RESUMO
OBJECTIVES: Acute kidney injury (AKI) after cardiac surgery is associated with adverse patient outcome. A new definition and staging system for AKI based on creatinine kinetics (CKs) has been proposed recently. Their proponents hypothesize that early absolute increases in serum creatinine (sCr) after kidney injury are superior to percentage increases, especially in patients with chronic kidney disease (CKD). The aims of our study were to measure agreement between CK definition and the current consensus definition [risk, injury, failure, loss and end-stage renal disease (RIFLE) system], and to compare time to diagnosis and prognostic value between both systems. METHODS: Retrospective cohort study. Agreement on AKI diagnosis by both classifications, time to diagnosis and prognostic value of both systems were compared in cardiac surgeries performed during a 6-year period (2002-2007) in a single centre. RESULTS: We found substantial agreement between both classifications (0.67). More patients were diagnosed with AKI by the CK definition than by RIFLE criteria both globally (28.2 vs 13.9%) and in every category (16.5 vs 8.4% for CK-1 vs RIFLE-R; 8.4 vs 3.6% for CK-2 vs RIFLE-I and 3.2 vs 2.0% for CK-3 vs RIFLE-F). Time to diagnosis was shorter for the CK definition (1.8 vs 2.5 days). Prognostic value in terms of information about in-hospital death and need for renal replacement was comparable between classifications. CONCLUSIONS: In cardiac surgery, the CK definition and classification system showed substantial agreement with the current standard, was more sensitive than RIFLE and detected AKI earlier without loss of prognostic information.
