| Using AI and Healthcare System Data to Forecast Antimicrobial Resistance Patterns: A Review of Current Approaches |
| Paper ID : 1287-IGA |
| Authors |
|
Sara Najafi * Department of Biology, Faculty of Basic Sciences, Hamedan Branch, Islamic Azad University, Hamedan, Iran |
| Abstract |
| Background and Aim: Antimicrobial resistance (AMR) presents a growing threat to global health, leading to increased morbidity, mortality, and healthcare costs. Traditional surveillance methods are often reactive and limited by delays in data collection and analysis. The integration of artificial intelligence (AI) with real-world healthcare system data offers a novel pathway to proactively predict AMR trends. However, current literature lacks a focused synthesis of how AI models are being applied to clinical and hospital data for AMR forecasting. This review aims to evaluate existing AI approaches using healthcare system data to predict AMR patterns and guide clinical decision-making. Methods: A structured literature search was conducted using PubMed, Scopus, and Web of Science. Keywords included “artificial intelligence,” “machine learning,” “antimicrobial resistance,” “electronic health records,” “hospital data,” and “AMR prediction.” Studies published in English from 2020 to 2025 were included. Selected articles employed AI models trained on hospital or clinical data to predict resistance trends or guide antimicrobial prescribing. Results: The review identified diverse AI techniques—including random forest, support vector machines, and deep learning—used to analyze electronic health records, microbiology reports, and prescription patterns. These models demonstrated moderate to high accuracy in predicting resistance at both pathogen and population levels. Key challenges included data heterogeneity, model interpretability, and integration into clinical workflows. Conclusion: AI, when combined with healthcare system data, holds great promise for predicting AMR patterns and supporting antimicrobial stewardship. Further research is needed to enhance model generalizability, transparency, and real-time application in clinical settings. |
| Keywords |
| Antimicrobial Resistance (AMR), Artificial Intelligence (AI), Healthcare Data, Machine Learning, Predictive Modeling |
| Status: Accepted |
