Efficacy and safety of antibiotics in the treatment of multidrug-resistant Acinetobacter baumannii infections: a systematic review and Bayesian network meta-analysis

OBJECTIVES: This study aims to compare the efficacy and safety of antimicrobial regimens for multidrug-resistant Acinetobacter baumannii (MDR A. baumannii) infections using a Bayesian network meta-analysis (NMA). METHODS: January 2025, for randomized controlled trials (RCTs) and cohort studies. The primary outcome was all-cause mortality; secondary outcomes included clinical improvement, microbiological cure, and nephrotoxicity. A random-effects NMA estimated odds ratios (ORs) with 95% credible intervals (CrIs) and ranked regimens using the surface under the cumulative ranking curve (SUCRA). RESULTS: The Bayesian NMA included 10 RCTs and 38 cohort studies (14 regimens; 7,300 patients), with cohorts contributing 80.6% of the data. Compared with tigecycline, cefiderocol-based therapies (combination: OR = 0.41, 95% CrI: 0.22-0.77; monotherapy: OR = 0.44, 95% CrI: 0.27-0.76) and sulbactam-durlobactam (OR = 0.42, 95% CrI: 0.15-1.20) were associated with reduced mortality (very low certainty). Sulbactam-durlobactam also ranked highest for clinical improvement (OR = 5.79, 95% CrI: 1.62-21.46, SUCRA 91.0%) and microbiological cure (OR = 8.66, 95% CrI: 1.92-41.01, SUCRA 98.2%; very low certainty). Colistin-based regimens had the highest nephrotoxicity (combination: OR = 8.21, 95% CrI: 1.41-59.35; monotherapy: OR = 7.75, 95% CrI = 1.81-40.82; very low certainty). CONCLUSIONS: Sulbactam-durlobactam ranked highest for clinical improvement and microbiological cure, while cefiderocol-based therapies scored highest for mortality reduction in our NMA. However, the evidence was largely derived from observational studies with very low certainty. Further A. baumannii-specific RCTs are needed to confirm these findings.