A comparison of bedside renal function estimates and measured glomerular filtration rate (Tc99mDTPA clearance) in cancer patients

Abstract

Background: The aim of this study was to compare measured glomerular filtration rate (GFR) with estimates of GFR derived from the population pharmacokinetic methods of Martin and Wright, and the creatinine clearance (CrCl) estimates of Cockcroft and Gault, and Jelliffe. Patients and methods: GFR was determined by technetium-99m diethyl triamine penta-acetic acid (Tc99mDTPA) clearance in adult cancer patients. Height, actual body weight and serum creatinine were measured, and GFR and CrCl estimates calculated. Results: One hundred and twenty-two patients were included. The mean measured GFR was 87 ml/min (range 30-174 ml/min). The mean bias (mean percentage error) was 2, 1, -10 and -17%, and the mean precision (mean absolute percentage error) was 18, 19, 21 and 23% for the Wright, Martin, Cockcroft and Gault, and Jelliffe formulas, respectively. The Martin formula significantly underestimates GFR for females (mean bias -10%) and overestimates GFR for males (mean bias 8%) (P <0.001 for bias of males versus females). The Wright and Martin formulas significantly overestimate GFR <50 ml/min (mean bias 39 and 30%; P = 0.03 and 0.05, respectively) and all formulas underestimate GFR <100 ml/min (mean bias -18, -16, -24 and -32% for Wright, Martin, Cockcroft and Gault, and Jelliffe formulas, respectively; P <0.001). Conclusions: All the assessed estimates for renal function were found to have significant limitations. © 2002 by the Infectious Diseases Society of America.