Value Of Creatinine Clearance Rate Estimated Based On Serum Cystatin C In Acute Kidney Injury

Objective To compare diagnostic value of creatinine clearance rate estimation based on serum cystatinC (SCysC) with creatinine (Scr) in acute kidney injury, and whether they can predict the need for renal replacement therapy (RRT) .Methods The patients without AKI on admission and required the length of ICU stay of at least 72 h was collected from August 2010 to December 2010. As patients admitted, SCysC, Scr, Ccr estimated based on SCysC, urine output were monitored, meanwhile Ccr estimated based on Scr used by Cockcroft-Gault Equation. Diagnostic values of Ccr estimated based on SCysC and Scr were compared with area under curve(AUC) analysis of receptor operative curve(ROCC). The Scr on admission served as baseline renal function. According to the RIFLE classification system for AKI .Results 51 patients were enrolled for analysis. The patients were divided to two groups: AKI group (n=21), developed AKI after admission; and non-AKI group, never developed AKI after admission, served as controls. Compare Ccr estimated based on SCysC, Ccr estimated based on Scr on admission, the 2 days prior to AKI, the 1 days prior to AKI, the day to AKI of AKI group and on admission, the 2 days prior to the highest Scr, the 1 days prior to the highest Scr, the day to the highest Scr of non-AKI group. Ccr estimated based on SCysC and Scr in AKI group was significantly higher than non-AKI (P<0.01). On the 2 days prior to AKI Ccr estimated based on SCysC were significantly higher than on admission (P<0.01). Ccr estimated based on Scr were not significantly higher than non-AKI before the day to AKI. ROC analysis indicated that Ccr estimated based on SCysC would find AKI earlier than Ccr estimated based on Scr, the AUC of SCysC and Ccr estimated based on SCysC on the 2 days prior to AKI was 0.859 and 0.664. In AKI group 6 patients required renal replace therapy (RRT), the AKI patients requiring RRT had significantly higher APACHE II scores on admission (P = 0.01) and produced less urine(P<0.01). Ccr estimated based on SCysC did not predict the need of RRT on the day to AKI, the AUC were 0.65.Conclusions The sensitivity of Ccr estimated based on SCysC were higher than Ccr estimated based on Scr, so they could detect the change of renal function in an early stage. But the specificity was low. Therefore, in a critically ill population with high prevalence of AKI, SCysC and Ccr estimated based on SCysC may be helpful to rule out AKI. Ccr estimated based on SCysC can not predict the need for renal replacement therapy (RRT).