Title: Large Language Model Influence on Diagnostic Reasoning A Randomized Clinical Trial
Authors: Ethan Goh, Robert Gallo, Jason Hom, Eric Strong, Yingjie Weng, Hannah Kerman, Joséphine A. Cool, Zahir Kanjee, Andrew S. Parsons, Neera Ahuja, Eric Horvitz, Daniel Yang, Arnold Milstein, Andrew P. J. Olson, Adam Rodman, Jonathan H. Chen
Published: 2024-10-28
Link: https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2825395
Abstract
Key Points
Question  Does the use of a large language model (LLM) improve diagnostic reasoning performance among physicians in family medicine, internal medicine, or emergency medicine compared with conventional resources?
Findings  In a randomized clinical trial including 50 physicians, the use of an LLM did not significantly enhance diagnostic reasoning performance compared with the availability of only conventional resources.
Meaning  In this study, the use of an LLM did not necessarily enhance diagnostic reasoning of physicians beyond conventional resources; further development is needed to effectively integrate LLMs into clinical practice.
Abstract
Importance  Large language models (LLMs) have shown promise in their performance on both multiple-choice and open-ended medical reasoning examinations, but it remains unknown whether the use of such tools improves physician diagnostic reasoning.
Objective  To assess the effect of an LLM on physiciansâ diagnostic reasoning compared with conventional resources.
Design, Setting, and Participants  A single-blind randomized clinical trial was conducted from November 29 to December 29, 2023. Using remote video conferencing and in-person participation across multiple academic medical institutions, physicians with training in family medicine, internal medicine, or emergency medicine were recruited.
Intervention  Participants were randomized to either access the LLM in addition to conventional diagnostic resources or conventional resources only, stratified by career stage. Participants were allocated 60 minutes to review up to 6 clinical vignettes.
Main Outcomes and Measures  The primary outcome was performance on a standardized rubric of diagnostic performance based on differential diagnosis accuracy, appropriateness of supporting and opposing factors, and next diagnostic evaluation steps, validated and graded via blinded expert consensus. Secondary outcomes included time spent per case (in seconds) and final diagnosis accuracy. All analyses followed the intention-to-treat principle. A secondary exploratory analysis evaluated the standalone performance of the LLM by comparing the primary outcomes between the LLM alone group and the conventional resource group.
Results  Fifty physicians (26 attendings, 24 residents; median years in practice, 3 [IQR, 2-8]) participated virtually as well as at 1 in-person site. The median diagnostic reasoning score per case was 76% (IQR, 66%-87%) for the LLM group and 74% (IQR, 63%-84%) for the conventional resources-only group, with an adjusted difference of 2 percentage points (95% CI, â4 to 8 percentage points; Pâ=â.60). The median time spent per case for the LLM group was 519 (IQR, 371-668) seconds, compared with 565 (IQR, 456-788) seconds for the conventional resources group, with a time difference of â82 (95% CI, â195 to 31; Pâ=â.20) seconds. The LLM alone scored 16 percentage points (95% CI, 2-30 percentage points; Pâ=â.03) higher than the conventional resources group.
Conclusions and Relevance  In this trial, the availability of an LLM to physicians as a diagnostic aid did not significantly improve clinical reasoning compared with conventional resources. The LLM alone demonstrated higher performance than both physician groups, indicating the need for technology and workforce development to realize the potential of physician-artificial intelligence collaboration in clinical practice.
Trial Registration  ClinicalTrials.gov Identifier: NCT06157944
Found via this article in the NY Times: ChatGPT Defeated Doctors at Diagnosing Illness - The New York Times