Validation of RIFLE, AKIN, and a modified AKIN definition ("backward classification") of acute kidney injury in a general ICU: Analysis of a 1-year period.

The aim of this study was to validate Acute Kidney Injury Network (AKIN) and to develop a modified AKIN aimed at an improved classification of patients without baseline creatinine versus Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease (RIFLE) in general intensive care unit (ICU) patients.We retrospectively analyzed medical records of general ICU patients over a 1-year period. We compared the grading of severity as well as the prediction of mortality and renal replacement therapy by AKIN and RIFLE. Furthermore, a modified AKIN score was evaluated.A total of 321 patients were included. In 87% of cases, the 2 definitions classified patients in the concordant severity group. Higher scores of AKIN and RIFLE were associated with increased ICU- and 28-day mortality. Both definitions provided large receiver operating characteristics (ROC)-area under the curve (AUCs) for the prediction of mortality, which were comparable to the ROC-AUC of unclassified serum creatinine. Modification of the AKIN score with a "backward classification" of baseline creatinine based on its time course resulted in a higher AKIN score in 32 patients.RIFLE and AKIN definitions had a high concordance in staging the severity of acute kidney injury. There was a strong relationship between the stages and need for dialysis, ICU, and 28-day mortality. However, unclassified serum creatinine values were at least comparable. Standardized observation of the creatinine time course allows for "ex-post" AKIN-classification in a substantial number of patients with missing baseline creatinine values.