8th International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV San Francisco, California - September 24 - 26, 2006

SYSTEMATIC ERRORS IN ESTIMATING RENAL FUNCTION BY COCKCROFT-GAULT (CG) OR MODIFICATION OF DIET IN RENAL DISEASE (MDRD) EQUATIONS

Antiviral Therapy 2006; 11:L13 (abstract no. 18)

DP Kotler¹, A Glyptis¹, C Grunfeld², M Pang², ES Engelson¹ and EM Sordillo^1
1St. Luke’s-Roosevelt Hospital Center, New York, NY, USA; ²University of California San Francisco, San Francisco, CA, USA

INTRODUCTION: Assessment of renal function is important in clinical medicine when drug safety and efficacy are affected by variations in renal function. Several prediction equations have been applied clinically, and are based upon age, sex, race, height, weight, and serum creatinine concentration. We examined the effects of nutritional alterations on the accuracy of these predictions.

METHODS: Studies were performed in 122 subjects without intrinsic renal disease, 97 men and 25 women, of whom 79 were HIV-infected inpatients or outpatients and 43 were healthy controls. The values for BMI ranged from malnourished to obese (12.9–40.6 kg/m²). Creatinine clearance (CrCl) was estimated using the CG and MDRD equations which were compared to each other and to CrCl determined from a 24 hour urine collection (CrCl 24). For the MDRD calculation, the prediction was adjusted to the estimated total body surface area, calculated by Mosteller’s formula, rather than normalized to 1.73 m². Body cell mass was estimated using bioelectrical impedance analysis (BIA).

RESULTS: CrCl 24 was significantly associated with estimates of body cell mass (r²=0.26, P<0.001), but was not significantly associated with BMI (r²=0.02, P=0.1). CrCl 24 was significantly associated with MDRD or CG estimates, using either actual or ideal body weight (r² range 0.31–0.35, P<0.001), while CG and MDRD estimates correlated closely with each other (r²=0.78, P<0.001). The prediction errors by CG and MDRD were directly related to the magnitude of CrCl 24, with overestimation at low values and underestimation at high values (r² range 0.34–0.38, P<0.001). The prediction error for CrCl by CG was related to BMI (P=0.001), and the prediction errors for CrCl by CG and MDRD were related to body cell mass, (P=0.002 and P<0.001, respectively), with overestimation at low values and underestimation at high values.

CONCLUSIONS: CrCl is affected by nutritional status, specifically body cell mass, of which skeletal muscle is the largest component. Estimations of CrCl using height, weight, serum creatinine concentration, and demographics contain systematic errors related to nutritional status, which may lead to overestimation of CrCl in underweight or malnourished subjects and underestimation in obese subjects, with potential effects on drug safety and efficacy.

Download PDF of this abstract.

2006-09-24
18

Copyright © 2006 - International Medical Press Ltd. Reproduction of this abstract (other than one copy for personal reference) must be cleared through the Medical Editor, International Medical Press, 36 St Mary-at-Hill, London EC3R 8DU, United Kingdom.