由於科技日新月異的進步,現今的醫療資訊發展相當的完善,對於醫院各部門的電腦化、資料電子化都有相當大量的資料累積。而其中在這大量的資料中,也隱藏著相當多在醫療過程未知的資訊。透過資料探勘技術能夠從資料中找出其中隱含或不明顯的資訊,對於醫療 管理人員來說,有著相當大的幫助。 急診醫療是面對緊急病患的第一線,不論急救、留置觀察或開刀都可進行,然而當病人進入急診隨即依照病患危急程度篩選等級,之後依照科別進入各科診察室就診,但由於國人醫療習性的偏好,大醫院的急診人數總是一般的數十倍,造成急診的人力指派是否足以應付病患需求,以及急診資源分配不均,影響病患生命以及醫療品質,因此醫師排班人力預測確實影響醫療品質與醫療成本評估,進而對病患 生命、滿意度產生重要性的影響。 本研究採資料探勘的分類方法,取得急診的病患預測需求模組,結合六標準差管理之DMAIC推動手法,提出一套績效矩陣,來評估資料探勘後的需求模組預測模式與原排班模式比較其排班的品質績效,最後提出一套適用於不同科別之醫療人力派遣之預測模式。 在決策樹最後的輸出變數是採用,依照現有病患人數,應需要醫師人數,而一位醫師一天之中應服務多少個檢傷病患。而病患劃分,本研究個案的排班規則中,按內科總就診人數11,509人次來算,一位醫師一天之中需服務15.76人次的病患。如果將條例當成最佳診療滿意度時,已達最佳服務水準的95.5%。
Thanks to the development of information technology, the daily routines in hospitals have been computerized and the accumulation of all kinds of medical electronic records is in an unprecedented pace. It would be extremely useful to the hospital management, if the collected information can be analyzed. The data-mining technology is one of the approaches that can be applied to identify important and useful messages in health care administration. The emergency department is sometimes the first place a patient make contacts with medical care. In Taiwan, under the National Health Insurance, enrollees have complete freedom in choosing providers and they have more trust on large-scale hospitals (most of them teaching hospitals). As a result, emergency departments in large hospitals are always overwhelmed by a large number of patients seeking medical treatments. This has caused an uneven distribution of medical resources and not enough manpower to care so many patients. It becomes crucial to have a efficient arrangement of on-shift physicians and nurses based on the patient flows in a emergency department. In this study, the so-called Performance Evaluation Matrix, Classification, Association Rule, and Clustering methods in data mining technology were used to sort out patients visiting an emergency department in regional teaching hospital in central Taiwan. A prediction model was developed to estimate the demands for emergency department use. All the simulated models were compared to the original physicians and nurses shifting schedule in the emergency department for their performance. Finally, a final model was selected and it can be also applied to other departments in the hospital.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。