ABSTRACT
AIM: Fluid balance (FB) is an emerging predictor of acute kidney injury (AKI). We investigated the comparative utility of FB with conventional and novel biomarkers to predict AKI in cardiovascular surgery patients. METHODS: Data collected in a prospective, observational study designed to investigate the relationship between FB and AKI in an academic medical center were utilized for analyses. FB, routine clinical parameters, conventional and novel biomarkers in 100 consecutive cardiovascular surgery patients was analyzed. RESULTS: Each variable studied was divided into quartiles and the lowest quartile served as the referent quartile. The adjusted OR for AKI for the highest vs. lowest quartile of FB was 4.98 (CI95%1.38-24.10, P=0.046), serum creatinine (SCr) 11.54 (CI95% 1.37-97.18, P=0.024), urine NGAL 2.76 (CI95% 0.48-15.93, P=0.255) and IL-18 2.31 (CI95% 0.41-13.16, P=0.346, and serum MCP-1 4.93 (CI95% 0.81-30.09, P=0.084) and TNF-alpha 15.59 (CI95% 1.19-204.19, P=0.036). Comparison of ROC curves demonstrated that the diagnostic performance of FB and SCr to predict AKI were comparable, as were FB with urine NGAL and IL-18 and serum MCP-1 and TNF-alpha.. While there was a graded relationship with the risk for AKI according to quartiles for FB, SCr and serum TNF-alpha, the remaining biomarkers including urine NGAL were not independent predictors of AKI. CONCLUSION: At 24 hours postoperatively, the performance of FB to predict AKI was comparable to that of preoperative conventional and postoperative 24-hour novel biomarkers.
