Novel AI applications in systematic review: GPT-4 assisted data extraction, analysis, review of bias
The DOI matches its Crossref record; 49 of the 49 references it cites were found in a registry or an index.
Abstract
Objective: To assess custom GPT-4 performance in extracting and evaluating data from medical literature to assist in the systematic review (SR) process. Design: A proof-of-concept comparative study was conducted to assess the accuracy and precision of custom GPT-4 models against human-performed reviews of randomised controlled trials (RCTs). Setting: Four custom GPT-4 models were developed, each specialising in one of the following areas: (1) extraction of study characteristics, (2) extraction of outcomes, (3) extraction of bias assessment domains and (4) evaluation of risk of bias using results from the third GPT-4 model. Model outputs were compared against data from four SRs conducted by human authors. The evaluation focused on accuracy in data extraction, precision in replicating outcomes and agreement levels in risk of bias assessments. Participants: Among four SRs chosen, 43 studies were retrieved for data extraction evaluation. Additionally, 17 RCTs were selected for comparison of risk of bias assessments, where both human comparator SRs and an analogous SR provided assessments for comparison. Intervention: Custom GPT-4 models were deployed to extract data and evaluate risk of bias from selected studies, and their outputs were compared to those generated by human reviewers. Main outcome measures: Concordance rates between GPT-4 outputs and human-performed SRs in data extraction, effect size comparability and inter/intra-rater agreement in risk of bias assessments. Results: When comparing the automatically extracted data to the first table of study characteristics from the published review, GPT-4 showed 88.6% concordance with the original review, with <5% discrepancies due to inaccuracies or omissions. It exceeded human accuracy in 2.5% of instances. Study outcomes were extracted and pooling of results showed comparable effect sizes to comparator SRs. A review of bias assessment using GPT-4 showed fair-moderate but significant intra-rater agreement (ICC=0.518, p<0.001) and inter-rater agreements between human comparator SR (weighted kappa=0.237) and the analogous SR (weighted kappa=0.296). In contrast, there was a poor agreement between the two human-performed SRs (weighted kappa=0.094). Conclusion: Customized GPT-4 models perform well in extracting precise data from medical literature with potential for utilization in review of bias. While the evaluated tasks are simpler than the broader range of SR methodologies, they provide an important initial assessment of GPT-4's capabilities.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Moreover, several studies have relied on LLMs to assess their own outputs (12, 14, 21–24), raising concerns about reliability and bias.”
From Performance of large language models and prompt engineering strategies for data extraction in systematic reviews (2026) · cites this paper as [23]
Citations
Cites
A guideline of selecting and reporting intraclass correlation coefficients for reliability research
Koo, T. K., & Li, M. Y. (2016) · Journal of Chiropractic Medicine · doi:10.1016/j.jcm.2016.02.012 · Cited by 31,709
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
Interrater reliability: The kappa statistic
McHugh, M. L. (2012) · Biochemia Medica · doi:10.11613/bm.2012.031 · Cited by 19,877
Kaugeranna, R., Kaugeranna, E., & Aion, (. S. 4. 6. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.giscience.2023.43 · Cited by 13,690
Chain-Of-Thought Prompting Elicits Reasoning in Large Language Models
Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E., Le, Q. V., & Zhou, D. (2022) · Advances in Neural Information Processing Systems 35 · doi:10.52202/068431-1800 · Cited by 4,505
Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E., Le, Q., & Zhou, D. (2022) · arXiv · doi:10.48550/arxiv.2201.11903 · Cited by 4,263
Schröder, F. H., Hugosson, J., Roobol, M. J., Tammela, T. L. J., Zappa, M., Nelen, V., Kwiatkowski, M., Lujan, M., Määttänen, L., Lilja, H., et al. (2014) · The Lancet · doi:10.1016/s0140-6736(14)60525-0 · Cited by 1,566
Balanced crystalloids versus saline in critically ill adults.
