臺灣博碩士論文加值系統

臺灣自1995年三月開辦全民健保以來，於近二十年間隨著全民健保普及化、民眾就醫習慣改變、與急診本身的24小時營運及快速服務等特性，急診的醫療量呈現大幅成長。急診已成為人民重要就醫第一道窗口，而由於急診所遇疾病常具多樣性、急迫性、與潛在危險等特性，且急診醫師常需在短時間內即對病情做下判斷與處置，故急診品質好壞攸關全民健康福祉。針對急診醫療現有諸多監控機制，而從醫院各項急診相關評鑑，到各種急診醫療品質指標，總有著「急診72小時返診比率」這一項目。且72小時內返診之病患常易有併發症，會增加民眾與醫院後續醫療成本。
現今資訊系統是醫院醫療運作基本架構，其已累積了龐大有用資料於儲存裝置中，但醫療人員常坐擁寶山卻不知如何開發運用。本研究先行建立資料倉儲(Data Warehouse)，再執行線上分析處理(Online Analytical Processing ,OLAP)。利用OLAP多層次操作功能，在各資料維度中運用向上整合、向下分析、橫向比較等方法而獲得許多重要結論。
本研究獲得結果:病患返診後檢傷級數一級與二級之重症病患比率增加；而五級輕症病患比率亦有增加。返診後住院率為30.63%遠高於一般急診病患住院率。返診前檢傷一級病患，返診後有最高住院率(80.95%)。返診病患前次離急診動態為「自動出院」者，有高住院率，較嚴重者有死亡個案。返診後前三大診斷依次為：腹痛、發燒、非傳染性胃腸炎及大腸炎，但醫師間有個別差異。返診診斷為發燒者有高住院率，逾五成患者須進一步醫療處置。同時有急專、內專訓練者返診率較平均返診率低。結論與建議：返診病患病情變化呈現輕重症二端比率皆增加，當設法降低重症返診與安排輕症非急診醫療管道。原檢傷一級病患與自動出院病患返診後病情最為嚴重，對此類病患離院時宜作更多評估與處置。每位急診醫師常見返診的診斷略有差異，經由大數據分析可得醫師個人化資料，作為改善與監測工具。



關鍵字：急診、72小時返診、線上分析處理

Since the Taiwan national health insurance system was established starting 1995, the population of the emergency department has increased tremendously in the recent 20 years due to the following: the popularization of the national health insurance system, the change of health seeking customs and the 24 hour fast service offered by the emergency department. The emergency department has become the public’s first priority medical choice, together with the variety, promptness and potential danger that the emergency department might come upon, the quality of the emergency department crew’s medical judgment and treatment is crucial for assuring public health. Therefore, various emergency department quality control systems, such as emergency department related evaluations and emergency department medication criteria include “72-hour Emergency Department Return Rate” unexceptionally. Moreover, 72-hour emergency department revisiting patients usually return with complications, which is likely to raise further public and hospital medical costs.
Nowadays, ‘the data system’ used to store enormous patient information has become the basic structure for hospital operation in most hospitals, however efficient application still remains unexplored. This research will start with the establishment of ‘Data Warehouse’, following the execution of ‘Online Analytical Processing’, then integrate, analyze and compare data by manipulating the multi dimensional data function of OLAP in order to pursue prominent conclusions.
The results are as follows: the ratio of critical medical condition patients among triage level 1 and 2 revisiting patients is raised. In addition, the ratio of level 5 non-urgent patients is also raised. Revisiting patients’ admission rate reached 30.63%, much higher than that of the emergency department. The admission rate of previous triage level 1 revisiting patients is the highest (80.95%). Revisiting patients who discharged against medical advice (AMA) during the previous visit show high admission rate, even more severe, death cases. The top three priority of revisiting diagnosis are: abdominal pain, fever and non-transmissible gastroenteritis, yet diagnosis between medical crew show individual difference. Revisiting diagnosis with fever show high admission rate, up to 50% patients require further health treatment. The rate of revisiting patients treated by the emergency doctor with both emergency and internal medicine training is lower than average.
Conclusions and suggestions of this research are as follows: the rates of revisiting patient’s condition among critical medical condition and non-urgent patients are raised at both extremes, thus decreasing critical medical condition revisiting patients and optional medical assistance arrangement for non-urgent patients should be made. Triage level 1 and AMA discharge patients suffer most severe condition, so more evaluation and management should be executed prior to the patient’s discharge. Common diagnosis among revisiting patients directed by each emergency doctor show difference. However, individual information can be sought via Big Data analysis, serving simultaneously as an improvement and monitoring tool.




Keywords: Emergency Department, 72-hour revisit, Online Analytical Processing

1. 緒論 1
1.1研究背景與動機 1
1.2研究目的 2
1.3研究問題與假設 3
1.4研究架構 4
1.5研究範圍與限制 5
2. 文獻探討 7
2.1急診醫療回顧 7
2.2急診檢傷分類 7
2.3急診醫療品質 8
2.4資料探勘 11
2.5資料倉儲(DATA WAREHOUSE)與線上分析處理(OLAP) 13
3. 研究方法 15
3.1研究資料與欄位整理 15
3.2研究資料分析之流程 18
3.3系統架構 23
3.4研究工具 25
4. 研究結果與分析 26
4.1病患性別比率 26
4.2檢傷級數分佈 27
4.3病患留滯時間 29
4.4各檢傷級數病患於返診後檢傷級數之變化 31
4.5急診各科別與返診比率 35
4.6急診各班別返診率 38
4.7各星期返診率 40
4.8時段與返診率 42
4.9疾病分類與返診率 44
4.10疾病與返診率 45
4.11醫師與返診率 56
4.12專科證照與返診率 65
4.13返診前次就診每位醫師平均留滯病患時間 67
4.14返診前次就診每位醫師平均留滯病患時間與返診率 67
4.15返診病患離急診狀況 68
4.16各科十大返診疾病 75
4.17各檢傷級數病患返診後動向 78
4.18疾病與病患返診後動向 84
4.19急診五大返診疾病返診後動向 89
5. 結論與建議 92
5.1結論 92
5.2建議 99
5.3未來研究方向 101
參考文獻 103
附錄 106
作者簡介 114

