本研究旨在應用機器學習方法預測到院前重大外傷患者，強調119緊急醫療服務與醫院急診室的橫向管理。外傷是全球年輕人口的主要死因之一，對社會和經濟帶來重大影響。透過分析高雄市政府消防局緊急醫療服務系統及重度急救責任醫院的外傷登錄資料，建立多種機器學習模型，包括隨機森林、神經網路、Gradient Boosting和XGBoost，以達到在到院前預測重大外傷患者之目標。研究結果顯示，這些機器學習模型於辨識重大外傷患者、提高到院前預警效能及改善檢傷不足和過度檢傷的情況方面具有顯著效果。
研究結果中另有發現，預測重大外傷模型中，採用只考慮年齡、受傷機轉及受傷部位之資料集，亦有良好的預測效果。洞見分析中，深入探討變數重要性和影響程度及其解釋意義等。本研究結論強調跨部門資料整合之重要性，以EMS觀點出發，藉由資料科學分析建立預測模型以提供緊急救護技術員決策之參考依據，達到提升緊急救護品質及醫療處置效能之目標。未來研究方向可著重考慮更多變項和資料來源，並探討結合通訊技術以邁向智慧醫療。

This study aims to apply machine learning methods to predict pre-hospital major trauma patients, emphasizing the horizontal management between 119 emergency medical services and hospital emergency departments. Trauma is a leading cause of death among young people worldwide, significantly impacting society and the economy. By analyzing the EMS system of the Kaohsiung City Government Fire Department and the trauma registry data of a designated heavy-duty emergency hospital, various machine learning models, including Random Forest, Neural Networks, Gradient Boosting, and XGBoost, were established to achieve the goal of predicting major trauma patients before hospital arrival. The results show that these machine learning models are effective in identifying major trauma patients, enhancing pre-hospital alert efficiency, and addressing under-triage and over-triage issues.
The research results revealed that using a dataset that only considers age, injury mechanism, and injury location also yields good predictive performance in the major trauma prediction models. The insight analysis delved into the importance and impact of variables and their interpretative significance. This study concludes by emphasizing the importance of cross-departmental data integration. From the EMS perspective, predictive models were established through data science analysis to provide decision-making references for emergency medical technicians, aiming to improve the quality of emergency care and the effectiveness of medical treatment. Future research directions may focus on considering more variables and data sources and exploring the integration of communication technology to advance towards smart healthcare.

中文文獻
王治元（2007）。電子病歷資訊儲存平台的未來（碩士論文）。國立臺灣大學管理學院高階公共管理組。
衛生福利部（2024）。112年國人死因統計結果。（擷取於2024年6月20日）https://www.mohw.gov.tw/cp-16-74869-1.html
衛生福利部中央健康保險署（2021）。重大傷病專區。衛生福利部中央健康保險署。（擷取於2024年5月25日） https://www.nhi.gov.tw/Content_List.aspx?n=C88841E119AC7392&topn=5FE8C9FEAE863B46
衛生福利部醫事司（2019a）。108年度醫院緊急醫療能力分級評定基準及評分說明與評量方法。
衛生福利部醫事司（2019b）。公告修正急診五級檢傷分類基準。（擷取於2024年5月17日）https://dep.mohw.gov.tw/DOMA/lp-5037-106.html
內政部消防署（2020）。緊急醫療救護智能平臺-救急救難一站通推動計畫4 年中程計畫。（擷取於2024年5月28日）https://www.nfa.gov.tw/cht/index.php?act=download&ids=12482
陳瑞杰（2012）。醫院緊急醫療能力分級─重大外傷病人照護。醫療品質雜誌，6(4)，19-23。
鍾亢（2017）。落實急診五級檢傷的重要性。長庚醫訊，38(8)，237-239。
