Accuracy of the estimation of glomerular filtration rate within a population of critically ill patients

Abstract

Accuracy of glomerular filtration rate (GFR) estimates has been questioned and several authors recommend routine use of measured renal creatinine clearance (CLCR) as a surrogate of GFR in the intensive care unit (ICU). Our purpose was to compare estimates of GFR using Cockroft-Gault (CG), Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) and Modification of Diet in Renal Disease Study (MDRD) equations with 8h-CLCR, within a population of critically ill patients with a wide range of measured CLCR. Through a prospective, observational study of 54 patients with normal serum creatinine (sCr) admitted to ICU, daily 8h-CLCR (reference method) and GFR estimates (644 paired samples) were matched and compared. Augmented renal clearance (ARC) was defined as 8h-CLCR >130 ml/min/1.73 m(2). No significant difference was found between mean 8h-CLCR (135.5 ml/min/1.73 m(2)) and CG equation (135.7 ml/min/1.73 m(2)), but significant differences (p < 0.01) were found for the MDRD (124.4 ml/min/1.73 m(2)) and CKD-EPI (107.6 ml/min/1.73 m(2)) equations. Correlation between 8h-CLCR and all estimates was weak (R = 0.2, 0.19 and 0.34, respectively). We observed poor agreement in terms of precision (40.9, 39.8 and 33.4%, respectively). Analysing subgroups, we observed that all equations significantly underestimated 8h-CLCR >120 ml/min/1.73 m(2) and overestimated 8h-CLCR <120 ml/min/1.73 m(2) (p < 0.05). The incidence of ARC patients was 55.6%. Estimates of GFR using CG, CKD-EPI and MDRD formulae are flawed in the critically ill with normal sCr, significantly underestimating renal function in those with ARC and overestimating it in those with normal or decreased 8h-CLCR. Globally, the population exhibited ARC on more than half of the ICU admission days.