Large Language Models for Therapy Recommendations Across 3 Clinical Specialties: Comparative Study
Open accessThe DOI matches its Crossref record; 10 of the 10 references it cites were found in a registry or an index.
Abstract
Background: As advancements in artificial intelligence (AI) continue, large language models (LLMs) have emerged as promising tools for generating medical information. Their rapid adaptation and potential benefits in health care require rigorous assessment in terms of the quality, accuracy, and safety of the generated information across diverse medical specialties. Objective: This study aimed to evaluate the performance of 4 prominent LLMs, namely, Claude-instant-v1.0, GPT-3.5-Turbo, Command-xlarge-nightly, and Bloomz, in generating medical content spanning the clinical specialties of ophthalmology, orthopedics, and dermatology. Methods: Three domain-specific physicians evaluated the AI-generated therapeutic recommendations for a diverse set of 60 diseases. The evaluation criteria involved the mDISCERN score, correctness, and potential harmfulness of the recommendations. ANOVA and pairwise t tests were used to explore discrepancies in content quality and safety across models and specialties. Additionally, using the capabilities of OpenAI’s most advanced model, GPT-4, an automated evaluation of each model’s responses to the diseases was performed using the same criteria and compared to the physicians’ assessments through Pearson correlation analysis. Results: Claude-instant-v1.0 emerged with the highest mean mDISCERN score (3.35, 95% CI 3.23-3.46). In contrast, Bloomz lagged with the lowest score (1.07, 95% CI 1.03-1.10). Our analysis revealed significant differences among the models in terms of quality (P<.001). Evaluating their reliability, the models displayed strong contrasts in their falseness ratings, with variations both across models (P<.001) and specialties (P<.001). Distinct error patterns emerged, such as confusing diagnoses; providing vague, ambiguous advice; or omitting critical treatments, such as antibiotics for infectious diseases. Regarding potential harm, GPT-3.5-Turbo was found to be the safest, with the lowest harmfulness rating. All models lagged in detailing the risks associated with treatment procedures, explaining the effects of therapies on quality of life, and offering additional sources of information. Pearson correlation analysis underscored a substantial alignment between physician assessments and GPT-4’s evaluations across all established criteria (P<.01). Conclusions: This study, while comprehensive, was limited by the involvement of a select number of specialties and physician evaluators. The straightforward prompting strategy (“How to treat…”) and the assessment benchmarks, initially conceptualized for human-authored content, might have potential gaps in capturing the nuances of AI-driven information. The LLMs evaluated showed a notable capability in generating valuable medical content; however, evident lapses in content quality and potential harm signal the need for further refinements. Given the dynamic landscape of LLMs, this study’s findings emphasize the need for regular and methodical assessments, oversight, and fine-tuning of these AI tools to ensure they produce consistently trustworthy and clinically safe medical advice. Notably, the introduction of an auto-evaluation mechanism using GPT-4, as detailed in this study, provides a scalable, transferable method for domain-agnostic evaluations, extending beyond therapy recommendation assessments.
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.
Integrity signals
Registry
Cites
Statistics
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.
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
Rajpurkar, P., Chen, E., Banerjee, O., & Topol, E. J. (2022) · Nature Medicine · doi:10.1038/s41591-021-01614-0 · Cited by 3,212 · cited 1×
“Artificial intelligence (AI) will have a far-reaching impact on medicine and has the potential to make health care more efficient, precise, and accessible for patients [1].”
Internet Health Information Seeking and the Patient-Physician Relationship: A Systematic Review
Tan, S. S.-L., & Goonawardene, N. (2017) · Journal of Medical Internet Research · doi:10.2196/jmir.5729 · Cited by 1,101
Johnson, D., Goodman, R., Patrinely, J., Stone, C., Zimmerman, E., Donald, R., Chang, S., Berkowitz, S., Finn, A., Jahangir, E., et al. (2023) · doi:10.21203/rs.3.rs-2566942/v1 · Cited by 598 · cited 1×
“Through an internal validation process, questions that received lower ratings were subjected to retesting after a period of 8 to 17 days, resulting in a significant enhancement of answer quality [11].”
Fine-Tuning Language Models from Human Preferences
Ziegler, D. M., Stiennon, N., Wu, J., Brown, T. B., Radford, A., Amodei, D., Christiano, P., & Irving, G. (2019) · arXiv · doi:10.48550/arxiv.1909.08593 · Cited by 413
AI in Health: State of the Art, Challenges, and Future Directions
Wang, F., & Preininger, A. (2019) · Yearbook of Medical Informatics · doi:10.1055/s-0039-1677908 · Cited by 398
Crosslingual Generalization through Multitask Finetuning
Muennighoff, N., Wang, T., Sutawika, L., Roberts, A., Biderman, S., Le Scao, T., Bari, M. S., Shen, S., Yong, Z. X., Schoelkopf, H., et al. (2023) · Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/2023.acl-long.891 · Cited by 361 · cited 2×
“BigScience’s Bloomz [10] model is a collaborative AI project emphasizing research, ethical considerations, and application development in diverse domains.”