1. 周美儀、林淑美、黃莉蓉、趙子傑（2012）。2011年 [醫院評鑑持續性監測制度] 試辦作業成果分享。醫療品質雜誌，6(2)，P58 - 62。
2. 邱曉彥、 陳麗琴、 林琇珠、桑潁潁、康巧娟、 邱艷芬（2008）。台灣急診檢傷新趨勢-五級檢傷分類系統。護理雜誌，55(3)，P.87-91。
3. 翁慈宗(2009)。資料探勘的發展與挑戰。科學發展期刊，(442)，
P.34-37.
4. 袁梅宇（2015）。王者歸來WEKA機器學習與大數據聖經。佳魁資訊。
5. 郝沛毅、李御璽、黃嘉彥(譯) （2014）。資料探勘（原著者Jiawei Han,Micheline Kamber,Jian Pei）台北：高立圖書（原著出版年：2012）
6. 張齡芝（2006）。北台灣醫院急診於白天, 傍晚及夜間醫療利用之分
析。元智大學資訊管理學系碩士論文
7. 莊岳泉（2014）。使用線上分析處理與資料探勘技術分析癌症病患之就醫與治療－以屏東某區域教學醫院為例。國立屏東科技大學高階經營管理碩士在職專班，碩士論文。
8. 陳淑如、盧美秀、陳靜敏、沈宗奇（2003）。醫院評鑑制度的探討。台灣醫界， 46(11)。
9. 曾勁順 （2005）。整合醫療資料之物件關聯式資料倉儲架構。義守大學資訊管理研究所碩士論文。
10. 黃宗賢（2005）。急診病患非計劃性返診之相關因素探討。長庚大學醫務管理研究所，碩士論文。
11. 葉時烊、胡百敏、廖浩欽、林作彥、王少谷、廖訓禎(2010)。急診病人三日內重返急診之分析: 全民健康保險研究資料庫之抽樣歸人檔資料分析。台灣急診專科醫師期刊，2(1), P.6-13。
12. 廖熏香（2011）。指標系統再進化-臺灣臨床成效指標介紹。醫療品質
雜誌，5(3)，P. 26-29。
13. 劉彥芳（2011）。探討急診病人非計畫性返診之重要影響因素。中國醫藥大學醫務管理學系碩士論文。
14. 戴國豪、郭義、陳盈岑、吳義男（2010），資料倉儲與線上分析處理技術之應用- 以員工健檢資料分析為例。醫檢會報，25(2)。
15. Yen, Z. S., Shih, F. Y., Tsai, K. C., Chen, S. C., & Chen, W. J. (2003). 回
診的急診病患: 分析危險因子和回診原因。中華民國急救加護醫學會雜誌。14(3)，P. 93-98。
16. Duseja, R., Bardach, N. S., Lin, G. A., Yazdany, J., Dean, M. L., Clay, T.
H., ... & Dudley, R. A. (2015). Revisit rates and associated costs after an emergency department encounter: a multistate analysis.Annals of internal medicine,162(11), 750-756.
17. Gabayan, G. Z., Asch, S. M., Hsia, R. Y., Zingmond, D., Liang, L. J., Han, W., ... & Sun, B. C. (2013). Factors associated with short-term bounce-back admissions after emergency department discharge. Annals of emergency medicine,62(2), 136-144.
18. Jiménez, J. G., Murray, M. J., Beveridge, R., Pons, J. P., Cortés, E. A.,
Garrigós, J. B. F., & Ferré, M. B. (2003). Implementation of the Canadian Emergency Department Triage and Acuity Scale (CTAS) in the Principality of Andorra: Can triage parameters serve as emergency department quality indicators? Canadian Journal of Emergency Medicine,5(05), 315-322.
19. Palaniappan, S., & Ling, C. (2008). Clinical decision support using OLAP with data mining. International Journal of Computer Science and Network Security, 8(9), 290-296.
20. Rajakumar, E., & Raja, R. (2015). An overview of data warehousing and OLAP technology. Advances in Natural and Applied Sciences, 9(6 SE), 288-297.
21. 急診人次統計。衛生福利部(2011)。取自
http://www.mohw.gov.tw/cht/DOS/Statistic.aspx?f_list_no=312&fod_list_no=2536(瀏覽日期2016/02/01)
22. 衛生福利部電子病歷推動專。衛生福利部(2014)。取自
http://emr.mohw.gov.tw/introduction.aspx (瀏覽日期2016/02/01)
23. 「如何改善急診壅塞、重症病人滯留無法於6小時內轉入加護病房、
各醫學中心住院床位管控調度制度、國內急重症醫護人力短缺及關床現況」。衛生福利部（2015）。取自http://www.mohw.gov.tw/MOHW_Upload/doc/1040608-%E8%A1%9B%E7%92%B0-%E6%80%A5%E8%A8%BA(%E8%A1%9B%E7%A6%8F%E9%83%A8)_0049525004.odt(瀏覽日期2016/03/01)