Supplementary Material for: Serum Creatinine and Chronic Kidney Disease-Epidemiology Estimated Glomerular Filtration Rate: Independent Predictors of Renal Replacement Therapy following Cardiac Surgery

Background: Most studies evaluating predictors of renal replacement therapy (RRT) following cardiac surgery use arbitrary defined limits of preoperative serum creatinine. The aim of this study was to evaluate the effect of preoperative renal function using either estimated-glomerular filtration rate (eGFR) derived using Chronic Kidney Disease-Epidemiology (CKD-EPI) or serum creatinine alone as a predictor for RRT after cardiac surgery. Methods: In this prospective cohort study, baseline, intraoperative, and postoperative data of all patients who underwent an elective, urgent, or emergency cardiac surgery between 2012 and 2016 in a single center were analyzed in order to identify multivariate parameters determining the need for RRT after surgery. For preoperative renal function, we used serum creatinine levels and eGFR-derived CKD-EPI equation. We also divided our cohort into eGFR groups following the thresholds of the currently proposed CKD classification. Results: From the 1,614 patients (mean age: 65.4 ± 10.6 years; male: 77.6%) that constituted the study population, 42 (2.6%) underwent RRT postoperatively. EUROSCORE II, cardiopulmonary bypass time, cross clamp time, red blood cell (RBC) units transfused, type and urgency of surgery, combined/non combined operation, peripheral vascular disease, heart failure, chronic obstructive pulmonary disease, dyslipidemia, and preoperative renal function were all univariately associated with RRT use. Multivariate regression with bootstrap utilization indicated that CKD-EPI eGFR (OR 0.979; 95% CI 0.956–0.998), heart failure with the New York Heart Association class ≥2 (OR 4.695; 95% CI 1.756–14.061) and RBC units transfused (OR 1.287; 95% CI 1.081–1.850) were independently associated with RRT need. When serum creatinine (OR 2.920, 95% CI1.056–8.074) was used in the model, the associations with RRT were also significant. Conclusion: Preoperative renal function, defined by serum creatinine or eGFR by CKD-EPI, NYHA class II–IV, and the number of blood units transfused were all independent predictors of RRT postoperatively.