ChatGPT, Enhanced with Clinical Practice Guidelines, is a Superior Decision Support Tool
Wang, Y.; Visweswaran, S.; Kappor, S.; Kooragayalu, S.; Wu, X.
Show abstract
ChatGPT has gained remarkable traction since its inception in November 2022. However, it faces limitations in generating inaccurate responses, ignoring existing guidelines, and lacking reasoning when applied in clinical settings. This study introduces ChatGPT-CARE, a tool that integrates clinical practice guidelines with ChatGPT, focusing on COVID-19 outpatient treatment decisions. By employing in-context learning and chain-of-thought prompting techniques, ChatGPT-CARE enhances original ChatGPTs clinical decision support and reasoning capabilities. We created three categories of various descriptions of patients seeking COVID-19 treatment to evaluate the proposed tool, and asked two physicians specialized in pulmonary disease and critical care to assess the responses for accuracy, hallucination, and clarity. The results indicate that ChatGPT-CARE offers increased accuracy and clarity, with moderate hallucination, compared to the original ChatGPT. The proposal ChatGPT-CARE could be a viable AI-driven clinical decision support tool superior to ChatGPT, with potential applications beyond COVID-19 treatment decision support.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- Evaluating large language model workflows in clinical decision support: referral, triage, and diagnosis 96%
- A Framework to Assess Clinical Safety and Hallucination Rates of LLMs for Medical Text Summarisation 96%
- Development and Prospective Implementation of a Large Language Model based System for Early Sepsis Prediction 95%
Similar papers in this journal
- Evaluating Semantic Similarity Methods for Comparison of Text-derived Phenotype Profiles 93%
- MelAnalyze: Fact-Checking Melatonin claims using Large Language Models and Natural Language Inference 93%
- ARDSFlag: An NLP/Machine Learning Algorithm to Visualize and Detect High-Probability ARDS Admissions Independent of Provider Recognition and Billing Codes 93%
Similar papers in this journal
Similar papers in this journal
- Comparative Effectiveness of Medical Concept Embedding for Feature Engineering in Phenotyping 95%
- A Study of Calibration as a Measurement of Trustworthiness of Large Language Models in Biomedical Research 95%
- Automating Evaluation of LLM-generated Responses to Patient Questions about Rare Diseases 94%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.