RESUMEN
INTRODUCTION: Acinetobacter baumannii is a leading agent of healthcare-associated infection. The objective of this study was to evaluate cases of colonization or infection with polymyxin-resistant A. baumannii (PRAB) in liver transplant recipients and to identify the risk factors for the acquisition of PRAB. METHODS: We evaluated all patients undergoing liver transplantation (LT) between January and November of 2011. The exclusion criterion was death within the first 72 h after transplant. Patients were screened for PRAB through weekly rectal and inguinal swabs during their stay in the intensive care unit (ICU) and at ICU discharge. Patients who came from other hospitals or had been treated in the emergency room for >72 h were screened at ICU admission. The minimum inhibitory concentrations (MICs) for polymyxins were determined by broth microdilution, and clonality was determined by pulsed-field gel electrophoresis. The stepwise logistic regression was used to identify risk factors related to acquisition of PRAB, and Cox forward regression used to identify risk factors for 60-day mortality. RESULTS: We evaluated 65 patients submitted to LT, among whom PRAB was isolated in 7, 4 of whom developed infection. The MICs for polymyxin E ranged from 16 to 128 mg/mL. All patients with PRAB required dialysis. The median time of polymyxin use before PRAB isolation was 21 days. These 4 included 1 case of primary bloodstream infection (BSI), which was treated with the carbapenem-polymyxin combination; 1 case of surgical site infection, which was treated with gentamicin, polymyxin, ampicillin-sulbactam, and tigecycline; and 2 cases of pneumonia, treated with the combination of carbapenem-polymyxin. In the case of BSI and in 1 of the cases of pneumonia, the treatment was considered successful. Mortality was 71% among the cases, compared with 33% among the non-cases. CONCLUSION: In the final model of the survival analysis, PRAB colonization or infection after LT was independently associated with mortality. One predominant clone was identified. The only risk factor identified in the multivariate analysis was polymyxin use. PRAB was an agent with high mortality, and the most important risk factor associated with colonization or infection for such bacterium was polymyxin use.