鍾侑庭（2022）。台灣推動到院前五級檢傷的過去、現在與未來展望-建構與急診一致的到院前檢傷分類系統。台灣急診醫學通訊Taiwan Emergency Medicine Bulletin，5(3)。https://tsem.azurewebsites.net/EJournal/Detail/401
鍾侑庭、李彬州、洪士強、鄭宏熙、蕭雅文、林晏任、劉健佑、呂育碩、陳昱峰、李宗翰、林志豪（2024）。建構與急診一致的到院前五級檢傷制度：案例深度解析、預期成效與推動建議。消防月刊，2024年2月號，30-41。
顏淑婷、黃詩芳、陳仁隆、王志榮、沈延盛、應純哲（2015）。外傷嚴重度和住院天數相關性之探討。醫務管理期刊，16(4)，315-330。doi:10.6174/JHM2015.16(4).315

英文文獻
Bakidou, A., Caragounis, E.-C., Hagiwara, M. A., Jonsson, A., Sjöqvist, B. A., & Candefjord, S. (2023). On Scene Injury Severity Prediction (OSISP) model for trauma developed using the Swedish Trauma Registry. BMC Medical Informatics and Decision Making, 23(206). https://doi.org/10.1186/s12911-023-02290-5
Baker, S. P., O'Neill, B., Haddon, W., & Long, W. B. (1974a). The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. Journal of Trauma, 14(3), 187-196.
Bouzat, P., Legrand, R., Gillois, P., Ageron, F.-X., Brun, J., Savary, D., Champly, F., Albaladejo, P., & Payen, J.-F. (2016). Prediction of intra-hospital mortality after severe trauma: Which pre-hospital score is the most accurate? Injury, 47(1), 14-18. https://doi.org/10.1016/j.injury.2015.10.035
Boyd, D. R., Rappaport, D. M., Marbarger, J. P., Baker, R. J., & Nyhus, L. M. (1971). Computerized trauma registry a new method for categorizing physical injuries. Aerospace Medicine and Human Performance, 42(6), 607-615.
Bull, J.P. (1975). The injury severity score of road traffic casualties in relation to mortality, time of death, hospital treatment time and disability. Accident Analysis & Prevention, 7(4), 249-255.
Carter, A. J. E., Davis, K. A., Evans, L. V., & Cone, D. C. (2009). Information loss in emergency medical services handover of trauma patients. Prehospital Emergency Care, 13(3), 280-285. https://doi.org/10.1080/10903120802706260
Cassignol, A., Markarian, T., Cotte, J., Marmin, J., Nguyen, C., Cardinale, M., Pauly, V., Kerbaul, F., Meaudre, E., & Bobbia, X. (2019). Evaluation and comparison of different prehospital triage scores of trauma patients on in-hospital mortality. Prehospital Emergency Care, 23(4), 543-550. https://doi.org/10.1080/10903127.2018.1549627
Chen, J., & Li, H. (2021). Flow diagram of gradient boosting machine learning method. ResearchGate. Retrieved from https://www.researchgate.net/figure/Flow-diagram-of-gradient-boosting-machine-learning-method-The-ensemble-classifiers_fig1_351542039
Chen, T., & Guestrin, C. (2016). A general architecture of XGBoost. ResearchGate. Retrieved from https://www.researchgate.net/figure/A-general-architecture-of-XGBoost_fig3_335483097
Copes, W. S., Champion, H. R., Sacco, W. J., Lawnick, M. M., Keast, S. L., & Bain, L. W. (1988). The injury severity score revisited. The Journal of Trauma and Acute Care Surgery, 28(1), 69-77.
