Correlates of acute renal failure in patients receiving parenteral amphotericin B

Background, While parenteral amphotericin B is an effective therapy for ser ious fungal infections. it frequently causes acute renal failure (ARF). Thi s study identified correlates of ARF in amphotericin B therapy and used the m to develop clinical prediction rules. Methods. All 643 inpatients receiving parenteral amphotericin B therapy at one tertiary care hospital were included. Data regarding correlates were ob tained both electronically and from manual chart review in a subsample of 2 31 patients. ARF was defined as a 50% increase in the baseline creatinine w ith a peak greater than or equal to2.0 mg/dL. Results. Among 643 episodes. ARF developed in 175 (27%). In the larger grou p, the only independent correlate of ARF was male gender (OR = 2.2. 95% Cl. 15 to 3.3). In the subsample (N = 231), independent correlates of ARF were maximum daily amphotericin dosage, location at the time of initiation of a mphotericin therapy, and concomitant use of cyclosporine. These data were u sed to develop two clinical prediction rules, A rule using only data availa ble at initiation of therapy stratified patients into groups with probabili ty of ARF ranging from 15 to 54%. while a rule including data available dur ing therapy (maximum daily dose) stratified patients into groups with proba bility of ARF ranging front 4 to 80%. Conclusions. Acute renal failure occurred in a quarter of the patients. Cor relates of ARF at the beginning and during the course of amphotericin thera py were identified and then combined to allow stratification according to A RF risk. These data also provide evidence for guidelines for the selection of patients for alternative therapies.