透過您的圖書館登入
IP:18.188.40.207
  • 學位論文

運用決策樹技術探索住院護理評估臨床路徑資訊化研究

Exploration of Computerized Clinical Pathway in Nursing Assessment with Decision Tree Technique

指導教授 : 劉德明

摘要


分析並探討在越來越多的護理紀錄撰寫背景下,入院護理評估與健康問題之相關性,並據此整理為護理診斷支援系統建置之建議;期望未來能將研究成果實作為護理資訊系統之自動護理建議機轉,減少護理記錄時間與人為疏失,進而提升護理照護品質與效益。 運用決策樹演算法技術,回溯、解析國內南部某醫學中心積累數年之護理評估與健康問題紀錄資料:63,808 人次之入院護理評估紀錄、護理健康問題紀錄 90,606 筆);將前述資料整理後,利用Weka 資料探勘軟體整理各類健康問題之可能相關之入院評估模樣(pattern),並將結果呈現以決策樹後,歸納出前11類健康問題及與相關性之入院護理評估內容。同時邀請護理專家,針對所選取之11類健康問題單張,審查有相關性的入院護理評估項目。最後,將兩種分析方法之結果相互比對。 本研究擷取探勘分析後之11份護理健康問題,共歸納有效入院護理評估項目計46項,平均各護理健康問題紀錄有 5 項正相關之入院護理評估項目。同步邀請三位專家(分別為護理長、資訊護理師、護理副主任),針對11份護理健康問題檢視對應之入院護理評估內容必填範圍,與資料探勘結果比對相識度;運用Excel T檢定可得知,11份護理健康問題,資料探勘出有相關欄位與三位護理專家提出之有效評估結果,分別得出的P-Value 均在.05水平上其行比例上彼此之間並無明顯不同。 根據本研究結果,以主要的11份護理健康問題表單,印證並整理出其與入院護理評估之正相關項目,這些正向關係資料,當可為護理診斷支援建議系統的引導規則;透過這些建議規則,可以協助在入院護理評估之後,提出適切的健康問題照護建議。 另外,護理專家審查階段進行初期,對於各問題對應之入院評估內容,護理專家間意見仍有差異,這些現象可能需要透過進一步的研究以臨床護理更實用的護理評估項目內容,進而提供更精準的醫療照護。

並列摘要


This study analyzed and explored the relevance of nursing admission assessments and health problems in the context of the growing nursing records. We also figured some recommendations for the implementation of nursing diagnosis support system. The research results will be used as an automatic nursing advice system for nursing information systems in the future. It can reducer time for data entry and also the human error, thereby improving the quality and efficiency of nursing care. This study used data mining technology, collecting and analyzing the records of nursing admission assessments and health problems accumulated in a southern Taiwan medical center between 2014 and 2018 (63,808 person-time admission assessment records, nursing records of health problems 90,606 in total). We use the Weka to explore the possible evaluation patterns of various health problems, and present the results as decision trees. We summarized the top 11 health problems and the relevance of the admissions assessment. Three nursing experts are invited to review the relevant admissions assessments for the selected ten health problem forms. Finally, the results of the two analytical methods are compared to each other. In this study, 11 nursing health problems were collected after exploratory analysis. A total of 46 effective admission nursing assessment items were summarized. On average, there were 5 positive related admission nursing assessment items in each nursing health record. Three experts were invited at the same time (chief nurse, information nurse and deputy director of nursing). The experts had different opinions on the required items of admission nursing assessment for 11 nursing health problems medical forms. Among them, the chief nurse put forward an effective evaluation result which was 89.67% close to the analysis, and the information nurse and deputy director of nursing were 42.83% and 63.33% respectively. The average consensus of the three experts was 65.28% and close to the analysis. This study suggested that a further study should be carried out in view of the differences between the three experts in their examinations, which may be due to their different job responsibilities. In this study, 11 nursing health problems were obtained after the exploration and analysis. Three nursing experts (a nursing director, an information nurse and a deputy director of nursing respectively) were invited to check the required scope of the corresponding nursing assessment content for 11 nursing health problems. Then Excel T verification were used to verify the results from this study and the experts. The P values obtained were at the .05 level and showed they were not significantly different from each other. This study found that the main 11 health problem forms are used to confirm and collate the positive correlations with the nursing admissions assessment. These positive relationship data can be used as guidelines for the nursing diagnosis support system. They can assist in the development of appropriate health care advice after the admission assessment. There are still differences in the opinions of the nursing experts on the admission assessment content corresponding to each question. These phenomena may need further research, clinical care and more practical care assessment project content, in order to provide more accurate medical care.

參考文獻


侯宜菁; 游惠珠; 陳建民; 吳榮軒; 藍敏方; 黃月嬌. (2010). 護理評估資訊系統使用前後對護理紀錄時間及紀錄品質成效之探討. 源遠護理, 11-20.
周主照. (2011). 以資料探勘技術探討護理人員工作績效與教育訓練之相關性—以某醫學中心為例.
周歆凱. (2004). 利用『資料探勘技術』探討急診高資源耗用者之特性.
尤銓駿. (2007). 應用資料探勘於門診處方藥物交互作用之分析.
李政霖. (2011). 以資料探勘分析改善急診醫師排班作業-以某區域教學醫院為例.

延伸閱讀