R. S. Elsayed1, B. Abt1, W. J. Mack2, A. Liu1, J. K. Siegel1, M. L. Barr1, R. G. Cohen1, C. J. Baker1, V. A. Starnes1, M. E. Bowdish1 1University Of Southern California,Cardiothoracic/Surgery/Keck School Of Medicine,Los Angeles, CA, USA 2University Of Southern California,Preventative Medicine,Los Angeles, CA, USA
Introduction: The need for hemodialysis is a known risk factor for mortality after isolated coronary artery bypass grafting (CABG). This study evaluated outcomes after isolated CABG in hemodialysis-dependent (HDD) and non-HDD patients.
Methods: A retrospective cohort study of 778 patients undergoing isolated CABG between 2006-2016. Patients were grouped by presence or absence of preoperative hemodialysis (696 non-HDD, 82 HDD). Mean follow-up was 20.8±33.3 months. Multivariable logistic regression models were developed to predict 30-day mortality and major adverse cardiovascular events (stroke, myocardial infarction (MI), death, and need for coronary reintervention i.e. MACE). Kaplan-Meier analysis was used to assess survival and multivariable Cox proportional hazard modeling was used to identify factors associated with overall mortality. Propensity scores and 1:1 Greedy matching (1:1) was used to create two groups of 65 non-HDD and 65 HD. Matched groups were compared for the primary outcomes.
Results: Overall survival was 97.9, 96.5, 96.5, and 95.6% at 1, 3, 5, and 7 years. Thirty-day mortality was 2.2% (n=17). On multivariable analysis, 30-day mortality was increased in those with diabetes (OR 7.3, 95% CI 1.0-52.2), COPD (OR 4.5, 95% CI 1.1-18.5), on preoperative inotropes (OR 4.8, 95% CI 1.1-21.2), with increasing cross clamp times (OR 1.04, 95% CI 1.0-1.1). MACE at 30 days was 4.1% (n-32). On multivariable analysis, MACE at 30-days was more common in those with diabetes (OR 4.1, 95% CI 1.3-12.5), COPD (3.7, 1.3-11.0), MI within 30 days (2.8, 95% CI 1.0-7.6), on preoperative inotropes (OR 6.1, 95% CI 1.8-19.8), and with increasing cross clamp times (OR 1.4, 95% CI 1.0-1.1). Thirty-day mortality was 1.9 and 4.9% in the non-HDD and HD groups, respectively (odds ratio (OR) 2.7, 95% confidence intervals (95% CI) 0.86-8.47, p=0.09). Median time on hemodialysis in the HD group was 78 weeks (IQR 48-156). Kaplan-Meier estimates of survival between non-HDD and HD patients showed a significant difference in survival (log-rank p = 0.0008, figure). After multivariable adjustment for age, sex, presence of diabetes, presence of COPD, and history of previous cardiac surgery, mortality was higher in the HDD group as compared to the non-HDD group (Hazard Ratio (HR) 3.1, 95% CI 1.05-9.1, P=0.04). After propensity matching, no survival difference was found between groups (unadjusted HR 4.0, 95% CI 0.47-35.1, p = 0.20).
Conclusions: Overall survival after isolated CABG remains excellent, with decreased survival in those with diabetes, COPD, needing preoperative inotropes, and those with longer aortic cross clamp times. The need for preoperative hemodialysis remains a significant risk factor for long term mortality.
