Academic Journal
DispatchMAS: fusing taxonomy and artificial intelligence agents for emergency medical services.
| Τίτλος: | DispatchMAS: fusing taxonomy and artificial intelligence agents for emergency medical services. |
|---|---|
| Συγγραφείς: | Li X; Shandong University, Jinan, Shandong, China., Yu H; The Chinese University of Hong Kong, Sha Tin, NT, Hong Kong SAR, China., Wang W; Shandong University, Jinan, Shandong, China., Wu Y; Pennsylvania State University, University Park, PA, USA., Zhou J; Stanford University School of Medicine, Stanford, CA, USA., Hua W; University of California, Santa Barbara, CA, USA., Lin X; The Chinese University of Hong Kong, Sha Tin, NT, Hong Kong SAR, China., Tan W; University of Macau, Macau SAR, China., Zhu L; New York University, New York, NY, USA., Chen B; The Chinese University of Hong Kong, Sha Tin, NT, Hong Kong SAR, China., Chen G; Broad Institute of MIT and Harvard, Cambridge, MA, USA.; The University of Hong Kong, Hong Kong SAR, China., Chen ML; Stanford University School of Medicine, Stanford, CA, USA., Zhou Y; Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China., Li Z; Chinese Academy of Medical Sciences, Beijing, China., Assimes TL; Stanford University School of Medicine, Stanford, CA, USA., Zhang Y; Rutgers University, New Brunswick, NJ, USA., Wu Q; Pennsylvania State University, University Park, PA, USA., Ma X; Shandong University, Jinan, Shandong, China. maxin@sdu.edu.cn., Li L; University of South Florida, Tampa, FL, USA. lingyaol@sfu.edu., Fan L; The Chinese University of Hong Kong, Sha Tin, NT, Hong Kong SAR, China. leofan@cuhk.edu.hk. |
| Πηγή: | BMC emergency medicine [BMC Emerg Med] 2026 Mar 18; Vol. 26 (1). Date of Electronic Publication: 2026 Mar 18. |
| Τύπος έκδοσης: | Journal Article |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: BioMed Central Country of Publication: England NLM ID: 100968543 Publication Model: Electronic Cited Medium: Internet ISSN: 1471-227X (Electronic) Linking ISSN: 1471227X NLM ISO Abbreviation: BMC Emerg Med Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: London : BioMed Central, [2001- |
| Ιατρικοί όροι (MeSH): | Artificial Intelligence* , Emergency Medical Services* , Emergency Medical Service Communication Systems* , Emergency Medical Dispatch*, Humans ; Large Language Models ; Intelligent Systems ; Electronic Health Records |
| Περίληψη: | BACKGROUND: Emergency medical dispatch is a critical, high-stakes process where dispatcher decisions directly impact patient outcomes. While standardized protocols exist, they are challenged by factors like caller distress, ambiguous symptom descriptions, and high cognitive load. The convergence of Large Language Models (LLMs) and Multi-Agent Systems (MAS) offers a novel opportunity to augment human dispatchers. This study aimed to develop and evaluate a taxonomy-grounded, LLM-powered multi-agent system for simulating realistic clinician’s medical dispatch scenarios. METHODS: We first constructed a clinically curated taxonomy and fact commons for emergency dispatch, defining 32 Chief Complaints based on national standards, six distinct caller identities derived from real-world electronic health records (Medical Information Mart for Intensive Care III [MIMIC-III]), and a standardized six-phase call protocol. Using this framework, we developed a multi-agent simulation system featuring a Caller Agent and a Dispatcher Agent. The system, built on the AutoGen multi-agent framework for large language models (AutoGen), grounds agent interactions in the fact commons to ensure clinical plausibility and mitigate misinformation. We designed a hybrid evaluation combining expert clinical assessment, automated linguistic analysis, and operational performance dynamics auditing. Four physicians evaluated 100 simulated dispatch cases for “Guidance Efficacy” and “Dispatch Effectiveness” using a structured questionnaire. Automated metrics assessed sentiment, emotion, readability, and politeness of agent-generated dialogue. Operational performance dynamics analyses showed phase-dependent efficiency