FIRE-EVIDENCE:基于大语言模型的女性健康机器可读证据合成框架及性能评估
FIRE-EVIDENCE: enabling LLM-driven, machine-readable evidence synthesis for women’s health
npj Digital Medicine于2026年11月发表一篇研究论文,提出FIRE-EVIDENCE大语言模型框架,将全文临床研究报告自动转换为符合FHIR标准的结构化可计算证据。
- Article
- Open access
- Published: 17 November 2026
npj Digital Medicine (2026) Cite this article
We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.
Abstract
Clinical evidence is abundant in the published literature yet remains fragmented and underutilized, particularly in women’s health and menopause, where timely, personalized decision-making is critical. Conventional evidence synthesis relies on resource-intensive manual review, limiting scalability and delaying translation into practice, while automated approaches often produce outputs that are not directly computable or integrable into decision-support workflows. We present FIRE-EVIDENCE, a large language model framework that transforms full-text clinical studies into structured, Fast Healthcare Interoperability Resources (FHIR)-aligned computable evidence. The pipeline was optimized on seven development documents and evaluated on two datasets: a primary dataset of 26 expert-annotated randomized controlled trials and a secondary dataset of 18 studies derived from five systematic reviews. Across expert-annotated trials, FIRE-EVIDENCE achieved 94.35% completeness, 94.61% correctness, 97.36% source faithfulness, and an F1 score of 94.43%. FHIR-compliance testing in a 10-output native FHIR-R5 subset showed that all validated outputs passed base HL7 FHIR validation without errors. FIRE-EVIDENCE demonstrates a feasible pathway for standard-based, machine-readable evidence synthesis.
Acknowledgements
The authors gratefully acknowledge funding for the article processing charge through the International Science Partnerships Fund (ISPF) Fellowship programme, administered by the British Council (application reference: 1690). The research contributes to the ISPF project’s aim of strengthening early-career researchers’ expertise in large language models (LLMs) and retrieval-augmented generation (RAG) systems.
Funding
The authors received funding for the article processing charge through the International Science Partnerships Fund (ISPF) Fellowship programme, administered by the British Council (application reference: 1690).
Author information
Authors and Affiliations
Department of Computer Science, Birmingham City University, Birmingham, UK
Muhammad Ammar Shahid & Muhammad Afzal
Health Information Research Unit, Department of Health Research Methods, Evidence, and Impact, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada
Cynthia Lokker
Department of Engineering, Birmingham City University, Birmingham, UK
Wenyan Wu
College of Computing and Informatics University of Sharjah Sharjah, Sharjah, UAE
Huseyin Seker
Authors
- Muhammad Ammar Shahid
- Muhammad Afzal
- Cynthia Lokker
- Wenyan Wu
- Huseyin Seker
Corresponding author
Correspondence to Muhammad Afzal.
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Shahid, M.A., Afzal, M., Lokker, C. et al. FIRE-EVIDENCE: enabling LLM-driven, machine-readable evidence synthesis for women’s health. npj Digit. Med. (2026). https://doi.org/10.1038/s41746-026-03411-w
Received: 02 April 2026
Accepted: 04 October 2026
Published: 17 November 2026
DOI: https://doi.org/10.1038/s41746-026-03411-w
来源:Nature 健康护理主题 · nature.com