AI-Driven Prediction and Management of Infection Risks in Intensive Care Units: A Systematic Review

Background and Objective: Healthcare-associated infections (HAIs) in intensive care units (ICUs) pose serious risks due to patient vulnerability and invasive procedures. Traditional infection control strategies often rely on retrospective monitoring. The objective of this systematic review is to evaluate the applications and effectiveness of AI-driven predictive and management tools for infection risk in ICUs. Methods: Following PRISMA guidelines, a systematic search of PubMed, Scopus, Web of Science, and IEEE Xplore was conducted for articles published between ۲۰۱۴ and ۲۰۲۵. After screening and applying inclusion criteria, ۳۶ high-quality studies were selected for analysis. Results: Of the selected studies, ۷۰% focused on AI-based risk prediction, ۵۵% on early intervention strategies, and ۴۰% on continuous infection management. Data sources included electronic health records, vital signs, laboratory results, and monitoring devices. Machine learning algorithms, particularly deep learning models, achieved accuracy between ۸۳% and ۹۲% in predicting infection risk, enabling timely preventive measures and reducing incidence of HAIs. AI integration improved patient safety, optimized antibiotic stewardship, and enhanced ICU workflow. Conclusion: AI-driven prediction and management systems offer promising solutions for proactive infection prevention and control in ICUs. Future research should focus on standardizing methodologies, improving data integration, and evaluating long-term effectiveness in diverse clinical settings.

کلیدواژه ها:

Artificial intelligence ، machine learning ، infection prevention ، intensive care units ، healthcare-associated infections