Das, A. (2018). Designing your neural networks. Towards Data Science. Retrieved from https://towardsdatascience.com/designing-your-neural-networks-a5e4617027ed
Daugherty, J., Waltzman, D., Sarmiento, K., & Xu, L. (2019). Traumatic brain injury–related deaths by race/ethnicity, sex, intent, and mechanism of injury — United States, 2000–2017. Morbidity and Mortality Weekly Report, 68(46), 1050-1056. US Department of Health and Human Services/Centers for Disease Control and Prevention. https://www.cdc.gov/nchs/products/databriefs/db328.htm
Dehouche, N. (2022). The injury severity score: An operations perspective. BMC Medical Research Methodology, 22(48). https://doi.org/10.1186/s12874-022-01528-6
Ditshuizen, J. C. V., Sewalt, C. A., Palmer, C. S., Van Lieshout, E. M. M., Verhofstad, M. H. J., & Den Hartog, D. (2021). The definition of major trauma using different revisions of the Abbreviated Injury Scale. Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine, 29(71). https://doi.org/10.1186/s13049-021-00873-7
Gianola, S., Castellini, G., Biffi, A., Porcu, G., Fabbri, A., Ruggieri, M. P., Stocchetti, N., Napoletano, A., Coclite, D., D’Angelo, D., Fauci, A. J., Iacorossi, L., Latina, R., Salomone, K., Gupta, S., Iannone, P., & Chiara, O. (2021). Accuracy of pre-hospital triage tools for major trauma: A systematic review with meta-analysis and net clinical benefit. World Journal of Emergency Surgery, 16(31). https://doi.org/10.1186/s13017-021-00372-1
Granieri, S. S., Reitano, E. E., Bindi, F. F., Renzi, F. F., Sammartano, F. F., Cimbanassi, S. S., Gupta, S. S., & Chiara, O. O. (2020). Motorcycle-related trauma: Effects of age and site of injuries on mortality. A single-center, retrospective study. World Journal of Emergency Surgery, 15(18). https://doi.org/10.1186/s13017-020-00297-1
Hassanzadeh, R., Farhadian, M., & Rafieemehr, H. (2023). Hospital mortality prediction in trauma patients: Comparing different SMOTE-based machine learning algorithms. BMC Medical Research Methodology, 23(101). https://doi.org/10.1186/s12874-023-01920-w
Hong, K. J., Song, K. J., Shin, S. D., Ro, Y. S., Park, J. H., Lee, S. C., & Kim, C. H. (2020). Development of a modified trauma and injury severity score to predict disability in acute trauma patients. Clinical and Experimental Emergency Medicine, 7(4), 281-289. https://doi.org/10.15441/ceem.19.097
Huang, L., Shea, A. L., Qian, H., Masurkar, A., Deng, H., Liu, D. (2019). Patient clustering improves efficiency of federated machine learning to predict mortality and hospital stay time using distributed electronic medical records. Journal of Biomedical Informatics, 99, 103291. https://doi.org/10.1016/j.jbi.2019.103291
Hung, K. K. C., Kifley, A., Brown, K., Jagnoor, J., Craig, A., Gabbe, B., Derrett, S., Dinh, M., Gopinath, B., Cameron, I. D., & FISH Investigators. (2021). Impacts of injury severity on long-term outcomes following motor vehicle crashes. BMC Public Health, 21(602). https://doi.org/10.1186/s12889-021-10638-7
James, M. K., Clarke, L. A., Simpson, R. M., Noto, A. J., Sclair, J. R., Doughlin, G. K., & Lee, S.-W. (2019). Accuracy of pre-hospital trauma notification calls. American Journal of Emergency Medicine, 37(4), 620-626. https://doi.org/10.1016/j.ajem.2018.06.058
Kang, B. H., Jung, K., Kim, S., Youn, S. H., Song, S. Y., Huh, Y., & Chang, H.-J. (2022). Accuracy and influencing factors of the Field Triage Decision Scheme for adult trauma patients at a level-1 trauma center in Korea. BMC Emergency Medicine, 22(101). https://doi.org/10.1186/s12873-022-00637-1
