RESUMO
Clinical guidelines recommend using Kidney Disease Improving Global Outcomes (KDIGO) criteria for the diagnosis and classification of acute kidney injury (AKI) in patients with chronic liver disease (CLD). Concerns have been raised about the use of urine output (UO) criteria in CLD. We examined the significance of oliguria meeting the urine output criteria for AKI (AKI-UO) and examined its association with clinical outcomes in CLD patients. Using an 8-year clinical database from a large university medical center, 3458 patients with CLD were identified. AKI occurred in 2854 (82.5%) patients when they fulfilled any KDIGO criteria. When serum creatinine (SC) and UO criteria were used, 604 patients (17.5%) had no evidence of AKI and had the lowest hospital mortality rate (5%). Using AKI-UO criteria alone, 2103 patients (60.8%) were classified as stage 2-3 AKI. When only SC criteria were applied, 1281 (61%) of those patients with stage 2-3 AKI-UO were misclassified as either no AKI or AKI stage 1. Patients reclassified with AKI according to UO criteria (AKI-UO) had nearly a 3-fold increased rate of hospital mortality compared with patients without any AKI (14.6% versus 5%; P < 0.001) and more than a 50% increased mortality compared with stage 1 AKI-SC (14.6% versus 9%; P < 0.001). Patients with transient oliguria (AKI-UO stage 1) had increased mortality rates compared with patients without oliguria (14.9% versus 6.9%; P < 0.001). CONCLUSION: CLD patients have a high incidence of AKI. Compared with creatinine criteria alone, incorporating UO into the diagnostic criteria increased the measured incidence of AKI. Stage 2-3 AKI-UO has a high negative impact on hospital mortality. (Hepatology 2017;66:1592-1600).