以全民健保資料庫探討不同醫療層級的服務差異

Using National Health Insurance Database to Explore the Attributes of Different Levels of Medical Institutes

林芃彣(Lin Peng-Wen)

指導教授 ： 余清祥

國立政治大學/商學院/統計學系/碩士(2018年)

https://doi.org/10.6814/THE.NCCU.STAT.002.2019.B03

摘要

我國實施全民健康保險制度（以下簡稱健保）已逾20年，民眾納保率超過99%，達到了健保第一階段目標醫療可近性，全國所有鄉鎮市區層級都有一家以上的健保特約醫療院所，而且國人對就診、取藥等相關規定非常熟悉，健保已成為日常生活的一部分。但健保近來也面臨不少挑戰，如醫護人員過勞、健保給付範圍縮減、城鄉醫療資源差距過大等隱憂，近年政府推出的二代健保即針對這些缺失提出修正。其中分級醫療是二代健保的主要特色之一，在善用醫療資源的前提下，藉由分工合作讓國人獲得完善的醫療照顧。本文也以評估分級醫療為目標，分析醫療供給面，透過個人就醫習性等紀錄推敲醫療機構之營運特色，檢視醫療供給的現況與趨勢，並且評估其中潛在的風險與挑戰。
本研究將透過健保資料探討臺灣醫療院所的服務差異性，使用素材為2005年百萬人抽樣資料，包括就醫門診處方及治療明細檔（CD）及醫事機構資料檔（HOSB）。由於採用的是抽樣資料，我們將先檢查樣本代表性，確定資料庫的分析結果與官方（衛福部）資料一致，檢查項目包括就醫次數、就醫金額等。醫療院所的服務差異性主要在於比較四個層級的醫療院所，探討病患的人口特性、看診疾病等是否存在明顯不同，透過辛普森指標等不均度指標找出主要差異。分析發現：近年國人有傾向醫學中心及區域醫院就醫的趨勢，前三層級醫療院所的疾病不均度相似，很難區隔它們扮演的角色；而基層醫療院所的門診主要為呼吸系統疾病，且疾病不均度在都會地區的差異性最大，但近年不均度有下降的傾向。本文研究結果除了可用於評估各級醫療院所的經營績效，包括二代健保的分級醫療實施作法，也可檢視病患的就醫屬性或偏好，作為增設、調整臺灣醫療資源配置的參考。

關鍵字

全民健保 ； 醫療供給 ； 大數據 ； 不均度 ； 市場區隔

並列摘要

Taiwan’s National Health Insurance (NHI) was implemented in 1995 and now more than 99% Taiwan’s citizens participate in the NHI. Taiwan’s NHI is world-famous because of good health service, including medical accessibility, i.e., at least one medical institution for every township in Taiwan, and the NHI becomes part of everyone’s daily life. However, the NHI also faces many potential challenges, such as too much workload for medical staff, reduced NHI coverage, and large gap of medical resource between urban and rural areas. Thus, Taiwan government proposes a series of modifications to the NHI in order to let the people have better medical care. In this study, we also aim to explore the potential challenges of NHI by analyzing the services given by different levels of medical institutes, and provide policy suggestions to Taiwan government.
In this study, we investigate the attributes of different levels of medical institutes in Taiwan by using the sample data from NHI Research Database. The sample consists of one millions of people randomly drawn in 2005 and their medical records such as the out-patient visits (CD) and medical institution data files (HOSB). We first check the sample representativeness of the random sample, making sure that they can represent Taiwan’s total population. The goal of our analysis is to compare the attributes of four different levels of medical institutes, including the patient's demographic characteristics, medical usages, and medical inequality. We found that it is difficult to differentiate the services provided by the first three levels of medical institutes (i.e., medical centers, regional hospitals, and local hospitals), and more people chose to visit medical centers and regional hospitals. On the other hand, people would choose the primary medical institutes (i.e., the fourth level) for respiratory diseases. Also, it seems there is medical inequality in the metropolitan areas. The analysis results can serve as a reference for the policy planning and medical resources allocation.

並列關鍵字

National Health Insurance ； Medical service ； Big sata ； Medical inequality ； Market segmentation

參考文獻

一、 中文部分

1. 何雍慶、吳文貴(2005)。民眾就醫選擇之研究-品牌轉換模型之應用。產業管理學報，vol.5(2)，431-445。

2. 吳肖琪、朱慧凡、黃麟珠、雷秀麗(2003)。從國際比較探討台灣每千人口需要多少醫師？，台灣公共衛生雜誌，vol.22(4)，279-286。

3. 李妙純、沈茂庭(2008)。全民健保下不同所得群體醫療利用不均因素分析，臺灣衛誌，vol.27(3)，223-231。

4. 林民浩、楊安琪、溫在弘(2011)。利用地區差異與人口學特徵評估全民健保資料庫人口居住地變項之推估原則。台灣衛誌，vol.30，347-361。