Surgical versus conservative treatment for lumbar disc herniation: a prospective cohort study
Gugliotta, M., da Costa, B. R., Dabis, E., Theiler, R., Jüni, P., Reichenbach, S., Landolt, H., & Hasler, P. (2016) · BMJ Open · doi:10.1136/bmjopen-2016-012938 · Cited by 215 · cited 1×
“…benefits (Q7), and effect on the quality of life (Q8) may be attributed to the intuitive and relatively simple nature of conservative therapies, such as rest, ice, compression, and elevation, as well as common treatment protocols involving physical therapy and pharmacological interventions [31-33].”
Patients’ recollection and understanding of informed consent: a literature review
Sherlock, A., & Brownie, S. (2014) · ANZ Journal of Surgery · doi:10.1111/ans.12555 · Cited by 191 · cited 1×
“Only 21%-86% of patients can recall the possible risks and complications of the procedures, and patient understanding appears to decrease with age [17].”
Evaluating the use of large language model in identifying top research questions in gastroenterology
Lahat, A., Shachar, E., Avidan, B., Shatz, Z., Glicksberg, B. S., & Klang, E. (2023) · Scientific Reports · doi:10.1038/s41598-023-31412-2 · Cited by 111
On the cusp: considering the impact of artificial intelligence language models in healthcare
Goodman, R. S., Patrinely, J. R., Osterman, T., Wheless, L., & Johnson, D. B. (2023) · Med · doi:10.1016/j.medj.2023.02.008 · Cited by 100 · cited 1×
“Further research is imperative to harness the full potential of LLMs as safe and dependable tools within the health care domain [14].”
Cited by
Bedi, S., Liu, Y., Orr-Ewing, L., Dash, D., Koyejo, S., Callahan, A., Fries, J. A., Wornow, M., Swaminathan, A., Lehmann, L. S., et al. (2024) · doi:10.1101/2024.04.15.24305869 · Cited by 738
“Using GPT for therapy recommendations in mental health (Patient Information in Periodontology (Wilhelm et al) 24”
Testing and Evaluation of Health Care Applications of Large Language Models: A Systematic Review
Bedi, S., Liu, Y., Orr-Ewing, L., Dash, D., Koyejo, S., Callahan, A., Fries, J. A., Wornow, M., Swaminathan, A., Lehmann, L. S., et al. (2025) · JAMA · doi:10.1001/jama.2024.21700 · Cited by 662
Tam, T. Y. C., Sivarajkumar, S., Kapoor, S., Stolyar, A. V., Polanska, K., McCarthy, K. R., Osterhoudt, H., Wu, X., Visweswaran, S., Fu, S., et al. (2024) · npj Digital Medicine · doi:10.1038/s41746-024-01258-7 · Cited by 424
A systematic review of large language model (LLM) evaluations in clinical medicine
Shool, S., Adimi, S., Saboori Amleshi, R., Bitaraf, E., Golpira, R., & Tara, M. (2025) · BMC Medical Informatics and Decision Making · doi:10.1186/s12911-025-02954-4 · Cited by 346
Development and Evaluation of a Retrieval-Augmented Large Language Model Framework for Ophthalmology
Luo, M.-J., Pang, J., Bi, S., Lai, Y., Zhao, J., Shang, Y., Cui, T., Yang, Y., Lin, Z., Zhao, L., et al. (2024) · JAMA Ophthalmology · doi:10.1001/jamaophthalmol.2024.2513 · Cited by 70
Metadata
- Authors
- Wilhelm, T. I., Roos, J., & Kaczmarczyk, R.
- Venue
- Journal of Medical Internet Research
- Publisher
- JMIR Publications Inc.
- Published
- 2023-10-30
- Volume
- 25
- Pages
- e49324
- ISSN
- 1438-8871 (ISSN-L)
- OpenAlex
- W4388014051
- Cited by
- 133
- References
- 10
- FWCI
- 4.60
- Type
- journal article
- Language
- en
- Open access
- gold
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: Large Language Models for Therapy Recommendations Across 3 Clinical Specialties: Comparative Study. https://app.cochat.ai/papers/10.2196/49324
- The paper
- Wilhelm, T. I., Roos, J., & Kaczmarczyk, R. (2023). Large Language Models for Therapy Recommendations Across 3 Clinical Specialties: Comparative Study. Journal of Medical Internet Research. https://doi.org/10.2196/49324
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.