Outcome definitions in non-dialysis intervention and prevention trials in acute kidney injury (AKI)

BACKGROUND: The risk, injury, failure, loss-of-function, end-stage-renal-failure (RIFLE) and acute kidney injury network (AKIN) consensus definitions of acute kidney injury (AKI) were established in part to facilitate comparison of trials. Contrast-induced nephropathy (CIN) has traditionally used a less demanding definition. OBJECTIVES: To review use of RIFLE and AKIN as AKI trial outcome variables and contrast these with outcomes for CIN. METHODS: We conducted a search of PubMed from 1 January 2005 to 31 December 2008 and 9 trial registries for randomized control trials for preventional or interventional treatment of AKI and CIN. RESULTS: RIFLE or AKIN were outcome variables in 36% (n = 8) of the published (n = 22) and 18% (n = 4) of the current (n = 22) AKI trials. RIFLE was used to triage to intervention in three trials. The urine output definition of RIFLE and AKIN was an outcome in only two trials. In 18% (n = 8) of AKI trials, the CIN definition (increase in serum creatinine of > or =25% and/or > or =44 micromol/l) was the primary outcome. This was also the primary outcome in 56% (n = 13) of published (n = 12) and current (n = 11) CIN trials. Three published CIN trials used RIFLE or AKIN as an outcome (13%). The duration over which outcomes were determined varied from 24 h to 7 days. CONCLUSIONS: Considerable heterogeneity remains in outcome variables of AKI and CIN clinical trials. Even when the RIFLE or AKIN criteria were used, they were not applied consistently. There is a need for further consensus on surrogate outcome variables.