Abstract

BACKGROUND: Antibiotic use drives antibiotic resistance. OBJECTIVES: To systematically review the literature and estimate associations between prior exposure to antibiotics across WHO AWaRe categories (Access, Watch, Reserve) and isolation of critical and high-priority multi-drug resistant organisms (MDROs) on the WHO priority pathogen list. METHODS: Data sources: Embase, Ovid Medline, Scopus, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials, and ClinicalTrials.gov (from inception to 20/08/2020). STUDY ELIGIBILITY CRITERIA: Case-control, cohort or experimental studies that assessed the risk of infection/colonization with MDROs. PARTICIPANTS: Inpatients or outpatients of any age and sex. INTERVENTIONS: Prior exposure to antibiotics that could be categorized into the AWaRe framework.Assessment of risk of bias: Tailored design-specific checklists applied to each included study. DATA SYNTHESIS: For each antibiotic/class, crude odds ratios (ORs) were pooled through random-effects meta-analyses, both overall and by MDRO. Heterogeneity was examined. RESULTS: We identified 349 eligible studies. All were observational, prone to bias due to design and lack of adjustment for confounding, and not primarily designed to compare associations across AWaRe categories. We found statistically significant associations between prior exposure to almost all antibiotics/classes across AWaRe categories and colonization/infection with any MDRO. We observed higher ORs for Watch and Reserve antibiotics than with Access antibiotics. First generation cephalosporins (Access) had the least association with any MDRO colonization/infection (58 studies; OR=1.2 [95% CI: 1.0-1.4]), whereas strongest associations were estimated for linezolid (Reserve) (22 studies; OR=2.6 [95% CI: 2.1-3.1]), followed by carbapenems (Watch) (237 studies; OR=2.3 [95% CI: 2.1-2.5]). There was high heterogeneity for all antibiotic/MDRO associations. CONCLUSION: Optimising use of Access antibiotics is likely to reduce the selection of MDROs and global antibiotic resistance. Despite data limitations, our study offers a strong rationale for further adoption of AWaRe as an important tool to improve antibiotic use globally.