Background: Antibiotic cycling or rotation of antimicrobial agent classes has been proposed to combat antimicrobial resistance.
Methods: A prospective cohort study was conducted in a medical intensive care unit (ICU) of a university hospital between December 1, 2000, and September 30, 2002, as part of a three-center trial under the aegis of the U.S. Centers for Disease Control and Prevention. Patients admitted to the medical ICU for > 48 h were enrolled, and demographic and microbiological data were collected until discharge or death. Baseline data were collected for four months (12/1/00 to 3/31/01) and compared with data collected after institution of a quarterly cycling regimen (cycle order: Cefepime, ciprofloxacin, piperacillin-tazobactam, imipenem-cilastatin) for the empiric treatment of gram-negative infections (4/01/01 to 9/30/02).
Results: Of 1,074 consecutive admissions, 301 were enrolled, 59 during baseline and 242 during the cycling periods. An outbreak of multi-drug resistant Pseudomonas aeruginosa followed cycle 2 (cefepime), coinciding with cycles 3 and 4 (ciprofloxacin and piperacillin-tazobactam) (80.0 and 73.7 vs. 37.3 isolates/100 patients enrolled for cycles 3/4 and baseline, respectively; p = 0.04). Acinetobacter spp. were isolated less frequently during the cycling periods (15.3 vs. 1.2 isolates/100 patients for baseline and cycling periods, respectively; p > or = 0.01). The crude hospital mortality rate was similar (24/59 [41%] baseline vs. 73/242 [30%] cycling; p = 0.16) between periods. However, the percentage of patients admitted to the medical ICU who subsequently acquired an infection followed by in-hospital death was higher at baseline than during cycling: 15/59 (25.4%) vs. 33/242 (13.6%)(p = 0.04).
Conclusions: In this study, the cycling strategy was not definitively associated with beneficial changes in unit epidemiology and in fact may have contributed to an outbreak of multi-drug resistant P. aeruginosa.