Machine Learning Models for Predicting Patient Risk of Hospital Acquired Infections

Abstract

Hospital-acquired infections (HAIs) pose a significant challenge to patient safety and healthcare systems worldwide. To address this, we propose the development of machine learning models to assess patient risk of HAIs, enabling targeted interventions for prevention and control. Our study focuses on leveraging electronic health record (EHR) data to train and validate these models, incorporating a range of clinical and demographic variables. We evaluate the performance of various machine learning algorithms, including logistic regression, random forests, and gradient boosting, in predicting HAIs across different patient populations. Our results demonstrate promising predictive capabilities, with the potential to enhance infection control measures and improve patient outcomes.