Objective: To assess renal dysfunction and outcome in patients treated exclusively with colistin vs. other antibiotics.
Design and setting: Prospective cohort study in a mixed ICU in a university-affiliated hospital.
Patients: 185 patients infected with Acinetobacter baumannii and Pseudomonas aeruginosa after an ICU stay longer than 48 h: 55 in the colistin group and 130 in the noncolistin group, similar in age, APACHE II, medical status, and SOFA score.
Measurements and results: We recorded data on epidemiology and severity of illness, site of infection, renal function before and after treatment, clinical cure, and mortality. Clinical cure was defined as simultaneous normalization of central temperature (< or = 38 degrees), leukocyte count (< or = 10,000/mm3), and PaO2/FIO2 ratio (>187). Before treatment creatinine was 0.9+/-0.2 in the colistin group and 0.9+/-0.1 in the noncolistin group; after treatment the value was 1.0+/-0.3 in both groups. The most frequent infection was ventilator-associated pneumonia: 53% vs. 66% in colistin and noncolistin groups, respectively, Acinetobacter was the cause in 65% and 60% and Pseudomonas in 35% and 53%. In the noncolistin group 81% of patients were treated with carbapenems. Inadequate empirical antimicrobial treatment was more frequent in the colistin group (100% vs. 8%), but there were no differences in the frequency of clinical cure on day 6 of treatment (15% and 17%) or in mortality (29% and 24%).
Conclusions: Colistin appears to be as safe and as effective as other antimicrobials for treatment of sepsis caused by Acinetobacter and Pseudomonas in critically ill patients.