Comparison of bedside estimates of renal function and measured glomerular filtration rate (GFR) in 510 adult oncology patients

Abstract

19512 Background: Various bedside formulae have been used in practice to estimate renal function to predict drug dosing. Recently the ‘4-variable Modification of Diet in Renal Disease’ (4-v MDRD) equation1 was derived in patients with chronic renal disease and has been advocated for application in oncology. The aim of this study was to compare measured GFR with estimates from formula based equations in adult oncology patients. Methods: GFR was determined using technetium-99m diethyl triamine penta-acetic acid (Tc99mDTPA) clearance, serum creatinine (Jaffe method) was measured and renal function estimates calculated using 4-v MDRD, Cockcroft and Gault (CGF), Wright, Martin, and Jelliffe (JF) formulae. Accuracy, bias (mean % error (MPE)) and precision were assessed for varying levels of GFR, body mass index (BMI), age and gender. Results: In 510 adult oncology patients (323 male, 187 female, mean age 63years, range 17–87years) GFR was determined using Tc99mDTPA clearance (mean 84mL/min, range 16–205mL/min). The mean (range) calculated GFR was 72mL/min/1.73m2(9–162mL/min/1.73m2), 71ml/min (11–267mL/min), 78mL/min (12- 195mL/min), 81mL/min (12–279mL/min), 64mL/min/1.73m2 (10–165mL/min/1.73m2) for 4-v MDRD, CGF, Wright, Martin, and JF formula respectively. Bias, precision and accuracy (%within 30% and 50% of true GFR) of estimates are shown in the table . The Wright and Martin formulae had greater bias relating to degree of renal function and gender respectively. The 4-v MDRD equation provided a less biased estimate compared to the CGF across all levels of renal function and BMI. Conclusions: When compared to measured GFR, the 4-v MDRD equation provides a less biased estimate compared to the other formulae evaluated across a range of variables including degree of renal function and BMI. The limitations of all the bedside estimates must be understood to allow appropriate clinical utility. 1. Levey AS, et al. Ann Intern Med 2006; 145: 247–254. [Table: see text] No significant financial relationships to disclose.