本研究目的主要是在探討當民眾面臨攸關生命之重大考量時，跨區就醫之行為是否有更好的醫療結果。本文資料來源是由國家衛生院所發行的「2000 ?百萬人次抽樣檔」，由1998年至2009年新增之肝癌病患做為研究樣本，利用病患居住地與就醫地不同且其距離超過20公里認定病患有進行跨區就醫，但健保資料庫中並未記錄病患居住地之資料，因此我們將新增肝癌病患其前一年的輕症就醫記錄中，選取最多次就診紀錄之地區，做為病患之居住地；其次，若病患確診前一年內並未有輕微疾病之就醫記錄，則我們將該病患之投保地做為病患之居住地。因病患的疾病嚴重程度會影響病患是否會選擇跨區就醫以及醫治結果，使得病患會自我選擇去醫療資源較豐富的醫院就醫，而病患自我選擇的行為會使得跨區變數產生內生性問題，因此我們將藉由工具變數控制跨區變數的內生性，並利用兩階段最小平方法解決內生性所導致的估計偏誤，以利於比較跨區就醫行為之醫療結果。 實證結果發現相較於未跨區就醫者，病患選擇跨區就醫的醫療結果較佳，死亡率顯著較低，可得知病患居住地醫療資源較貧乏者，可藉由跨區就醫獲得更好的醫療結果，表示台灣醫療分配不均造成病患在當地無法獲得較好的醫療照顧因而選擇跨區就醫。 The purpose of this thesis is to investigate whether patients with liver cancer could have better treatment outcome if they seek care in areas far from their residence. The data resource is 1998-2009 liver cancer patients extracted from the 2000 Longitudinal Health Insurance Database (LHID2000) which is distributed by the f National Health Research Institutes. Patients are recognized as crossing area to seek care if the distances between patients’ residences and hospitals are over 20 kilometers and their areas are different. However, there is no information of patient’s residence available in the National Health Insurance Database. As such, we adopt following procedures to identify patient’s crossing-area-seeking-care behavior: First, we identify a patient’s residence as the area where the patient had several records of seeking care with minor diseases one year before cancer diagnosed. Second, if the patient did not have records of care for minor diseases, we use the area where the patient enroll in NHI as the patient’s residence. The decision of cross-area-seeking-care would be endogenous to the treatment outcome because the unobserved factors such as severity may affect patient’s treatment outcome and also motivate the patient to seek care in areas with more abundant medical resources. Hence, two stage least square (2SLS) estimation was adopted to take into account the endogeneity problem. Our empirical results show that mortality rate of the patients who used medical care by crossing region is significantly lower than those who didn’t. We can make a conclusion that patients who live in the area with poor medical resources are able to obtain better medical outcomes by crossing areas to seek better medical diagnosis and treatment. This raises a policy implication of equalizing medical resources across areas to secure the quality of care for each patient no matter where he/she lives.
