MétaCan
Menu
← Back to cohort
Record W4417498530 · doi:10.2196/84251

Evaluating the Accuracy of Medical Information Generated by ChatGPT and Gemini and Its Alignment With International Clinical Guidelines From the Surviving Sepsis Campaign: Comparative Study

2025· article· en· W4417498530 on OpenAlexvenueno aff
Dina Kutbi, Ehab Abou-Bakr, Hassan Mousa Haidar

Bibliographic record

VenueJMIR Formative Research · 2025
Typearticle
Languageen
FieldMedicine
TopicArtificial Intelligence in Healthcare and Education
Canadian institutionsnot available
Fundersnot available
KeywordsSepsisMEDLINEClinical PracticeMedical informationMedical literature

Abstract

fetched live from OpenAlex

BACKGROUND: Assessment of medical information provided by artificial intelligence (AI) chatbots like ChatGPT and Google's Gemini and comparison with international guidelines is a burgeoning area of research. These AI models are increasingly being considered for their potential to support clinical decision-making and patient education. However, their accuracy and reliability in delivering medical information that aligns with established guidelines remain under scrutiny. OBJECTIVE: This study aims to assess the accuracy of medical information generated by ChatGPT and Gemini and its alignment with international guidelines for sepsis management. METHODS: ChatGPT and Gemini were asked 18 questions about the Surviving Sepsis Campaign guidelines, and the responses were evaluated by 7 independent intensive care physicians. The responses generated were scored as follows: 3=correct, complete, and accurate; 2=correct but incomplete or inaccurate; and 1=incorrect. This scoring system was chosen to provide a clear and straightforward assessment of the accuracy and completeness of the responses. The Fleiss κ test was used to assess the agreement between evaluators, and the Mann-Whitney U test was used to test for the significance of differences between the correct responses generated by ChatGPT and Gemini. RESULTS: ChatGPT provided 5 (28%) perfect responses, 12 (67%) nearly perfect responses, and 1 (5%) low-quality response, with substantial agreement among the evaluators (Fleiss κ=0.656). Gemini, on the other hand, provided 3 (17%) perfect responses, 14 (78%) nearly perfect responses, and 1 (5%) low-quality response, with moderate agreement among the evaluators (Fleiss κ=0.582). The Mann-Whitney U test revealed no statistically significant difference between the two platforms (P=.48). CONCLUSIONS: ChatGPT and Gemini both demonstrated potential for generating medical information. Despite their current limitations, both showed promise as complementary tools in patient education and clinical decision-making. The medical information generated by ChatGPT and Gemini still needs ongoing evaluation regarding its accuracy and alignment with international guidelines in different medical domains, particularly in the sepsis field.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.057
metaresearch head score (Gemma)0.270
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesMetaresearch
Consensus categoriesnone
DomainCandidate signal: Evaluation · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: Observational
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.943
Threshold uncertainty score0.302

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0570.270
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0040.002
Science and technology studies0.0010.002
Scholarly communication0.0020.002
Open science0.0010.003
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0020.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.597
GPT teacher head0.673
Teacher spread0.075 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

Study designObservational
DomainEvaluation
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same venueJMIR Formative Research→Same topicArtificial Intelligence in Healthcare and Education→French-language works237,207→