The recently proposed RIFLE classification of acute renal failure (ARF) (Acute Dialysis Quality Initiative, http://www.ADQI.net) has yet to be evaluated in a general ICU population. RIFLE defines three grades of severity of ARF on the basis of either urine output (U) or an acute increase in serum creatinine (C): risk = C × 1.5 or U < 0.5 ml/kg/hour × 6 hours; injury = C × 2 or U < 0.5 ml/kg/hour × 12 hours; and failure = C × 3, C ≥ 4 mg/dl, U < 0.3 ml/kg/hour × 24 hours, or anuria × 12 hours.

We prospectively collected data for all patients admitted to an ICU from 1 July 2000 to 30 June 2001 at the University of Pittsburgh Medical Center, a tertiary hospital with > 120 ICU beds serving medical, surgical, trauma, neurologic, and transplant patients. We tested two hypotheses: (1) increasing RIFLE class corresponds to decreasing ICU occurrence and increasing hospital mortality; and (2) patients classified by C or U criteria have similar occurrence and mortality. We classified patients according to their worst class on C or U criteria. For baseline C, we selected the lower of the admission C and the predicted C (based on age, sex, and race, using the Modification of Diet in Renal Disease formula).

RIFLE criteria appear clinically sensible with increasing mortality. Unfortunately, C and U criteria alone yield different results, and occurrence decreases only when C criteria are used alone.