Using interpretable machine learning to predict bloodstream infection and antimicrobial resistance in patients admitted to ICU: Early alert predictors based on EHR data to guide antimicrobial stewardship

The study introduces a machine learning (ML) approach based on Multi-Objective Symbolic Regression (MOSR) to predict bloodstream infections and Antimicrobial Resistance (AMR) in intensive care units. The approach uses clinical data and outperforms traditional ML models, ensuring reliable results regardless of training set balance. The MOSR approach can be implemented on a large scale, offering a new tool for addressing healthcare issues influenced by limited data availability.