AI-driven abstract generating: evaluating LLMs with a tailored prompt under the PRISMA-A framework
Open accessThe DOI matches its Crossref record; 20 of the 20 references it cites were found in a registry or an index.
Abstract
BACKGROUND: This study aimed to assess and compare ChatGPT-4o and Gemini Pro's ability to generate structured abstracts from full-text systematic reviews and meta-analyses in orthodontics, based on adherence to the PRISMA Abstract (PRISMA-A) Checklist, using a customised prompt developed for this purpose. MATERIALS AND METHODS: A total of 162 full-text systematic reviews and meta-analyses published in Q1-ranked orthodontic journals since January 2019 were included. Each full-text article was processed by ChatGPT-4o and Gemini Pro, using a PRISMA-A Checklist-aligned structured prompt. Outputs were scored using a tailored Overall quality Score OQS derived from 11 PRISMA-A checklist. Inter-rater and time-dependent reliability were assessed with Intraclass Correlation Coefficients (ICCs), and model outputs were compared using Mann-Whitney U tests. RESULTS: Both models yielded satisfactory OQS in generating PRISMA-A checklist compliant abstracts; however, ChatGPT-4o consistently achieved higher scores than Gemini Pro. The most notable differences were observed in the "Included Studies" and "Synthesis of Results" sections, where ChatGPT-4o produced more complete and structurally coherent outputs. ChatGPT-4o achieved a mean OQS of 21.67 (SD 0.58) versus 21.00 (SD 0.71) for Gemini Pro, a difference that was highly significant (p < 0.001). CONCLUSIONS: Both LLMs demonstrated the ability to generate PRISMA-A-compliant abstracts from systematic reviews, with ChatGPT-4o consistently achieving higher quality scores than Gemini Pro. While tested in orthodontics, the approach holds potential for broader applications across evidence-based dental and medical research. Systematic reviews and meta-analyses are essential to evidence-based dentistry but can be challenging and time-consuming to report in accordance with established standards. The structured prompt developed in this study may assist researchers in generating PRISMA-A-compliant outputs more efficiently, helping to accelerate the completion and standardisation of high-level clinical evidence reporting.
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
Figures





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.
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?
The cited source says what the claim attributes to it.
In the citing paper
“demonstrating that checklist-based evaluation and structured prompting can enhance output quality (31).”
From Performance of large language models and prompt engineering strategies for data extraction in systematic reviews (2026) · cites this paper as [31]
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
Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gotzsche, P. C., Ioannidis, J. P. A., Clarke, M., Devereaux, P. J., Kleijnen, J., & Moher, D. (2009) · BMJ · doi:10.1136/bmj.b2700 · Cited by 17,648
Large language models encode clinical knowledge
Singhal, K., Azizi, S., Tu, T., Mahdavi, S. S., Wei, J., Chung, H. W., Scales, N., Tanwani, A., Cole-Lewis, H., Pfohl, S., et al. (2023) · Nature · doi:10.1038/s41586-023-06291-2 · Cited by 4,083 · cited 1×
“By leveraging transformer architectures, these models assign weighted importance to different parts of a text, allowing them to prioritise key information while maintaining coherence and logical flow [4, 5].”
Siddaway, A. P., Wood, A. M., & Hedges, L. V. (2019) · Annual Review of Psychology · doi:10.1146/annurev-psych-010418-102803 · Cited by 2,550 · cited 3×
“Systematic reviews and meta-analyses involve multiple critical steps, all requiring a high level of methodological rigor [6, 7].”
Gao, C. A., Howard, F. M., Markov, N. S., Dyer, E. C., Ramesh, S., Luo, Y., & Pearson, A. T. (2023) · npj Digital Medicine · doi:10.1038/s41746-023-00819-6 · Cited by 1,099 · cited 1×
“Previous studies have evaluated the ability of LLMs to generate scientific abstracts without structured guidance [11, 12].”
Structured information extraction from scientific text with large language models
Dagdelen, J., Dunn, A., Lee, S., Walker, N., Rosen, A. S., Ceder, G., Persson, K. A., & Jain, A. (2024) · Nature Communications · doi:10.1038/s41467-024-45563-x · Cited by 771 · cited 3×
“Similarly, Dagdelen et al. [10] found that fine-tuning and structured prompting improved model performance in scientific text processing, outperforming traditional methods in accuracy and information extraction.”
Generating scholarly content with ChatGPT: Ethical challenges for medical publishing
Liebrenz, M., Schleifer, R., Buadze, A., Bhugra, D., & Smith, A. (2023) · The Lancet Digital Health · doi:10.1016/s2589-7500(23)00019-5 · Cited by 594 · cited 1×
“LLMs can serve as effective assistants, but they cannot replace the depth of domain expertise, critical judgement, and ethical responsibility that underpin high-quality scientific reporting [3, 19, 20].”
