Using Generative Artificial Intelligence for Healthcare-Associated Infection Surveillance

Comprehensive Summary

This study explores how generative AI can improve healthcare-associated infection (HAI) surveillance. HAIs, such as CLABSI, CAUTI, and SSI, are common but often preventable with effective HAI tracking and surveillance. However, current surveillance methods are resource intensive, making an alternative automated surveillance with GenAI beneficial. GenAI can analyze unstructured clinical notes, extract relevant infection criteria, and make diagnoses. The study proposes three levels of implementation, assistive where AI extracts information for human review, semi automated where AI diagnoses with human confirmation, and fully automated where AI makes final decisions. These approaches could improve efficiency and consistency in surveillance, however some challenges include bias in training the data and accuracy.

Outcomes and Implications

This study examines the implementation of GenAI in automating healthcare-associated infection surveillance, improving HAI prevention.