Predicting future hospital antimicrobial resistance prevalence using machine learning

The study uses machine learning to predict antimicrobial resistance (AMR) in hospitals in England. It uses historical AMR and antimicrobial usage to predict future AMR. The results show that Extreme Gradient Boosting (XGBoost) models achieve the best predictive performance, with limited year-to-year variability in AMR prevalence. XGBoost outperforms other methods in Trusts with larger changes in AMR prevalence. The study concludes that XGBoost models can improve predictions, enabling informed decisions, efficient resource allocation, and targeted interventions at the hospital level.