A simplified Cockcroft-Gault formula to improve the prediction of the glomerular filtration rate in diabetic patients.

Aim: The National Kidney Foundation recommends stratification of renal failure into moderate (Glomerular Filtration Rate: GFR = 30-60 mL/min/1.73 m2), severe (15-30) or terminal (<15) using the Cockcroft-Gault (CG) or the Modification of Diet in Renal Disease (MDRD) equations. We studied the biases in these methods in an attempt to improve the standard CG (MCG) and devise a strategy for stratification.

Methods: GFR was measured by 51Cr-EDTA clearance in 200 diabetic patients: 100 (Group 1: study of concordance) before 2003 and 100 thereafter (Group 2: validation of MCG). The CG was modified by replacing body weight by its mean value: 76.

Results: In group 1, the recommended equations only correctly stratified 50 patients. The CG, not the MDRD, underestimated GFR if BMI was normal, and overestimated it in obese patients. In group 2, the MCG was well correlated with GFR and not biased by weight. Over the whole population, the MCG and MDRD were more accurate for the diagnosis of moderate and severe renal failure. The MDRD showed the lowest differences with GFR, except if GFR > 60, where the MCG performed better. All formulae overestimated low GFR, the MDRD also underestimated high GFR. The best stratification (147/200) was obtained using the MCG if creatininemia < 120 micromol/l and the MDRD if creatininemia > or =120 micromol/l.

Conclusion: The CG is biased by weight, the MCG corrects this. The more accurate MDRD cannot be used in all patients as it underestimates high GFR. The best stratification was obtained using the MCG at low and the MDRD at high creatininemia.