The evaluation of acute renal failure (ARF) by the newly developed classification for ARF (RIFLE, standing for 'risk, injury, failure, loss, end-stage kidney disease') [1] in patients with sepsis has not yet been performed. We evaluated, retrospectively, the incidence of ARF and its risk factors, therapy, and outcome among patients with sepsis admitted to the Infectious Diseases Intensive Care Unit of the Hospital de Santa Maria between January 2005 and December 2006. ARF was defined by means of the RIFLE classification [1]. Sepsis was classified in accordance with the American College of Chest Physicians and the Society of Critical Care Medicine consensus [2]. In all, 182 patients (aged 56.2 ± 18.56 years (mean ± SD), 120 male, 162 Caucasian) were analyzed. Baseline characteristics of the patients are summarized in Table 1. Sixty-eight patients (37.4%) had ARF. By multivariate analysis, age more than 60 years (odds ratio (OR) 0.39, 95% confidence interval (CI) 0.17 to 0.87, P = 0.002), male (OR 5.5, 95% CI 2.2 to 13.5, P < 0.0001), chronic kidney disease (OR 0.2, 95% CI 0.06 to 0.79, P = 0.021), Gram-negative-related infection (OR 0.38, 95% CI 0.16 to 0.89, P = 0.027), and a Simplified Acute Physiology Score, version II (SAPS II) > 50 (OR 0.14, 95% CI 0.06 to 0.31, P < 0.0001) were independently associated with ARF. Thirteen patients (3 with injury and 10 with renal failure) had received renal replacement therapy (12 receiving continuous venovenous hemodiafiltration, and 1 receiving intermittent hemodialysis). The mortality rate was 37.4%, and increased from 'normal' to 'failure'. Patients who did not die had renal function recovery. Multivariate analysis including age more than 60 years, gender, SAPS II > 50, comorbidity (namely cardiovascular disease), and ARF showed that SAPS II > 50 (OR 0.12, 95% CI 0.05 to 0.29, P < 0.0001) and ARF (OR 0.26, 95% CI 0.11 to 0.63, P = 0.003) were independent predictors of mortality.

Table 1 Baseline characteristics
Full size table

Thus, ARF as determined by RIFLE is common among patients with sepsis, and increases mortality. Age, gender, chronic kidney disease, Gram-negative-related infection and severity of illness are independently associated with ARF in this setting.