定額支付制度下病例醫令之合適性研究; Study on the Fixed Amount Payment’s Order Items

NCU Institutional Repository > 管理學院 > 資訊管理研究所 > 博碩士論文 > Item 987654321/12862

請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/12862

題名:	定額支付制度下病例醫令之合適性研究;Study on the Fixed Amount Payment’s Order Items
作者:	蔣肇慶;Jaw-Ching Chiang
貢獻者:	資訊管理研究所
關鍵詞:	SFLI演算法;分解法;定額支付;基本醫令群;資料開採;Fixed Amount system;decomposition;SFLI algorithm;Basic Order Group;Data Mining
日期:	2004-12-27
上傳時間:	2009-09-22 15:16:31 (UTC+8)
出版者:	國立中央大學圖書館
摘要:	醫療院所的市場競爭日趨劇烈，醫院經營決策者在面臨此複雜的決策問題時，除了本身專業知識與過去的經驗外，必須應用資訊技術，以協助解決問題及輔助制訂決策之機制。自實施全民健保之後，醫療院所面臨著財務的壓力與外界的競爭力，可謂是有增無減。因此，如何提升醫院的競爭力，取得醫療市場的利基，是醫院經營決策者的任務。資料開採(Data Mining)係在豐富的資料庫中，搜尋出有價值之隱藏資料，並且加以分析，擷取有意義且有價值的資訊，或是歸納出結構化的模型，作為經營決策者決策時之參考。當資料開採已經逐漸成為企業的核心時，經營決策者要知道的是運用此技術實現在組織內，將組織的資料予以適當的分析及製作決策模式，提升在企業經營的利基。因此，在醫療市場的競爭之下，在定額支付制度下的病例，醫令執行項目內容合適性與費用的合理性問題，即非常重要。因此，我們可以藉著資料開採的技術設計出基本醫令群(Basic Order Group, BOG)，而予以解決。所以，我們提出以分解法（Decomposition）為基本構想的SFLI (standing for Suitable and Frequent Large Itemsets) 演算法尋找出雛型基本醫令群。為了更能準確地計算及驗證，因此我們再使用以自組織映射圖為主的基本醫令群演算法尋找相似基本醫令群做比較。再分別使用相對強度(Relative Strength)、趨近值(Approach Value)及值比率(Value Rate)三組的分別計算，產生出初稿型基本醫令群；接著，透過費用評估的「差異比(Different Rate)的驗證」。結果評估顯示，三組產生的基本醫令群，皆合乎成本效益。因此，三組中任一方法皆可達到我們的目標。同時，此基本醫令群非常具彈性，其一、該基本醫令群之子集合的組合亦可適用（減少醫令），其二、視病患的病情予以調整（增加醫令）。綜合言之，基本醫令群的產生，不但可以協助醫療支付單位在定額支付制度下的醫令項目研究，如精確的計算探討出住院（或門診）定額支付病例的必須執行基本診療醫令項目及門診AP-DRGs制給付基準之醫令的研究；而且在醫院方面，可以節省不必要的處置與成本支出，提升營運的績效；對醫師診療開立醫令而言，將更具成本的觀念，提升醫師之服務績效；於病患方面，可以獲得適當的診療，免除不必要的檢驗、藥品及處置，得到就醫權益的保障。因此，如何將資料開採的理論與實務導入醫療院所，是資訊管理的新議題；希望藉此研究能提供資料開採在醫療產業的新思維。即將此學術上的研究，帶給實務上的重要參考以達相輔相成之效。 Maintaining a financial balance given limited medical payments is essential for health insurance payment units and hospitals. The Bureau of National Health Insurance (BNHI) implemented the Prospective Payment System of the Global Budget System to assist hospitals in planning and controlling medical care costs and service quality. Meanwhile, the BNHI also devised various plans for strengthening the operational utilization of medical resources, such as Case Payment System. This system has two types of orders; one is basic required examinations and treatments, and the other is option. In this Case Payment System, hospitals must execute 65% basic required examinations and treatments. Under this condition, it is impossible to understand how ‘65% of basic required examinations and treatments’ are calculated? Is the ‘35% of exanimations and treatments’ unimportant? This study suggested SFLI (standing for Suitable and Frequent Large Itemsets) algorithm to solve the suitable number of necessary medical order items. The algorithm was based upon decomposition. Prototyping and Similar Basic Order Group (BOG) were generated by SFLI and SOM algorithm respectively. Acting on relative strength, approach value and value rate, draft BOG were generated using these Prototyping BOG and Similar BOG. The SFLI algorithm employed decomposition method. This makes the suitable frequent and large itemsets processed faster, and reduced CPU time of no generating candidate itemsets. Therefore, this performance was the more better of the Apriori and FP_Tree algorithm. Meanwhile, the inside of the draft BOG’s elements, which are order items, were allowed to employ a reporting fee to calculate the payments of every draft BOG. This study compared and tested statistical hypotheses between the experiment contrast payment for every draft BOG and the health insurance reporting payments of each hospital level. Through verification of different rate, draft BOG were transformed to BOG that had full of cost-benefit. Therefore, the BOG can not only assist payment units in reducing costs, but also can assist hospitals in operating efficiently using the Fixed Amount System. Consequently, this study demonstrates a standard of payment for current and future basic and suitable order items of Fixed Amount System references.
顯示於類別:	[資訊管理研究所] 博碩士論文

文件中的檔案:

檔案	大小	格式	瀏覽次數

在NCUIR中所有的資料項目都受到原著作權保護.

社群 sharing

資料載入中.....