本研究主要探討醫療可近性對癌症患者的存活率是否有影響。透過國家衛生院所發行的2000年及2005年的百萬人次抽樣檔，利用1996年至2013年新增之口腔癌病患做為研究對象，利用病患是否有跨區就醫與病患居住地與就醫之距離作為醫療可近性程度之衡量指標，由於全民健保資料庫中並無記錄病患居住地與就醫醫院之真實地址，因此本研究選取口腔癌病患之輕症與癌症治療就醫紀錄中最多次之地區做為病患居住地與就醫地。本文除了使用存活分析–Cox比例危險模型探討醫療可近性對病患的死亡危險率外，由於口腔癌非急症，病患可以自我選擇至醫療資源豐富的地區就醫，使就醫行為會與醫療結果產生內生性，因此本文亦利用二元常態機率模型加入工具變數控制內生性以進行實證分析。 實證結果發現病患就醫距離變數不顯著，表示距離並非造成口腔癌病患死亡主要原因，而選擇跨區（縣市）就醫的病患相對於未跨區就醫者與居住地之每十萬人口醫師數越多，其醫療結果較佳，會顯著降低死亡率，代表醫療可近性的程度確實會影響口腔癌病患之死亡率，因此本文建議政府及相關單位在規劃醫療保健政策上，應該將醫療資源分布不均納入考量，考慮各地區內醫療資源及人口分布，針對醫療資源較貧脊的地區適度調配醫事人力及設施分布，使醫療保健政策可以貼近國人的醫療需求，改善醫療資源較貧脊地區之病患就醫時的交通及成本障礙，也可減少病患在地區上造成的健康狀況差異。 ;The purpose of this paper is to investigate the relationship between the medical accessibility and the mortality of oral cancer patients. The data resource is from 1996-2013 oral cancer patients extracted from the 2000 & 2005 Longitudinal Health Insurance Database (LHID2000、LHID2005) which is distributed by the National Health Research Institutes. The medical accessibility in this paper means that whether patient with oral cavity cancer choose cross- regional to seek medical care and the distance from oral patients’ residence to the hospital. Since the National Health Insurance Database does not record the real address of the patient′s residence and the hospital, the study selected the most frequent areas of oral cancer patients′ minor ailment and cancer treatment records as the address of patient′s residence and medical treatment. First, we used survival analysis - Cox proportional hazards model to discuss the relationship between the medical accessibility and the hazard rate of oral cancer patients, but oral cancer is not an acute disease, patients can choose where to seek the medical care. The decision of cross-regional would be endogenous to the treatment outcome. Hence, bivariate probit model estimation was adopted to take into account the endogeneity problem and add instrumental variables to control endogeneity. The results of the experiment indicated that there is no significant association between the travel distance and mortality. However, we documented a statistically significant negative association between mortality, crossing region and the number of doctors per 100,000 population. We can make a conclusion that the degree of medical accessibility would indeed affect the mortality rate of oral cancer patients. This raises a policy implication of allocating the distribution of medical resources to the poor areas after considering the uneven distribution of medical resources.