Unleashing the potential of prompt engineering for large language models
Chen, B., Zhang, Z., Langrené, N., & Zhu, S. (2025) · Patterns · doi:10.1016/j.patter.2025.101260 · Cited by 490
Evaluating large language models on medical evidence summarization
Tang, L., Sun, Z., Idnay, B., Nestor, J. G., Soroush, A., Elias, P. A., Xu, Z., Ding, Y., Durrett, G., Rousseau, J. F., et al. (2023) · npj Digital Medicine · doi:10.1038/s41746-023-00896-7 · Cited by 423
Evaluating Large Language Models on Medical Evidence Summarization
Tang, L., Sun, Z., Idnay, B., Nestor, J. G., Soroush, A., Elias, P. A., Xu, Z., Ding, Y., Durrett, G., Rousseau, J., et al. (2023) · doi:10.1101/2023.04.22.23288967 · Cited by 423 · cited 1×
“By leveraging transformer architectures, these models assign weighted importance to different parts of a text, allowing them to prioritise key information while maintaining coherence and logical flow [4, 5].”
ChatGPT and Publication Ethics
Rahimi, F., & Talebi Bezmin Abadi, A. (2023) · Archives of Medical Research · doi:10.1016/j.arcmed.2023.03.004 · Cited by 158 · cited 1×
“LLMs can serve as effective assistants, but they cannot replace the depth of domain expertise, critical judgement, and ethical responsibility that underpin high-quality scientific reporting [3, 19, 20].”
Research works printed but not linked: 3
- Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Revista Esp De Nutricion Humana Y Dietetica. 2016;20:148–60. https://doi.org/10.1186/2046-4053-4-1/TABLES/4. · doi:10.1186/2046-4053-4-1/tables/4
- Lakens D. Calculating and reporting effect sizes to facilitate cumulative science: A practical primer for t-tests and ANOVAs. Front Psychol. 2013;4:62627. https://doi.org/10.3389/FPSYG.2013.00863/BIBTEX. · doi:10.3389/fpsyg.2013.00863/bibtex
- AlSagri HS, Farhat F, Sohail SS, Saudagar AKJ. ChatGPT or gemini: who makes the better scientific writing assistant? J Acad Ethics. 2024;1–15. https://doi.org/10.1007/S10805-024-09549-0/FIGURES/3. · doi:10.1007/s10805-024-09549-0/figures/3
Cited by
Oami, T., Okada, Y., Maeda, K., & Nakada, T.-A. (2026) · Frontiers in Digital Health · doi:10.3389/fdgth.2026.1799623 · Cited by 4
Uranbey, Ö., Özbey, F., & Kaygisiz, Ö. F. (2026) · Scientific Reports · doi:10.1038/s41598-026-50239-1
“Similarly, Demir et al. reported that PICO-based tailored prompts enhanced methodological coherence in LLM-generated orthodontic systematic review abstracts, in some instances exerting a stronger influence than the underlying model version itself 26 .”
InfluentialRomero, M. Á. M. (2026) · Dilemas contemporáneos Educación Política y Valores · doi:10.46377/dilemas.v14i1.4998
Adam, G. P., Davies, M. C., Caputo, E. L., Mai, H. J., Coppola, E. L., Holmer, H. K., Kuhn, E., Balk, E. M., Trikalinos, T. A., Ivlev, I., et al. (2026) · Research Synthesis Methods · doi:10.1017/rsm.2026.10118
Metadata
- Authors
- Demi̇r, G. B., Gökmen, Ş., Süküt, Y., Topsakal, K. G., & Görgülü, S.
- Affiliations
- Sağlık Bilimleri Üniversitesi (TR)
- Corresponding author
- Demi̇r, Gizem Boztaş
- Venue
- BMC Oral Health
- Published
- 2025-10-10
- Volume
- 25
- Issue
- 1
- Pages
- 1594-1594
- ISSN
- 1472-6831 (ISSN-L)
- PMC
- PMC12512598
- OpenAlex
- W4415024443
- Cited by
- 2
- References
- 17
- Citations by year
- 2026: 2
- FWCI
- 0.48
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by-nc-nd
- creativecommons.org/licenses/by-nc-nd/4.0 (tdm)
- creativecommons.org/licenses/by-nc-nd/4.0 (vor)
- Locations
- BMC Oral Health (publishedVersion, cc-by-nc-nd)
- Europe PMC (PubMed Central) (publishedVersion, cc-by-nc-nd)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Social Sciences › Library and Information Sciences › Academic Writing and Publishing
- Keywords
- large language models, abstract generation, GPT-4o, Gemini Pro, systematic reviews, meta-analysis, orthodontics, prompt engineering, reporting quality
- MeSH terms
- Abstracting and Indexing, Systematic Reviews as Topic, Humans, Orthodontics, Meta-Analysis as Topic, Reproducibility of Results, Checklist
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: AI-driven abstract generating: evaluating LLMs with a tailored prompt under the PRISMA-A framework. https://app.cochat.ai/papers/10.1186/s12903-025-06982-4
- The paper
- Demi̇r, G. B., Gökmen, Ş., Süküt, Y., Topsakal, K. G., & Görgülü, S. (2025). AI-driven abstract generating: evaluating LLMs with a tailored prompt under the PRISMA-A framework. BMC Oral Health. https://doi.org/10.1186/s12903-025-06982-4
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.