How reliable is eGFR when calculating drug dosage in acute medical admissions?

The Modification of Diet in Renal Disease-derived estimation of glomerular filtration rate (eGFR) is used widely. Although validated in stable chronic kidney disease (CKD) outpatients, it is not known how it performs in those presenting with acute medical illness.

We aimed to compare eGFR with Cockroft Gault (CG) - the renal function assessment tool available prior to eGFR - to assess the difference in clinical outcome that would occur when one over another estimation is used in practice. In particular, we wished to assess whether use of eGFR would have resulted in a change of dose of commonly used acutely administered medications.

Acute medical admissions presenting to a tertiary hospital between August and December 2008 were included. Serum creatinine concentration, age, sex, height and weight were collected. Renal function was estimated by both estimates. Movement from CKD class 3 to 4 or 5 was measured - a clinically used cut-off point for changes in management.

A total of 54 patients was included. eGFR values were higher than those estimated by CG. Almost half of patients categorized as CKD stage 4-5 using CG were only categorized as CKD stage 3 using eGFR.

Although we did not use a gold standard estimation of GFR, this study shows that estimates of renal function vary in a clinically significant manner. As estimates of GFR are used to adjust drug dosages and to stratify for many other treatments, it is imperative that we find a method of estimating kidney function that is readily available, consistent and accurate.