peaking during Assessment and faster pacing in life-critical events. RESULTS: Human evaluation, with substantial inter-rater agreement (Gwet’s AC1 [Formula: see text], confirmed the system’s high performance. It demonstrated excellent Dispatch Effectiveness (e.g., 94% contacting the correct potential other agents) and Guidance Efficacy (advice provided in 91% of cases), both rated highly by physicians. Algorithmic metrics corroborated these findings, indicating a predominantly neutral affective profile (73.7% neutral sentiment; 90.4% neutral emotion), high readability (Flesch 80.9), and a consistently polite style (60.0% polite; 0% impolite). Operational performance evaluation further showed urgency-adaptive pacing: for life-critical events, information completeness rose faster and plateaued earlier while converging to comparable end-of-call completeness across complaint types. The agent also responded more rapidly in life-critical scenarios (1.8 s per dispatcher turn vs 2.1–2.4 s), indicating accelerated early questioning without loss of overall coverage. CONCLUSION: Our LLM-based MAS simulates diverse, clinically plausible dispatch scenarios with high fidelity. The resulting platform provides a controlled environment for analyzing dispatcher–caller interactions, stress-testing protocol variants, and deriving structured design patterns that may inform future real-time decision support. Our simulation-based tools could serve as an intermediate step between offline method development and eventual integration into emergency response workflows. |
| Competing Interests: | Declarations. Ethics approval and consent to participate: This study was approved by Shandong University Qilu Hospital Ethics Committee Review Board (IRB No. KYLL-202505-005) and conducted in accordance with the principles of the Declaration of Helsinki. The requirement for informed consent was waived due to the retrospective analysis of publicly available anonymized data. Consent for publication: Not applicable. Competing interests: The authors declare no competing interests. |
| References: | J Emerg Med. 2012 Dec;43(6):1132-7. (PMID: 23047196) BMC Emerg Med. 2024 Jul 18;24(1):120. (PMID: 39020318) Emerg Med J. 2019 Oct;36(10):595-600. (PMID: 31439715) Front Public Health. 2023 Feb 20;11:1121779. (PMID: 36891343) BMC Emerg Med. 2024 Jul 9;24(1):111. (PMID: 38982356) Br J Gen Pract. 2019 Jan;69(678):e70-e79. (PMID: 30510099) BMJ Open. 2022 May 13;12(5):e059129. (PMID: 35568491) BMJ Open. 2019 Jul 1;9(7):e027477. (PMID: 31266836) CJEM. 2005 Jan;7(1):5-11. (PMID: 17355647) Ann R Coll Surg Engl. 2008 Mar;90(2):113-6. (PMID: 18325208) Ann Emerg Med. 1985 Nov;14(11):1055-60. (PMID: 4051269) JAMA Netw Open. 2024 Dec 2;7(12):e2448723. (PMID: 39625719) BMC Med Educ. 2025 May 7;25(1):666. (PMID: 40336004) J Gen Intern Med. 2022 May;37(7):1797-1798. (PMID: 34240287) Patient Educ Couns. 2024 May;122:108161. (PMID: 38308973) Nat Commun. 2024 Oct 8;15(1):8236. (PMID: 39379357) Resuscitation. 2020 Jan 1;146:5-12. (PMID: 31706968) World J Emerg Med. 2011;2(4):245-52. (PMID: 25215018) BMC Emerg Med. 2015 Oct 27;15:32. (PMID: 26507648) J Healthc Inform Res. 2024 Sep 14;8(4):658-711. (PMID: 39463859) Am J Emerg Med. 2025 Mar;89:174-181. (PMID: 39731895) Bull Emerg Trauma. 2017 Jul;5(3):171-178. (PMID: 28795061) BMJ Qual Saf. 2024 Aug 16;33(9):573-586. (PMID: 38050161) Sci Data. 2016 May 24;3:160035. (PMID: 27219127) Am J Emerg Med. 2025 Mar;89:27-29. (PMID: 39675178) Resusc Plus. 2022 Jun 28;11:100264. (PMID: 35801232) Am J Emerg Med. 1996 Jan;14(1):19-22. (PMID: 8630148) JMIR Med Inform. 2024 May 10;12:e53787. (PMID: 38728687) |
| Contributed Indexing: | Keywords: Emergency Medical Services; Fact Commons; Multi-Agent Systems; Prehospital Care; Taxonomy |
| Entry Date(s): | Date Created: 20260319 Date Completed: 20260628 Latest Revision: 20260821 |
| Update Code: | 20260822 |
| PubMed Central ID: | PMC13001355 |
| DOI: | 10.1186/s12873-026-01540-9 |
| PMID: | 41851670 |
| Βάση Δεδομένων: | MEDLINE |
| ISSN: | 1471-227X |
|---|---|
| DOI: | 10.1186/s12873-026-01540-9 |