Objective: To determine the predictors of postoperative AKI following non-emergent cardiac surgery among patients with variable preoperative eGFR levels. Methods: Retrospective study of patients who underwent elective or in-hospital cardiac surgical procedures performed between January 2006 and November 2015. The procedures included isolated CABG, isolated AVR or combined CABG and AVR. The primary outcome AKI (any stage) following non-emergent cardiac surgery utilizing the 2012 KDIGO criteria. Patients were categorized based the following renal outcomes: mild AKI, severe AKI (KDIGO stage 2 or 3) and post-operative dialysis.. Results: A total of 6713 patients were included in our study. The mean age was 66.8 years (SD ± 10.3), with 76.2% being males. A total of 4487 patients had normal or mildly decreased eGFR (G1 or G2) preoperatively (66.8%), while 1960 patients were in the G3 category (29.1%). Only 266 patients (3.9%) had G4 or worse renal function. A total of 1489 (28.5%) patients experienced post-operative AKI. The need for postoperative dialysis occurred in 4.2% of the AKI subgroup. In-hospital mortality was higher among the AKI subgroup (7.3% vs 0.5%, p<0.0001). In an adjusted model, a lower pre-operative eGFR category was the strongest predictor of AKI. A practical scorecard for the preoperative estimation of severe AKI for non-emergent cardiac procedures incorporating these parameters was developed. Conclusions: Preoperative eGFR is the strongest predictor of post-operative AKI in individuals undergoing non-emergent cardiac surgery. A practical scorecard incorporating preoperative predictors of AKI may allow informed decision making and to predict AKI following non-emergent cardiac surgery
