Creatinine Clearance in Acute Brain Injury: A Comparison of Methods.

ABSTRACT

BACKGROUND: Currently, the measurement of glomerular filtration rate is very complex and costly, so its daily evaluation is performed using endogenous markers, of which creatinine is the most frequently used. It allows the estimation of glomerular filtration rate by means of its clearance or by formulas based on its serum and urine concentration. Augmented renal clearance (ARC) is frequent among critically ill patients and is defined as creatinine clearance (CrCl) > 130 ml/min/1.73 m2. The aim of this study was to compare measured CrCl (MCC) and estimated CrCl obtained with the Cockcroft-Gault formula (CG), the Modification of Diet in Renal Disease Study equation (MDRD), and the Chronic Kidney Disease Epidemiology Collaboration formula (CKD-EPI) in patients with severe traumatic brain injury and nontraumatic subarachnoid hemorrhage. The second aim was to assess the incidence of ARC in this population of neurocritical patients. METHODS: This was a prospective, observational, single center study from a cohort of 74 patients admitted to the neurocritical intensive care unit due to traumatic brain injury or subarachnoid hemorrhage. Serum creatinine (at 7 a.m.) and a 6-h urine collection were analyzed, and CrCl was measured and estimated by using CG, MDRD, and CKD-EPI. The intraclass correlation coefficient (ICC) was evaluated for each pair, and Bland-Altman plots were used to assess clinical significance. RESULTS: Among 74 patients, the median age was 53 (interquartile range [IQR] 36-65), and the median Glasgow Coma Scale score at admission was 6. The median MCC at admission was 176 (IQR 135-214). The medians of CG, MDRD and CKD-EPI were, respectively, 129 ml/min/1.73 m2 (IQR 95-176), 158 (IQR 115-202), and 116 (97-132). An ICC was applied to evaluate the correlation between MCC and estimated methods and showed a weak correlation between MCC and estimated CrCl obtained with the three different methods. The strongest ICC statistical correlation was found between MCC and MDRD, and the weakest correlation was found between MCC and CKD-EPI. Bland-Altman plots showed that differences between each pair were not clinically acceptable. ARC was present in 78% of measurements, using MCC. A weak correlation was observed between MCC and calculated CrCl. CG, MDRD, and CKD-EPI overestimated MCC when MCC ≤ 130 ml/min/1.73 m2 and underestimated it when MCC > 130 ml/min/1.73 m2. CONCLUSIONS: In this population, there was a weak statistical correlation between measured and estimated methods. In patients with ARC, formulas underestimated MCC. MCC should probably be the preferred methodology for renal function assessment in the clinical setting to better adjust drug dosage and guarantee drug effectiveness.