Background: eGFR (estimated glomerular filtration rate) formulas may be inaccurate in overweight cardiac surgery patients, overestimating the kidney reserve. The aim of this study was to modify the eGFR formulas and to determine whether the modified eGFR is a more accurate predictor of acute kidney injury (AKI).
Materials and methods: The patients were assigned into 4 BMI groups as follows: normal weight (18.5–25 kg/m2), pre-obesity (25–30 kg/m2), class I obese (30–35 kg/m2), class II and III obese (≥35 kg/m2). Cockcroft–Gault (CG) eGFR formula was modified by using the fat-free mass (FFM) derived from bioelectrical impedance. ROC-AUC curves were analyzed to identify the accuracy of the eGFR formulas (CG, CG modified with FFM, Mayo Clinic Quadratic equation, CKD-EPI, MDRD) to predict the AKI in each group.
Results: Although all of the used equations showed similar predictive power in the normal weight and overweight category, Mayo formula had the highest AUC in predicting the occurrence of AKI (ROC-AUC 0.717 and 0.624, p<0.05). However, in the group of patients with class I obesity, only the CG formula modified with a fat-free mass appeared to be predictive of postoperative AKI (ROC-AUC 0.631 p<0.05). None of the equations were accurate in the group of BMI (>35 kg/m2).
Conclusions: eGFR is a poor predictor of AKI, especially in the obese patients undergoing cardiac surgery. The only equation with a moderate predictive power for the class I obese patients was the CG formula modified with the fat-free mass.