AI Tool Enhances Patient Safety by Analyzing Nurses’ Notes

Darius Baruo
Jun 28, 2025 02:18

An AI-powered tool, CONCERN EWS, developed by researchers, reduces patient risk and hospital stays by analyzing nurses’ shift notes for early health deterioration signs.

In a groundbreaking advancement in healthcare technology, researchers have developed an AI-powered tool that significantly enhances patient safety by analyzing nurses’ shift notes. The tool, known as the CONCERN Early Warning System (CONCERN EWS), has demonstrated its potential to identify early signs of patient health deterioration, thereby reducing associated risks and hospital stays, according to a report by NVIDIA.

Revolutionizing Patient Care

The CONCERN EWS, tested in clinical trials involving over 60,000 patients from 2020 to 2022, has shown promising results. The AI tool decreased patient mortality risk by more than 35% and reduced the average duration of hospital stays by over half a day. Furthermore, the system achieved a 7.5% reduction in sepsis risk among patients in hospitals where it was deployed.

AI Prioritizes Nurses’ Observations

Developed by a team of researchers from Columbia University and the University of Pennsylvania, the AI tool leverages machine learning to prioritize the critical yet often subtle observations made by nurses. Nurses frequently interact with patients and can detect nuanced changes in their health that may not be immediately apparent to others. By analyzing these observations, the AI tool provides early alerts to healthcare teams about potential health declines.

Technical Insights and Deployment

CONCERN EWS employs natural language processing to interpret nurses’ notes within electronic health records (EHRs) and examines metadata such as date, time, and location to detect patterns indicative of health issues. If it identifies unusual patterns, such as increased nurse visits at odd hours, it signals potential health risks to medical teams.

The AI system was tested in four hospitals across Massachusetts and New York, enabling medical staff to detect health concerns an average of 42 hours earlier than traditional methods, thus providing a critical window for intervention.

Future Developments and Recognitions

The success of CONCERN EWS has been acknowledged with a grant from the American Nurses Foundation’s “Reimagining Nursing Initiative.” The research team, co-led by Kenrick Cato of the University of Pennsylvania, plans to use the grant to develop a pediatric version of the tool in collaboration with Children’s Hospital Colorado.

This innovation underscores the transformative potential of AI in healthcare, particularly in enhancing patient safety and optimizing hospital operations. For further details, the original report can be accessed on the NVIDIA blog.

Image source: Shutterstock