Risk of de novo dialysis after elective cardiac surgery according to preoperative renal dysfunction

OBJECTIVES: The aim was to estimate the risk of dialysis postoperative de novo dialysis in patients undergoing elective cardiac surgery, according to varying degrees of pre-existing renal dysfunction, and to compare the outcomes with the expected prevalence of dialysis based on several risk scores. METHODS: A retrospective analysis was performed over a 5-year period (2012-16) from a series of 1332 adult patients who underwent elective cardiac surgery. Patients were divided into 3 estimated creatinine clearance (eCrCl) groups: eCrCl > 60 ml/min, eCrCl from 50 ml/min to 60 ml/min and eCrCl <= 49 ml/min. The primary outcome was any renal failure requiring first-time dialysis during the postoperative hospital stay. The expected risk for postoperative dialysis was calculated with 3 predictive scores: the Society of Thoracic Surgeons Bedside Risk Tool, the Cleveland Clinic Score and the Simplified Renal Index. The global accuracy of eCrCl and the different scores was evaluated in terms of calibration and discrimination. RESULTS: In the overall population, 26.6% of patients presented moderate or severe pre-existing renal dysfunction, and the need for de novo dialysis varied from 0.6% to 5.0% depending on the degree of preoperative eCrCl (P < 0.0001). Preoperative renal dysfunction with eCrCl < 50 ml/min was associated with nearly 8-fold increase of risk for postoperative dialysis; eCrCl between 50 ml/min and 60 ml/min showed a 2-fold increase, though this last difference did not reach statistical significance. CONCLUSIONS: In elective cardiac surgery, isolated preoperative eCrCl seemed to be an accurate indicator of risk for postoperative de novo dialysis. More complex models did not provide additional information to stratify that risk.