Kang, I. H., Lee, K. H., Youk, H., Lee, J. I., Lee, H. Y., & Bae, K. S. (2019). Trauma and Injury Severity Score modification for predicting survival of trauma in one regional emergency medical center in Korea: Construction of Trauma and Injury Severity Score coefficient model. Hong Kong Journal of Emergency Medicine, 26(4), 225-232. https://doi.org/10.1177/1024907918799910
Kuo, S. C. H., Kuo, P.-J., Chen, Y.-C., Chien, P.-C., Hsieh, H.-Y., & Hsieh, C.-H. (2017). Comparison of the new Exponential Injury Severity Score with the Injury Severity Score and the New Injury Severity Score in trauma patients: A cross-sectional study. PLoS ONE, 12(11), e0187871. https://doi.org/10.1371/journal.pone.0187871
Langlois, J. A., Rutland-Brown, W., & Wald, M. M. (2006). The epidemiology and impact of traumatic brain injury: A brief overview. Journal of Head Trauma Rehabilitation, 21(5), 375–378. https://doi.org/10.1097/00001199-200609000-00001
Lee, K.-C., Hsu, C.-C., Lin, T.-C., Chiang, H.-F., Horng, G.-J., & Chen, K.-T. (2022). Prediction of prognosis in patients with trauma by using machine learning. Medicina, 58(10), 1379. https://doi.org/10.3390/medicina58101379
Lin, G., Becker, A., & Lynn, M. (2012). Do pre-hospital trauma alert criteria predict the severity of injury and a need for an emergent surgical intervention? Injury, 43(8), 1381-1385. https://doi.org/10.1016/j.injury.2010.11.014
Liu, N. T., & Salinas, J. (2017). Machine learning for predicting outcomes in trauma. Shock, 48(5), 504-510. https://doi.org/10.1097/SHK.0000000000000898
McCoy, C. E., Menchine, M., Sampson, S., Anderson, C., & Kahn, C. (2012). Emergency medical services out-of-hospital scene and transport times and their association with mortality in trauma patients presenting to an urban level I trauma center. Annals of Emergency Medicine, 61(2), 167-174. https://doi.org/10.1016/j.annemergmed.2012.08.026
Mell, H. K., Mumma, S. N., Hiestand, B., Carr, B. G., Holland, T., & Stopyra, J. (2017). Emergency medical services response times in rural, suburban, and urban areas. JAMA Surgery. https://doi.org/10.1001/jamasurg.2017.2230
Newgard, C. D., Holmes, J. F., Haukoos, J. S., Bulger, E. M., Staudenmayer, K., Wittwer, L., Stecker, E., Dai, M., & Hsia, R. Y. (2016). Improving early identification of the high-risk elderly trauma patient by emergency medical services. Injury, 47(1), 19-25. https://doi.org/10.1016/j.injury.2015.09.010
Newgard, C. D., Staudenmayer, K., Hsia, R. Y., Mann, N. C., Bulger, E. M., Holmes, J. F., Fleischman, R., Gorman, K., Haukoos, J., & McConnell, K. J. (2013). The cost of overtriage: More than one-third of low-risk injured patients were taken to major trauma centers. Health Affairs, 32(9), 1591-1599. https://doi.org/10.1377/hlthaff.2012.1142
Nwomeh, B. C., Lowell, W., Kable, R., Haley, K., & Ameh, E. A. (2006). History and development of trauma registry: Lessons from developed to developing countries. World Journal of Emergency Surgery, 1(32). https://doi.org/10.1186/1749-7922-1-32
Palmer, C. S., Gabbe, B. J., & Cameron, P. A. (2016). Defining major trauma using the 2008 Abbreviated Injury Scale. Injury, 47(1), 109-115. https://doi.org/10.1016/j.injury.2015.07.003
Pinciroli, R., Pizzilli, G., Vassena, E., Checchi, S., Ghinaglia, M., & Bassi, G. (2019). Prehospital care and in-hospital initial trauma management. In P. Aseni, S. C. Sammartano, & S. C. Bonicalzi (Eds.), Operative Techniques and Recent Advances in Acute Care and Emergency Surgery (pp. 111-126). Springer. https://doi.org/10.1007/978-3-319-95114-0_8
