Impact of Estimated Glomerular Filtration Rate After Valve and Combined Valve and Coronary Surgery

Preoperative renal impairment predicts postoperative mortality in patients undergoing cardiac surgery. We examined the impact of three equations for glomerular filtration rate (GFR) estimation on mortality. This is a retrospective, observational study on patients undergoing valve (n = 594) and combined valve and coronary procedures (n = 412). Glomerular filtration rate was estimated with modification in diet and renal disease (MDRD), Cockcroft-Gault (CG) equation, and Mayo-Clinic (MC) equation. Mean age was 65 +/- 11 years and mean logistic EuroSCORE was 9.0 [95% confidence interval (CI): 8.4-9.7]. Preoperative kidney dysfunction (eGFR < 60 ml/min/1.73 m(2)) is associated with increased risk for mortality. Hazard ratios for 30-day mortality differed between eGFR equations [MC 2.6 (95% CI: 1.6-4.1); MDRD 2.9 (95% CI: 1.9-4.4); CG 3.9 (95% CI: 2.2-5.2)]. Discriminatory performance of GFR equations was comparable with logistic EuroSCORE, but CG and MC discriminated significantly better than MDRD. Median intensive care unit stay did not differ, but Bland-Altman plots showed considerable variation between eGFR equations, indicating that they cannot be used interchangeably. Substantial amount of agreement was achieved with CG and MDRD. Only one equation should be used in one institution. The bias between equations varied considerably. Impaired preoperative kidney dysfunction is associated with increased risk for mortality irrespective of the formula used. ASAIO Journal 2010; 56:543-549.