Clinical staff outperform ChatGPT AI at emergency department triage, but AI shows promise as a support tool for urgent cases.
Key Details
- 1Study compared six doctors, 44 nurses, and ChatGPT 3.5 on triaging 110 clinical cases using the Manchester Triage System.
- 2Doctors achieved 70.6% accuracy, nurses 65.5%, and AI 50.4%; AI had lower sensitivity for urgent cases (58.3% vs. nurses 73.8%, doctors 83%).
- 3AI outperformed nurses in the most urgent triage category for accuracy (27.3% vs. 9.3%) and specificity (27.8% vs. 8.3%).
- 4AI tended to over-triage, assigning higher urgency more often than staff.
- 5Authors advocate for AI as an adjunct to, not replacement for, clinical judgement.
- 6Study limitations include small sample size, single center, and non-real-world AI setting.
Why It Matters
These findings emphasize the current limitations of general AI models in patient triage compared to clinical professionals, highlighting the need for cautious integration, oversight, and further development before AI can play a leading role in acute medical decision-making.
Source
EurekAlert
Related News
•EurekAlert
UVA's $4M NIH Grant Backs AI to Optimize Cardiac MRI Therapies
UVA secures $4 million NIH funding to develop GenStrain, an AI model using MRI data to personalize cardiac pacing therapies.
•EurekAlert
Machine Learning Models Predict Cancer Risk with Routine Health Data
Researchers developed machine learning models using routinely collected health data to predict individual risk for 20 common cancers, outperforming traditional methods for several cancer types.
•EurekAlert
AI Deep Learning Model Enhances Echocardiogram RV Assessment and Prognosis
Researchers developed DROID-RV, a deep learning tool that automates right ventricular assessment from echocardiograms, showing strong prognostic links with cardiovascular outcomes.