CoChat

ChatGPT Influence on Medical Decision-Making, Bias, and Equity: A Randomized Study of Clinicians Evaluating Clinical Vignettes

Open access

The DOI matches its Crossref record; 12 of the 12 references it cites were found in a registry or an index.

Abstract

In a randomized, pre-post intervention study, we evaluated the influence of a large language model (LLM) generative AI system on accuracy of physician decision-making and bias in healthcare. 50 US-licensed physicians reviewed a video clinical vignette, featuring actors representing different demographics (a White male or a Black female) with chest pain. Participants were asked to answer clinical questions around triage, risk, and treatment based on these vignettes, then asked to reconsider after receiving advice generated by ChatGPT+ (GPT4). The primary outcome was the accuracy of clinical decisions based on pre-established evidence-based guidelines. Results showed that physicians are willing to change their initial clinical impressions given AI assistance, and that this led to a significant improvement in clinical decision-making accuracy in a chest pain evaluation scenario without introducing or exacerbating existing race or gender biases. A survey of physician participants indicates that the majority expect LLM tools to play a significant role in clinical decision making.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1101/2023.11.24.23298844

Cites

12 of 12 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 - Study Design: 50 US-licensed physicians were recruited for a remote video session where they were presented with a video of a standard patient actor depicting a case of chest pain in an outpatient setting. Participants were randomized to encounter an actor who was a White male or a Black female. The clinicians then responded to a series of four questions based on the vignette. For the first two questions, after providing their initial answers, they were presented with a pre-prepared LLM response based on the same vignette and questions. Clinicians were then offered an opportunity to modify their initial answers. For the final two questions, after their initial response, clinicians were allowed to directly interact with the LLM to ask any questions before considering whether or not to modify their answer.
Figure 1 - Study Design: 50 US-licensed physicians were recruited for a remote video session where they were presented with a video of a standard patient actor depicting a case of chest pain in an outpatient setting. Participants were randomized to encounter an actor who was a White male or a Black female. The clinicians then responded to a series of four questions based on the vignette. For the first two questions, after providing their initial answers, they were presented with a pre-prepared LLM response based on the same vignette and questions. Clinicians were then offered an opportunity to modify their initial answers. For the final two questions, after their initial response, clinicians were allowed to directly interact with the LLM to ask any questions before considering whether or not to modify their answer.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 12 references

Research works printed but not linked: 1

  1. Marshall C , Rossman GB . Designing Qualitative Research. 5th ed. Thousand Oaks, CA: SAGE Publications; 2010.

Cited by

See all 37

Metadata

Authors
Goh, E., Bunning, B., Khoong, E., Gallo, R., Milstein, A., Centola, D., & Chen, J. H.
Affiliations
  • Stanford University (US)
  • San Francisco General Hospital (US)
  • VA Palo Alto Health Care System (US)
  • University of Pennsylvania (US)
  • Stanford Medicine (US)
Corresponding author
Goh, Ethan
Venue
medRxiv
Publisher
openRxiv
Published
2023-11-27
OpenAlex
W4389041771
Cited by
37
References
12
Citations by year
2026: 6 · 2025: 16 · 2024: 14 · 2023: 1
Type
posted content
Language
en
Open access
Licence
cc-by
Locations
  • medRxiv (acceptedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (submittedVersion, cc-by)
Funders
  • U.S. Department of Veterans Affairs
  • Intermountain Healthcare
  • Peterson Center on Healthcare
  • National Institutes of Health
  • National Heart, Lung, and Blood Institute (K23HL157750)
  • National Institute on Drug Abuse (UG1DA015815)
  • U.S. National Library of Medicine (T15LM007033)
Topics
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Health Sciences › Medicine › Family Practice › Clinical Reasoning and Diagnostic Skills
  • Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
Keywords
ChatGPT, large language models, generative AI, clinical decision-making, physician bias, racial bias, gender bias, health care equity, chest pain evaluation, clinical vignettes, AI-assisted clinical decision support, physician attitudes toward AI

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: ChatGPT Influence on Medical Decision-Making, Bias, and Equity: A Randomized Study of Clinicians Evaluating Clinical Vignettes. https://app.cochat.ai/papers/10.1101/2023.11.24.23298844
The paper
Goh, E., Bunning, B., Khoong, E., Gallo, R., Milstein, A., Centola, D., & Chen, J. H. (2023). ChatGPT Influence on Medical Decision-Making, Bias, and Equity: A Randomized Study of Clinicians Evaluating Clinical Vignettes. medRxiv. https://doi.org/10.1101/2023.11.24.23298844

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.