RESUMO
OBJECTIVES: This study aimed to investigate antibiotic prescribing patterns and effectiveness of different anti-carbapenem-resistant Acinetobacter baumannii (CRAB) strategies for CRAB pneumonia. METHODS: We conducted a multicentre, retrospective study in three hospitals. During 2010-2015, adult ICU patients with CRAB pneumonia treated with at least one antimicrobial agent covering the CRAB isolate in vitro for more than 2 days were included. We used multivariate logistic regression to analyse the associations of anti-CRAB strategies with ICU mortality and other clinical outcomes. RESULTS: Among 238 patients with CRAB pneumonia, tigecycline monotherapy (84, 35.3%) was the most common antibiotic strategy, followed by tigecycline with colistin (43, 18.1%), colistin monotherapy (34, 14.3%), colistin combination without tigecycline (33, 13.9%), tigecycline combination without colistin (32, 13.4%), and sulbactam-based therapy without tigecycline and colistin (12, 5.0%). In multivariate analysis, tigecycline-based therapy was associated with higher ICU mortality than non-tigecycline therapy (adjusted OR 2.30, 95% CI 1.19-4.46). There was no difference between colistin-based therapy and non-colistin therapy. Compared with tigecycline monotherapy, colistin monotherapy was associated with lower ICU mortality (aOR 0.30, 95% CI 0.10-0.88). Treatment failure analyses showed similar trends. Tigecycline-based therapy was associated with higher treatment failure rate than non-tigecycline therapy (aOR 2.51, 95% CI 1.39-4.54), whereas colistin-based therapy was associated with lower treatment failure rate than non-colistin-based therapy (aOR 0.48, 95% CI 0.27-0.86). CONCLUSIONS: Tigecycline was commonly prescribed for CRAB pneumonia. However, tigecycline-based therapy was associated with higher ICU mortality and treatment failure. Our study suggests that colistin monotherapy may be a better antibiotic strategy for CRAB pneumonia.
