Acute Kidney Injury in Non-Critical Care Setting: Elaboration and Validation of an In-hospital Death Prognosis Score

Acute renal failure (AKI) is defined by a deterioration of kidney function over a short period. This definition was clarified recently in order to allow homogenization and optimal comparison of patients in clinical studies by the classifications RIFLE in 2004, AKIN in 2007 and KDIGO in 2013. These classifications decline several stages of ARI through the increase in the plasma creatinine level and the decrease in urine flow.

Even though AKI is a frequent pathology in all hospitalized patients, there are only few studies that are interested in this entity in conventional hospital services except intensive care or intensive care. Indeed, the recent meta-analysis including the 154 studies focusing on the ARI defined by the KDIGO criteria, only 7 have recruited patients in conventional nephrology services. However, patients admitted for an ARI which requires treatment in a medical service probably have epidemiological characteristics and a different prognosis than those requiring treatment in intensive care. In addition, the parameters of the RIFLE, AKIN or KDIGO scores are more difficult to establish in conventional hospital services than in intensive care, especially for hourly monitoring of urine flow, not allowing an optimal classification of the episode of IRA.

This study set out to develop a prognostic score for intra-hospital mortality in ARI based on a first historical cohort. The investigators then validated this score on a second prospective cohort obtained over an independent inclusion period and at a distance from the first.