Institutional spillover effects of antimicrobial use on antimicrobial resistance and Clostridioides difficile: a systematic review.

OBJECTIVE

To synthesize evidence on institutional spillover effects of antimicrobial use (AMU) on antimicrobial resistance (AMR) and Clostridioides difficile infections on individuals without direct antimicrobial exposure.

DESIGN

Systematic review.

METHODS

Three databases were searched through August 2024 for studies evaluating spillover effects of AMU on unexposed individuals in institutional settings. Study characteristics, AMU, and outcomes were extracted. Study quality was assessed based on underlying methodology to detect spillover effect. A hybrid synthesis, including effect direction and meta-analysis of studies reporting continuous AMU and non-aggregate outcomes was utilized. Reporting followed PRISMA guidelines.

RESULTS

Of 5916 screened studies, five observational studies met inclusion criteria. Three were conducted across 68 hospital wards (ward-level exposure), and two across 693 nursing homes (facility-level exposure). Three studies evaluated all antimicrobial classes; two focused on penicillins, fluoroquinolones, and carbapenems. Two studies examined C. difficile, one MRSA, one carbapenem-resistant Enterobacterales, and one reported combined AMR and C. difficile outcomes. Low to moderate quality evidence indicated a positive spillover effect direction with increasing facility AMU. Meta-analysis of three studies yielded a pooled IRR of 1.54 (95% CI 0.85-2.80) per 100 days of therapy per 1,000 patient-days, with significant heterogeneity (I 2 = 97.6%).

CONCLUSIONS

This review identified five studies suggesting a positive association between institutional AMU and collateral risks of AMR and C. difficile among unexposed individuals. Findings were limited by methodological heterogeneity and potential publication bias. Standardizing terminology, specifying spillover mechanisms, and adopting robust observational designs can enhance future research on spillover effects.