Modified diet in renal disease method overestimates renal function in selected elderly patients.

BACKGROUND the Modification of Diet in Renal Disease (MDRD) method of renal function estimation has not been extensively assessed in elderly patients. We needed to assess which renal function estimate was most suited for drug dose estimation in our population. METHOD we compared MDRD with an optimised version of the Cockcroft-Gault (CG(opt)) method in a hospital population, using gentamicin clearance as a baseline. RESULTS MDRD overestimated gentamicin clearance by 29% (P < 0.001, n = 68), while CG(opt) underestimated by 10% (P < 0.01). Overestimation by MDRD increased with increasing age. This was 12%, 26% and 69% in age groups <65, 65-80 and >80 years respectively (P < 0.001). CG(opt) underestimated renal function by -5%, -16% and -4% respectively (P = NS). Bias and precision of renal function estimations for the three age groups were less for CG(opt) than for MDRD. Age significantly influenced MDRD overestimation in this population (P = 0.037). CONCLUSION MDRD overestimated renal function as age increased. While CG(opt) underestimated renal function, this was of a smaller magnitude, consistent across age, and thus better suited for dose calculation, especially in the elderly. Larger-scale studies using gold standard markers of renal function estimation are urgently needed to determine the accuracy of MDRD in elderly hospitalised patients.