目的:利用資料的判定係數及 Rasch 模式,探討醫院層級間的護病比。 方法:下載中央健康保險署 2015 年各層級479家醫院各月全日平均護病比資訊公開的資料,利用文獻微軟 Excel 下開發的試題反應理論 Rasch 模式的資訊模組,先挑出護病比資料的判定係數最小者;再依據符合與不符合護病比評鑑標準的醫院進行分類;最後以護病比的不穩定性指標排序。本研究較傳統統計方法能提供更多訊息,挑出資料的異常現象,做為進一步分析及檢討改善的依據。結果:研究結果顯示:1.挑出護病比資料判定係數最小的層級醫院為醫學中心層級醫院(R2=0.71);2.醫學中心層級醫院中符合護病比評鑑標準的醫院有21家,不符合護病比的醫院,有1家;3.醫學中心層級醫院中,有2家醫院的護病比呈現偏高且不穩定的現象。結論與建議:利用Rasch 模式資訊模組來比較各層級醫院全日平均護病比資料,驗證醫學中心層級醫院護病比資料的判定係數並非最高,本研究提供的資料分析方法,提供醫院管理者的參考與應用。精簡工作內容、改善護病關係與薪資福利制度是提升護理人力的重要議題,值得醫院管理者的重視。 Objectives:The purpose of this study is to investigate nurse-patient ratios among Taiwan hospitals using coefficient of determination and Rasch model. Methods:We downloaded data of nurse-patient ratios of 479 hospitals in 2015 from the website of Taiwan National Health Insurance Administration and used the Rasch continuous module developed on Microsoft Excel to screen out the minimum of determination coefficient of nurse-patient ratios among hospital levels for improvement consideration, then to separate two groups according to the criterion of hospital accreditation on nurse-patient ratios, and to rank the far-away-from stability indicators for hospitals as well as demonstrating a method compared to the traditional statistical approach to provide more meaningful information to readers for further improvement made in future. Results: We found that the first the minimal determination coefficient (R2=0.71) among hospital levels is attributable to the medical center level, the second one medical center hospital is not met to the criterion (9 patients per nurse) set by hospital accreditation regulation and other 21 hospitals totally meeting to the criterion, the third two hospitals in medical centers are shown most significantly unstable, their outfit mean square error are greater than 2.0.Conclusions: A novel module downloaded from literature was demonstrated and used for confirming that nurse-patient ratios of medical center hospitals in 2015 suffer the minimal determination coefficient beyond our expectation and the data examined by Rasch model can be addible and constructed a unidimensional scale. Reducing work loadings, Improving relationships of patients and nurses, and changes of welfare policies are important issues for hospital managers to reduce and maintain nurse-patient ratios.