-risk population. Comparative analysis was conducted with Student t-test or Mann-whithney test fro numerical variable and Chi-Square for categorical variables. Survival were compared and reported using Log rank test and Kaplan Meier methods. A Cox Proportional Hazard model was developed to identify risk factors for long term mortality. Results 5016 patients with reduced LVEF were identifiedL1024 (20.4%) with poor and 3992 (79.6%) with moderate LVEF. After excluding reoperations and combined procedures, the final sample consisted of 3867 patients. Average age was68 years, 83% were male, 40% had NYHA IV; 12.5% of patients received IC and 44% underwent OPCAB surgery. 30-day mortality was 4.4%, stroke 1.4% and renal failure requiring dialysis of 2.9%. Survival at 1, 5, 10 and 20 years was 91.1%, 76.7%, 55.1% and 22.1% respectively (Figure). OPCAB surgery was not associated with long-term mortality. Independent predictors of late mortality as shown in Table. Additive Euroscore and Logistic Euroscore appeared to overestimate mortality in this patient population (AUC 0.78). Conclusions In patients with reduced LVEF the use of CABG is associated with low rates of in-hospital mortality, key post-operative complications and excellent long-term survival. Incomplete revascularization but not OPCAB surgery are surgical predictors of long-term mortality. Euroscore appears to overestimate the risk of mortality in this population. Funding Acknowledgement Type of funding sources: None. Cox Hazard Ratio Table