本研究以台灣代表性醫院之一的急診室資料分析近幾年的急診醫療資源情 況及問題。在健保制度的實施下,民眾對醫療資源的仰賴也逐漸提高,造成近 幾年全台急診室醫療費用的增加及滯留與壅塞的情況。因此透過急診室資料, 將病患的病情以 TTAS 急診五級檢傷分類標準,並分析病患醫療診斷碼之關聯 性,進而去探討病患在急診室中的滯留情況與行為。研究主要以關聯規則探勘 診斷碼樣式,對病患的 ICD-9 診斷碼以演算法 FP-Growth 找出具有相關性的醫 療診斷碼樣式。接著建立 J48 決策樹模型,預測病患滯留時間,再加入醫療診 斷屬性和病患屬性。分組探討在不同檢傷級數下病患的滯留行為,分析決策樹 規則和醫療與病患屬性的重要程度及正確率。研究結果發現,加入醫療診斷碼 樣式在決策樹預測分析中,能更容易了解急診室病患的情況。從決策樹規則可 以發現有些病情相當分散,且大部分會滯留的原因可能屬於個人行為。與醫師 的討論分析後,可能只有在病情中度嚴重的病患是較急需要急診醫療資源的群 眾。因此現今急診室的壅塞情況,目前大部分的急診室壅塞問題可能是病患。 所以多跟民眾溝通和宣導,建立正確的醫療觀念與制度,並協助醫院了解不同 病患的情況,才能將醫療資源用在真正有需要的人身上。

The research is aim to analyze the utilization situation and problems of emergency medical resources in Taiwan during the past years by exploring the information of representative hospital emergency rooms in northern Taiwan. Under the implementation of the health insurance system, the demand for medical resources increased gradually, which has resulted in an increase of medical expenses and the extension of length of stay in the emergency rooms recent years.In this research, the information of patient's condition in emergency rooms is classified according to the TTAS emergency five-level classification standard. By analyzing the relevance of the patient's medical diagnosis code to explore the situation of the length of stay in the emergency rooms.
The research examines mainly the diagnostic code style with association rules, and finds the relevant medical diagnosis code patterns with the algorithm FP-Growth for the patient's ICD-9 diagnostic code. Therefore, establishing the J48 decision tree model to predict the length of stay, then add medical diagnostic attributes and patient attributes to discuss the retention behavior of patients under different levels of injury ,analyzing decision tree rules and the importance and accuracy of medical and patient attributes. The study indicated that adding ICD codes can make it easier to understand the patients’ situation of emergency room by means of decision tree prediction analysis.Through the decision tree rules, it’s found that some conditions are quite scattered and extension of the length of stay mostly resulting from personal behavior. After discussing with the attedning physician, it’s also found that only those patients who are moderately ill conditions have urgent need of emergency medical resources. In summary, the congestion in the emergency rooms mostly depends on patients.
In order to make sure people who really need can utilize medical resources in time, it is essential to the enhance the communication between personnel and people, strengthen the publicity for correct medical concepts. Furthermore, it is important to establish comprehensive systems and make hospitals to understand the situation of different patients well.

王偉驎、林文燦、賴政皓、陳慧敏(2008)。應用資料探勘技術提升急診醫學檢 傷分類之一致性--以臺灣某醫學中心急診醫學部為例。品質學報,15(4), 頁 283-291。
吳國禎(2000)。資料探索在醫學資料庫之應用(碩士論文)。中原大學醫學工 程學系所。取自台灣碩博士論文系統(系統編號 088CYCU0530038)。
巫俊卿(2012)。一個植基於多類別關聯規則演算法之疾病分類代碼預測系統 (未出版之碩士論文)。中興大學資訊科學與工程學系所,臺中市。
陳子立、吳怡瑾、馮嚴毅、楊昌倫 (2016)。基於急診滯留時間探勘醫療行為 之實證研究。管理與系統,23(4),頁 527-561。
陳鎮揮、楊鎮嘉、吳效文(2008)。以資料探勘技術建立 ICD-9-CM 編碼決策系 統。光田醫學雜誌,3(5)。
周歆凱、黃興進、蔡明足、翁林仲、蘇喜、陳真吟(2009)。運用資料探勘之叢 集分析技術探討急診 72 小時再返診病患特性。澄清醫護管理雜誌,5(3)。
鄭凱今(2011)。應用文字探勘技術於疾病分類碼之研究–以胃腸科病歷摘要為 例(未出版之碩士論文)。國立雲林科技大學工業工程與管理研究所,雲林 縣。
簡禎富、許嘉裕(2014)。資料挖礦與大數據分析。新北市:前程文化。
Bellazzi, R.,&Zupan, B. (2008). Predictive Data Mining in Clinical Medicine: Current Issuesand Guidelines. International Journal of Medical Informatics. 77(2), pp. 81- 97.
