928

Abstract

development of acute kidney injury (AKI), as determined by RIFLE criteria, on day 7 of therapy. Patients who required renal replacement therapy (RRT) prior to colistin initiation or had a history of end-stage renal disease were excluded. Classification and regression tree analysis (CART) was used to determine the breakpoint for colistin dose with the strongest association with development of AKI. Results: Of 212 patients screened for inclusion, 65 met all inclusion and exclusion criteria. CART analysis identified a dose of > 3.6 mg/kg of IBW/ day (considered high-dose) to be significantly associated with the development of AKI, where 67% of patients who received high-dose colistin developed AKI, compared to 15% who received ≤ 3.6 mg/kg/day (p<0.001). There was also a significant correlation between colistin dose and the stepwise progression from no AKI to risk, injury, and failure categories, where the average daily colistin doses were 3.5, 4.3, 4.6, and 5.4 mg/kg of IBW/day, respectively (p<0.001). After adjusting for concomitant nephrotoxins, baseline SOFA score, and co-morbidities, receipt of high-dose colistin was the only factor associated with development of day 7 AKI (p<0.001). Despite higher rates of day 7 AKI, receipt of high-dose colistin was not associated with the persistent need for RRT at day 30 (18% vs. 12%, p=0.73). Conclusions: These data suggest that a colistin dose > 3.6 mg/kg of IBW/day may be associated with the development of early AKI. However, few patients progress to needing prolonged RRT.