Semler, M. W., Self, W. H., Wanderer, J. P., Ehrenfeld, J. M., Wang, L., Byrne, D. W., Stollings, J. L., Kumar, A. B., Hughes, C. G., Hernandez, A., et al. (2018) · New England Journal of Medicine · doi:10.1056/nejmoa1711584 · Cited by 1,433
Andriole, G. L., Crawford, E. D., Grubb, R. L., Buys, S. S., Chia, D., Church, T. R., Fouad, M. N., Isaacs, C., Kvale, P. A., Reding, D. J., et al. (2012) · JNCI Journal of the National Cancer Institute · doi:10.1093/jnci/djr500 · Cited by 1,121
Young, P., Bailey, M., Beasley, R., Henderson, S., Mackle, D., McArthur, C., McGuinness, S., Mehrtens, J., Myburgh, J., Psirides, A., et al. (2015) · JAMA · doi:10.1001/jama.2015.12334 · Cited by 728
Research works printed but not linked: 1
- Subramaniam, 2021, Single Dose of Oral Azithromycin With or Without Amoxicillin to Prevent Peripartum Infection in Laboring, High-Risk Women in Cameroon: A Randomized Controlled Trial, Obstet Gynecol, 138, 703
Cited by
Sebo, P., & Wang, T. (2025) · Family Practice · doi:10.1093/fampra/cmaf069 · Cited by 5
Oami, T., Okada, Y., Maeda, K., & Nakada, T.-A. (2026) · Frontiers in Digital Health · doi:10.3389/fdgth.2026.1799623 · Cited by 4
Park, Y., Zhang, H.-S., & Bai, S. W. (2026) · npj Digital Medicine · doi:10.1038/s41746-026-02431-w · Cited by 3
Fan, S., Chen, M., Doi, S. A. R., Ye, Z., Peng, Z., Tian, Y., Zhang, C., Furuya‐Kanamori, L., Lin, L., Mayo‐Wilson, E., et al. (2026) · BMJ Evidence-Based Medicine · doi:10.1136/bmjebm-2025-114044 · Cited by 5
Mentorship in Medical Writing: Shaping Future-Ready Publication Professionals
Toroser, D. (2025) · Trends in Scholarly Publishing · doi:10.21124/tsp.2025.54.58
Metadata
- Authors
- Kim, J. K., Chua, M. E., Li, T. G., Rickard, M., & Lorenzo, A. J.
- Affiliations
- University of Toronto (CA)
- Hospital for Sick Children (CA)
- Venue
- BMJ Evidence-Based Medicine
- Publisher
- BMJ
- Published
- 2025-04-08
- Volume
- 30
- Issue
- 5
- Pages
- bmjebm-2024-113066
- ISSN
- 1356-5524 (ISSN-L)
- 2515-446X
- 2515-4478
- OpenAlex
- W4409283522
- Semantic Scholar
- 56e583d66c7bbe3e64d2cec8f7ae47bbdd5c8daf
- Cited by
- 11
- References
- 46
- Citations by year
- 2026: 9 · 2025: 2
- FWCI
- 2.10
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- BMJ Evidence-Based Medicine (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › General Social Sciences › Computational and Text Analysis Methods
- Keywords
- large language models, systematic reviews, data extraction, risk of bias assessment, randomized controlled trials, AI-assisted literature review, interobserver agreement
- MeSH terms
- Systematic Reviews as Topic, Humans, Bias, Randomized Controlled Trials as Topic, Information Storage and Retrieval
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: Novel AI applications in systematic review: GPT-4 assisted data extraction, analysis, review of bias. https://app.cochat.ai/papers/10.1136/bmjebm-2024-113066
- The paper
- Kim, J. K., Chua, M. E., Li, T. G., Rickard, M., & Lorenzo, A. J. (2025). Novel AI applications in systematic review: GPT-4 assisted data extraction, analysis, review of bias. BMJ Evidence-Based Medicine. https://doi.org/10.1136/bmjebm-2024-113066
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.