Raita, Y., Goto, T., Faridi, M. K., Brown, D. F. M., Camargo, C. A. Jr., & Hasegawa, K. (2019). Emergency department triage prediction of clinical outcomes using machine learning models. Critical Care, 23(64). https://doi.org/10.1186/s13054-019-2351-7
Rapsang, A. G., & Shyam, D. C. (2013). Scoring systems of severity in patients with multiple trauma. Cirugía Española, 93(4), 213-221. https://doi.org/10.1016/j.cireng.2013.12.031
Rau, C.-S., Wu, S.-C., Chuang, J.-F., Huang, C.-Y., Liu, H.-T., Chien, P.-C., & Hsieh, C.-H. (2019). Machine learning models of survival prediction in trauma patients. Journal of Clinical Medicine, 8(6), 799. https://doi.org/10.3390/jcm8060799
Staudenmayer, K., Weiser, T. G., Maggio, P. M., Spain, D. A., & Hsia, R. Y. (2016). Trauma center care is associated with reduced readmissions after injury. Journal of Trauma and Acute Care Surgery, 80(3), 412-418. https://doi.org/10.1097/TA.0000000000000956
Svensson, A., Nilsson, B., Lantz, E., Bremer, A., Årestedt, K., & Israelsson, J. (2024). Response times in rural areas for emergency medical services, fire and rescue services, and voluntary first responders during out-of-hospital cardiac arrests. Resuscitation Plus, 17, 100548. https://doi.org/10.1016/j.resplu.2023.100548
Van Dillen, C. M., Tice, M. R., Patel, A. D., Meurer, D. A., Tyndall, J. A., Elie, M. C., & Shuster, J. J. (2016). Trauma simulation training increases confidence levels in prehospital personnel performing life-saving interventions in trauma patients. Emergency Medicine International, 2016, Article ID 5437490. https://doi.org/10.1155/2016/5437490
van Rein, E. A. J., Sadiqi, S., Lansink, K. W. W., Lichtveld, R. A., van Vliet, R., Oner, F. C., Leenen, L. P. H., & van Heijl, M. (2020). The role of emergency medical service providers in the decision-making process of prehospital trauma triage. European Journal of Trauma and Emergency Surgery, 46(1), 131-146. https://doi.org/10.1007/s00068-018-1006-8
Voskens, F. J., van Rein, E. A. J., van der Sluijs, R., Houwert, R. M., Lichtveld, R. A., Verleisdonk, E. J., Segers, M., van Olden, G., Dijkgraaf, M., Leenen, L. P. H., & van Heijl, M. (2018). Accuracy of prehospital triage in selecting severely injured trauma patients. JAMA Surgery, 153(4), 322-327. https://doi.org/10.1001/jamasurg.2017.4472
Wandling, M. W., & Cotton, B. A. (2020). Prehospital care is critical to improving outcomes after major trauma. British Journal of Surgery, 107(4), 329-331. https://doi.org/10.1002/bjs.11589
Wang, C.-J., Yen, S.-T., Huang, S.-F., Hsu, S.-C., Ying, J. C., & Shan, Y.-S. (2017). Effectiveness of trauma team on medical resource utilization and quality of care for patients with major trauma. BMC Health Services Research, 17(505). https://doi.org/10.1186/s12913-017-2429-3
Williams, T. A., Tohira, H., Finn, J., Perkins, G. D., & Ho, K. M. (2016). The ability of early warning scores (EWS) to detect critical illness in the prehospital setting: A systematic review. Resuscitation, 102, 35-43. https://doi.org/10.1016/j.resuscitation.2016.02.011
Zhang, T., Nikouline, A., Lightfoot, D., & Nolan, B. (2022). Machine learning in the prediction of trauma outcomes: A systematic review. Annals of Emergency Medicine, 80(5), 440-455. https://doi.org/10.1016/j.annemergmed.2022.05.011
Zhang, Z., Su, X., & Wu, X. (2019). Architecture of the Random Forest algorithm. ResearchGate. Retrieved from https://www.researchgate.net/figure/Architecture-of-the-Random-Forest-algorithm_fig1_337407116