模糊績效評估模式－以住院醫師為例

全民健保實施以來，使得台灣醫療市場產生激烈的質變，尤其在總額給付制度施行之後，造成醫院在內部的管理及外部的應變措施產生重大的改變。住院醫師除了是醫療品質與資源有效利用的主要掌控者外，亦是醫療團隊中臨床照顧活動施行的主要決定者，其醫療表現之優劣對病患的就診滿意度與醫院的聲譽、永續經營有著深層的影響力，如何多元、公平地評估其醫療表現即成為極重要且極富挑戰的課題。然而，在績效評估的過程中存在著許多的難題，其中包括管理者對住院醫師的醫療表現常常僅存模糊印象、所使用的準則彼此不全然獨立、無法同時兼顧個別醫療機構的組織文化差異及績效評定的客觀性等。本研究的主要目的即在於建構「模糊績效評估模式」，以克服上述難題並使評估結果更真實反應住院醫師的醫療表現。本評估模式之優勢包括：(1)運用大量問卷篩選出含有4個構面及20個準則的績效評估架構；(2)利用語意變數表達無法以明確分數表示的住院醫師醫療表現；(3)引用模糊測度測量準則之間的相關程度與模糊積分克服準則之間的不獨立特性，給予受評住院醫師適當的評判；以及(4)在兼顧個別組織文化差異之原則下，利用貝氏定理整合專家權重與主管權重而得到合宜的評估準則權重－「貝氏權重」，以增加評估結果的真實性。在實務驗證方面，本研究以署立M區域醫院三位主管評量其所屬四位住院醫師的醫療表現為例，就演算觀念、受評估者觀感和績效檢視能力等三個面向，將本研究所提新式模糊績效評估模式與傳統演算法和模糊期望值演算法做優劣之比較，並且印證此模式在評估績效表現上的優勢能力。

關鍵字

績效評估；貝氏權重；模糊積分；住院醫師

並列摘要

The implementation of global budget payment system of Taiwan's national health insurance program made the medical organizations putting most of managerial efforts on cost containment and quality assurance. In turn, interns who control 80% of the resource utilization decisions should have most of the responsibility for quality assurance and cost reduction. Therefore, the conduction of how to equitably evaluate intern performance is especially meaningful. However, there exist several difficulties, including the ambiguous identification on intern performances, the interrelationship existing between the criteria used, the embarrassment of not care concurrently evaluation objectivity and special considerations of individual medical organization. The objective of this study aims at constructing a Bayes' weighting fuzzy intern performance evaluation model to overcome these difficulties. This study makes several contributions to this field. First, following a series of screening procedures, a structure with 4 criteria and 20 subcriteria has been objectively constructed. Second, this study uses the fuzzy multiple criteria decision making to evaluate the intern performance, in which fuzzy linguistic settings, fuzzy measure and fuzzy integral are utilized to identify intern performance and overcome criterion interrelation, respectively. Third, we employ Bayes' theorem to obtain the Bayes' weightings of 4 criteria by integrating the weightings that come respectively from subjective appraisal and objective appraisal. By that, we propose a new algorithm, Bayes' weighting fuzzy integral method, to effectively rank intern performances. In empirical verification, this study focuses on the performance evaluation case of 4 interns by 3 senior managers in a local hospital. The excellent capability of the new algorithm on distinguishing intern performances is proved from several viewpoints.