Evaluation of glomerular filtration rate using iohexol plasma clearance in critically ill patients with augmented renal creatinine clearance: A single-centre retrospective study.

ABSTRACT

BACKGROUND: Augmented renal creatinine clearance (ARC) (≥130 ml min-1 1.73 m-2) is frequent in intensive care unit (ICU) patients and may impact patient outcome. OBJECTIVES: To compare glomerular filtration rate (GFR) measured with iohexol plasma clearance and creatinine clearance in critically ill patients with augmented renal clearance. DESIGN: Single-centre, retrospective study. SETTING: French University Hospital ICU from November 2016 to May 2019. PATIENTS Adult patients with augmented renal clearance who had a measurement of iohexol plasma clearance. MAIN OUTCOME MEASURE: Agreement between 6 h creatinine clearance (6 h CrCl) and iohexol plasma clearance (GFRio). RESULTS: Twenty-nine patients were included. The median 6 h creatinine clearance was 195 [interquartile range (IQR) 162 to 251] ml min-1 1.73 m-2 and iohexol clearance was 133 [117 to 153] ml min-1 1.73 m-2. Sixteen patients (55%) had hyperfiltration (clearance >130 ml min-1 1.73 m-2) measured with iohexol clearance. Mean bias between iohexol and creatinine clearance was -80 [limits of agreement (LoA) -216 to 56 ml min-1 1.73 m-2]. For Cockcroft and Gault Modification of Diet in Renal Disease equation (MDRD), Chronic Kidney Disease Epidemiology Collaboration equation (CKD-EPI) formulae, mean biases were, respectively -27 (LoA -99 to 45), -14 (LoA -86 to 59) and 15 (LoA -33 to 64) ml min-1 1.73 m-2. CONCLUSION: In the present study, we found that in patients with augmented renal creatinine clearance, half of the patients do not have hyperfiltration using iohexol clearance measurements. We observed an important bias between 6 h CrCl and GFRio with large LoA. In critically patients with ARC, 6 h CrCl does not reliably estimate GFR and 6 h CrCl nearly systematically overestimates renal function. Comparison of creatinine-based GFR estimations and GFRio show acceptable bias but wide LoA.