Problem: Hospital Acquired Infection (HAI) surveillance serves as the foundation for benchmarking acute care hospital performance against peer organizations and provides a focus for HAI reduction efforts, but consumes a considerable amount of time for the Infection Prevention (IP) team. Specifically, time spent on interpretation and application of the National Healthcare Safety Network (NHSN) Patient Safety Manual definitions limits the IP team’s ability to engage in proactive prevention efforts upstream of HAI occurrences.
Solution: We propose the development and implementation of an artificial intelligence-driven system to automate the surveillance of HAIs including CAUTI and CLABSI. This system will leverage tools such as large language models to analyze structured and unstructured data within the electronic health record (EHR) to adjudicate potential infections based on the NHSN criteria – all in real-time.
Impact: This project represents a novel application of AI in infection prevention and control, shifting the focus from retrospective analysis to real-time surveillance and prevention efforts. By automating the surveillance process, we aim to enhance the frequency, efficiency and effectiveness of proactive IP efforts, ultimately improving patient safety and clinical outcomes. Fewer HAI will result in decreased morbidity, mortality, length of stay, and cost.
