貝氏網路之醫療品質指標網之研究-以急性照護指標為例

Tunghai University Institutional Repository > 工學院 > 工業工程與經營資訊學系所 > 碩博士論文 > Item 310901/5124

Please use this identifier to cite or link to this item: http://140.128.103.80:8080/handle/310901/5124

Title:	貝氏網路之醫療品質指標網之研究-以急性照護指標為例
Other Titles:	A BN-Based Healthcare Quality Indicator Network －An Application in Acute Care Indicators
Authors:	楊仕莙 Yang, Shih-Chun
Contributors:	王文清;王偉華 Wang, Wen-Ching;Wang, Wei-Hua 東海大學工業工程與經營資訊學系
Keywords:	醫療品質;台灣醫療品質指標改善計畫;貝氏網路 Healthcare Quality;the Taiwan Quality Indicator Project;Bayesian Network
Date:	2005
Issue Date:	2011-05-19T07:46:15Z (UTC)
Abstract:	近年來，無論政府、各醫院及社會大眾，醫療品質已逐漸成為一項重要關心的議題。為了加以監控及改善醫療品質，財團法人醫院評鑑暨醫療品質策進會（簡稱醫策會）引進美國馬里蘭醫院協會所主導的「醫療品質指標計畫」(Quality Indicator Project , 簡稱QIP)，以「台灣醫療品質指標改善計畫」(Taiwan Quality Indicator Project,簡稱TQIP)，在國內致力於醫療品質指標計畫之推動。目前，台灣醫療品質指標改善計畫已引進的照護指標群，計有急性照護、長期照護、精神照護及居家照護等，各群組下有不同的指標。對醫院而言，使用統計輪廓 (Statistical Profiling)來評估醫療品質是常用的方式，在統計管制下，醫院可以預測判斷流程上影響及品質造成的原因變異來檢視對品質的影響。然而，這些指標並非獨立使用卻透過衡量結果反映出不同醫療照護層面的醫療品質現況。因此，本研究由急性照護TQIP 醫療指標群以貝氏網路來分析機制來衡量指標間相關程度及不確定性，並找出TQIP指標的非線性網路架構來協助醫院監控醫療品質且是否有隱含指標間機率性關係存在，同時，協助醫院在操作TQIP指標有客觀的衡量依據。 In recent years, healthcare quality in Taiwan has become an increasingly important issue for the government, for the hospitals, and for the public. In order to survey and ultimately improve healthcare quality, the Taiwan Joint Commission on Hospital Accreditation (TJCHA) has imported, from the Maryland Hospital Association (MHA) in the United States, the Quality Indicator Project (QIP). TJCHA has re-named the Maryland QIP the Taiwan Quality Indicator Project (TQIP), and has been devoted to implementing the TQIP program in domestic hospitals.At present, the TQIP is comprised of multiple sets of performance indicators for measuring four different pattern care settings of the hospital system: Acute Care, Psychiatric Care, Long-term Care, and Home Care. Each care setting uses numerous TQIP indicators for measuring quality. Among hospitals, statistical profiling is a generally accepted method for assessing healthcare quality. Statistical profiling allows hospitals to predict certain outcomes of the healthcare process and to thus survey the various causes that affect quality performance. However, the data from these indicators cannot be used, independent of an appropriate analysis method, to determine relationships among healthcare quality settings. TQIP indicators merely provide data for various settings within the healthcare system. Therefore, this study proposes a method for analyzing the relevance and uncertainty among data gathered from TQIP indicators. In addition, we devise a feasible mechanism, involving Bayesian Networks, to solve problems of structure and parameter requirements for the TQIP indicators. We provide hospitals with an objective auxiliary-assessment for indicator-data analysis and check whether it implies certain probability relationships among these TQIP indicators.
Appears in Collections:	[工業工程與經營資訊學系所] 碩博士論文

Files in This Item:

File	Size	Format
093THU00030009-001.pdf	989Kb	Adobe PDF	198	View/Open

Loading...