Reassessment of kidney function equations in predicting long-term survival in cardiac surgery

Background/objectives: Chronic kidney disease (CKD) is a risk factor for long-term survival in cardiac surgery. The Cockcroft-Gault, Modification of Diet in Renal Disease (MDRD) study, CKD Epidemiology Collaboration (CKD-EPI), revised Lund-Malmö (LM), and full age spectrum equations are used to estimate glomerular filtration rates (eGFR), but each have advantages and disadvantages. Our objective was to determine which equation better predicts long-term survival.

Methods: Data on 1492 consecutive patients who underwent isolated off-pump coronary artery bypass surgery between September 1996 and December 2008 were prospectively collected. Preoperative and postoperative eGFR were calculated using the five equations and compared using Cox regression analyses and time-dependent receiver operating characteristic (ROC) curves at 10 years.

Results: In a Cox regression model after correction for significant predictors of long-term mortality, adjusted hazard ratios (HR) for one standard deviation increase in preoperative eGFR were 0.661 (P < .0001), 0.844 (P = .0166), 0.787 (P = .0002), 0.746 (P < .0001), and 0.717 (P < .0001) for the CG, MDRD, CKD-EPI, LM, and FAS equations, respectively. The areas under the time-dependent ROC curve at 10 years also showed that the CG formula has a better predictive value. Postoperative eGFR at discharge were also significant predictors of long-term mortality (HR = 0.603, P < .0001; HR = 0.725, P < .0001; HR = 0.688, P < .0001; HR = 0.673, P < .0001; HR = 0.632, P < .0001 for the CG, MDRD, CKD-EPI, LM, and FAS equations, respectively).

Conclusions: The CG formula was shown to better predict survival in cardiac surgery, though the FAS equation has a comparable prognostic value. Additionally, postoperative eGFR at discharge also predicted long-term survival.