EUVIMEDObjevte EUVIMED
Multizentrische StudieValidierungsstudieOpen access

Large Language Model–Based Clinical Decision Support for Antibiotic Selection and Dose Recommendation in Hospitalized Patients With Pneumonia: Multicenter Retrospective Study

Yang Zhang, Li Li, Chunting Tan iD, Mengyuan Ji, Xican Tian, Xiangdong Mu, Jun Li, Yu Gu, Honglei Liu iD

JMIR Medical Informatics · 2026-08-04 · 14 · e98207-e98207

DOI: 10.2196/98207 ↗

At a glance

Multizentrische Studie in JMIR Medical Informatics, published online on 2026-08-04. Written by Yang Zhang and 8 co-authors. Institution: School of Biomedical Engineering, Capital Medical University, No. 10 Xitoutiao, You'anmenwai, Fengtai District, Beijing, 100069, China, 86 010-839115…. Major MEDLINE topics: Anti-Bacterial Agents, Decision Support Systems, Clinical, Large Language Models, Pneumonia. Cited 3 times so far (Crossref); 31 references. The full text is freely available in PubMed Central.

AbstraktSubjectsBibliografické údajeCiteRelated studies

Abstract (original language)

Background

Pneumonia is a common infectious disease, and antibiotic treatment in hospitalized patients must balance efficacy, safety, and resistance risk. However, antibiotic selection and dose adjustment still rely heavily on clinician experience. Although large language models (LLMs) are promising for clinical reasoning, their direct use for antibiotic selection and dose recommendation is limited by hallucinations and weak adherence to clinical constraints.

Objective

This study aimed to develop and externally validate a constrained LLM-based clinical decision support pipeline for antibiotic selection and dose recommendation in hospitalized patients with pneumonia.

Methods

We conducted a multicenter retrospective study using electronic health record narratives, antibiotic orders, and laboratory indicators of hepatic and renal function from 331 hospitalized patients with pneumonia from 2 hospitals in China. The development cohort included 233 patients, and the external validation cohort included 98 patients. The pipeline integrated dual-branch retrieval (similar-case vector retrieval plus guideline-based knowledge graph retrieval), clinician-defined rule constraints, and hybrid-context reasoning. DeepSeek-V3, GLM-4.6, and GPT-4o were evaluated using F 1 -score and Jaccard accuracy.

Results

On the internal test set, the full pipeline using DeepSeek-V3 achieved the best performance, with an F 1 -score of 0.8110 (95% CI 0.7371-0.8762) and Jaccard accuracy of 0.7624 (95% CI 0.6810-0.8386) for antibiotic selection and an F 1 -score of 0.7538 (95% CI 0.6671-0.8329) and Jaccard accuracy of 0.7076 (95% CI 0.6145-0.7938) for joint antibiotic selection plus dosing recommendation. On the external validation set, performance remained high, with an F 1 -score of 0.8605 (95% CI 0.7891-0.9252) and Jaccard accuracy of 0.8571 (95% CI 0.7857-0.9184) for antibiotic selection, and an F 1 -score of 0.8503 (95% CI 0.7789-0.9150) and Jaccard accuracy of 0.8469 (95% CI 0.7755-0.9133) for antibiotic selection plus dosing recommendation. The system also provided traceable evidence and rule trigger information to support clinician review.

Conclusions

A constrained, retrieval-augmented LLM pipeline improved the consistency and interpretability of antibiotic selection and dose recommendation for hospitalized patients with pneumonia and provided preliminary evidence of cross-site generalizability.

Subjects

MeSH terms (MEDLINE)

Anti-Bacterial AgentsDecision Support Systems, ClinicalLarge Language ModelsPneumoniaAgedFemaleHumansMaleMiddle AgedChinaElectronic Health RecordsHospitalizationRetrospective Studies

Author keywords

knowledge graphlarge language modelsmedication recommendationpneumoniaretrieval-augmented generation

Substances

Anti-Bacterial Agents

Authors and institutions

  • School of Biomedical Engineering, Capital Medical University, No. 10 Xitoutiao, You'anmenwai, Fengtai District, Beijing, 100069, China, 86 010-83911542.
  • Beijing Key Laboratory of Clinical Engineering Solutions for Mental Health, Capital Medical University, Beijing, China.
  • Department of Respiratory and Critical Care Medicine, School of Clinical Medicine, Tsinghua Medicine, Beijing Tsinghua Changgung Hospital, Tsinghua University, Beijing, China.
  • Department of Respiratory Medicine, Beijing Friendship Hospital, Capital Medical University, Beijing, China.

ORCID profiles: Chunting Tan, Honglei Liu

Cite

APA

Zhang, Y., Li, L., Tan, C., Ji, M., Tian, X., Mu, X., Li, J., Gu, Y., & Liu, H. (2026). Large Language Model–Based Clinical Decision Support for Antibiotic Selection and Dose Recommendation in Hospitalized Patients With Pneumonia: Multicenter Retrospective Study. JMIR Medical Informatics, 14, e98207-e98207. https://doi.org/10.2196/98207

Vancouver

Zhang Y, Li L, Tan C, Ji M, Tian X, Mu X, et al. Large Language Model–Based Clinical Decision Support for Antibiotic Selection and Dose Recommendation in Hospitalized Patients With Pneumonia: Multicenter Retrospective Study. JMIR Medical Informatics. 2026;14:e98207-e98207. doi:10.2196/98207

BibTeX
@article{Zhang2026,
  author = {Zhang, Yang and Li, Li and Tan, Chunting and Ji, Mengyuan and Tian, Xican and Mu, Xiangdong and Li, Jun and Gu, Yu and Liu, Honglei},
  title = {Large Language Model–Based Clinical Decision Support for Antibiotic Selection and Dose Recommendation in Hospitalized Patients With Pneumonia: Multicenter Retrospective Study},
  journal = {JMIR Medical Informatics},
  year = {2026},
  volume = {14},
  pages = {e98207--e98207},
  doi = {10.2196/98207}
}
Start for free →