Bertsimas, D., Bjarnadóttir, M. V., Kane, M. A., Kryder, J. C., Pandey, R., Vempala, S., & Wang, G. (2008). Algorithmic Prediction of Health-care Costs, Operations Research, 56(6), pp.1382-1392.
Borkar, S., & Rajeswari, K. (2013). Predicting Students Academic Performance Using Education Data Mining. International Journal of Computer Science and Mobile Computing, 2(7), pp.273-279.
Delen, D., Walker, G., & Kadam, A. (2005). Predicting Breast Cancer Survivability: A Comparison of Three Data Mining Methods. Artificial Intelligence in Medicine, 34(2), pp. 113-127.
Olson,D.L., Delenb,D., & Menga,Y. (2011). Comparative Analysis of Data Mining Methods for Bankruptcy Prediction. Decision Support Systems, 52, pp. 464-473.
Farkas, R., & Szarvas, G. (2008). Automatic Construction of Rule-based ICD-9-CM Coding Systems. BMC Bioinformatics. 9(Suppl 3): S10.
Fayyad, U., Piatetsky-Shapiro, G., Smyth, P., &. Uthurusamy, R. (1996). Advances in Knowledge Discovery in Databases. AAAI Press.
Jiawei, H., Pei, Jian & Yiwen, Y. (2000). Mining Frequent Patterns without Candidate Generation: A Frequent-Pattern Tree Approach. Data Mining and Knowledge Discovery. 8(1), pp. 53-87.
Jiawei H., Jian P., & Yiwen Y. (2000). Mining Frequent Patterns Without Candidate Generation. SIGMOD '00 Proceedings of the 2000 ACM SIGMOD International Conference on Management of Data, pp.1-12. Doi:10.1145/335191.335372
100
Karhade, P. P., Shaw, M. J., & Subramanyam, R. (2015). Patterns in Information Systems Portfolio Prioritization: Evidence from Decision Tree Induction. MIS Quarterly. 39(2), pp. 413-433.
Kavuluru, R., Rios A., Lu, Y. (2015). "An Empirical Evaluation of Supervised Learning Approaches in Assigning Diagnosis Codes to Electronic Medical Records". Artificial Intelligence Medicine. 65, pp. 155-166.
Khan, M. S., Muyeba, M., & Coenen, F. (2008) “A Weighted Utility Framework for Mining Association Rules."Proceedings of European Modeling Symposium (IEEE), pp. 87-92. Doi: 10.1109/EMS.2008.73
Lin, W. T., Wu, Y. C., Zheng, J. S., & Chen, M. Y. (2011). Analysis by Data Mining in the Emergency Medicine Triage Database at a Taiwanese Regional Hospital. Expert Systems with Applications. 38(9), pp. 11078-11084.
Moro, S., Cortez, P., & Rita, P. (2014). A Data-driven Approach to Predict the Success of Bank Telemarketing, Decision Support Systems. 62, pp. 22-31.
Pereira, L., Rijo, R., Silva, C., & Agostinho, M. (2013), ICD9-Based Text Mining Approach to Children Epilepsy Classification. Procedia Technology. 9, pp. 1351- 1360.