台灣的健康及醫療照護的社會差距

Social Disparities in Health and Medical Care in Taiwan

劉容華(Jung-Hua Liu)

指導教授 ： 江東亮

國立臺灣大學/公共衛生學院/衛生政策與管理研究所/博士(2006年)

https://doi.org/10.6342/NTU.2006.10461

摘要

背景與目的：臺灣於1995年3月1日開辦全民健康保險，旨在確保所有國民的就醫權利，以增進其身心健康。本研究從醫療體系的結構、過程，與結果三個面向，來檢視與分析1986-2004年臺灣的健康與醫療照護地理差異的變化，尤其著重在全民健康保險的影響。具體而言，本研究的目標有三：(1) 分析 1986-2004 年間，醫師人力地理分布的變化；(2) 檢視1994-2004 年間，地區社經別醫療服務利用的變化；以及 (3) 探討 1986-2004 年間，地區社經別可避免死亡率的變化。

方法：觀察期間由1986年到2004年，資料來源為臺灣地區公私立醫療院所現況調查、家庭收支調查、1994與2001年國民健康訪問調查、死亡登記檔、臺閩地區人口統計，以及1980年臺閩地區戶口及住宅普查。在探討醫師人力地理分布與可避免死亡方面，採區位分析，分析單位包括縣市與鄉鎮市區兩類；在醫療服務利用方面則採多層級分析，分析單位包括個人與區位層次。使用的分析方法包括描述統計、分節迴歸分析、多層級受限迴歸與多層級對數複迴歸分析。

結果：1986-2004年間，總醫師人力的供給由 15,852名增加為33,259名，成長109%。其中，診所醫師人力與醫院醫師人力比由0.87 減少為0.60。分節複迴歸分析結果顯示：(1) 無論是縣市層級或鄉鎮市區層級，醫院醫師人力及診所醫師人力的供給皆與地區社經發展有顯著相關；以及(2) 全民健保實施以後，醫院醫師人力的地理差異仍繼續減少，相較於診所醫師人力的地理差異反而擴大。

由國民保健調查發現：2001年的門診服務、住院服務，與急診服務的利用率分別為37.2%、8.0 % 與 10.6%。而1994年則分別為 23.8%、10.5 % 與 9.3%。多層級受限迴歸與多層級對數複迴歸分析顯示：住院服務與地區社經發展無顯著相關，但居住在越剝奪的地區者，西醫門診利用次數越少，且較少利用急診服務。然而進一步以有無保險分層分析，在控制個人社經地位後，僅2001年有保險者的急診服務利用與地區社經發展達顯著統計水準。

1986-2004年間，5-64歲標準化總死因死亡率由2.8‰下降為1.3‰，共減少25%。其中，可避免死因死亡率減少34%，而不可避免死因死亡率減少 8%。分節複迴歸分析結果顯示：(1) 無論是縣市層級或鄉鎮市區層級，可避免死因及不可避免死因死亡率皆與地區社經發展有顯著相關；並且(2)相對於不可避免死因死亡率的地理差異在全民健保開辦後加速惡化，可避免死因死亡率地理差異的擴大速度減緩，尤其是初級預防死因死亡率。

結論：實施全民健保有助於縮小臺灣健康的社會差距，但與改善醫師人力之地理分布及醫療利用的社會差距無顯著相關。

關鍵字

地理差異 ； 醫師人力 ； 醫療服務利用 ； 可避免死亡率

並列摘要

Background and Objectives: On March 1, 1995, the Taiwan government inaugurated a national health insurance to assure health care for all citizens and reduce disparities in health. This study applied a structure-process-outcome framework to examine the changing geographic disparities in health and medical care in Taiwan between 1986 and 2004, with an emphasis on the impact of the national health insurance scheme. In specific, the objectives of this study are: (1) to analyze the changing geographic distribution of physician manpower between 1986 and 2004, (2) to examine the changing geographic differences in medical services utilization between 1994 and 2004, and (3) to investigate the changing geographic variation of avoidable mortality between 1986 and 2004.

Methods: The study period was from 1986 to 2004. Data for the analysis came from the annual national surveys of hospitals and clinics, the annual family income and expenditure survey, the 1994 and 2001 national health interview survey, the databases of death certificate, the annual demographic fact book, and the 1980 population and housing census. For the geographic disparities of physician manpower and avoidable mortality, we used area as the unit of analysis, including the level of city/county and the level of township. For medical services utilization, we applied a multilevel method by taking into account both individual and area levels. Statistical methods used included descriptive statistics, segmental regression, multilevel censored regression, and multilevel logistic regression.

Results: Between 1986 and 2004, the total supply of physician manpower in Taiwan increased by 109% from 15,852 to 33,259, with the ratio of office-based physician to hospital physician decreasing from 0.87 to 0.60. By using the segmental regression analysis, we found that: (1) the supply of physician manpower was significantly associated with area socioeconomic development; and (2) after the implementation of the national health insurance, the geographic disparities of hospital physician manpower continuously decreased, in contrast to the widening geographic disparities of office-based physician manpower.

The results from the national health interview survey showed that, the likelihood of any use for physician services, inpatient services, and emergency services was 37.2%, 8.0%, and 10.6%, respectively, in 2001, while the comparable figures for 1994 were 23.8%, 10.5%, and 9.3%. By using the multilevel censored regression and multilevel logistic regression analysis, we found that, although hospital admission was not significantly associated with area socioeconomic development, people living in deprived areas tended to use more physician services and were less likely to use emergency services. However, further stratification analyses by health insurance coverage, only use of emergency services for the insured in 2001was significantly associated with area socioeconomic development, after adjustment for individual socioeconomic position.

Between 1986 and 2004, age-standardized all-cause mortality for aged 5-64 in Taiwan declined by 25% from 2.8‰ to 1.3‰. This was made up of a 34% fall in avoidable mortality and a 8% fall in non-avoidable mortality. By using the segmental regression, we found that: (1) avoidable mortality and non-avoidable mortality were significantly associated with area socioeconomic development; and (2) the geographic disparities of avoidable mortality increased at a slower pace, in contrast to the geographic disparities of non-avoidable mortality worsened more rapidly after the introduction of the national health insurance. This was especially significant for avoidable mortality associated with primary prevention.

Conclusions: The implementation of the national health insurance is helpful for closing the gaps in health disparities in Taiwan. However, the national health insurance was not found to be associated with amelioration of geographic disparities of physician manpower and medical services utilization.

並列關鍵字

geographic disparity ； physician manpower ； medical services utilization ； avoidable mortality

參考文獻

呂宗學、石台平、賴華山、李麗雪、李孟智、周明智：台灣地區死因統計準確度評估 (Ⅰ)：可能錯誤來源與評估方法。中華衛誌 1995；14：179-90。

石曜堂、葉金川、楊漢泉、羅紀瓊、張明正與吳正儀：台灣地區國民自付醫療費用調查：1992年國民醫療保健支出調查之初步發現。中華衛誌 1994；13：473-84。

劉容華、江東亮：臺灣醫療次區醫院產業擴張與診所醫師人力消長的關係。台灣衛誌 2004;23:32-6.

劉容華、江東亮：台灣小型醫院新設與歇業之影響因素。台灣衛誌 2001;20:27-33。

朱僑麗、薛亞聖、江東亮。醫師人力供給的多寡是否會影響被保險人的門診利用?- 1994年國民醫療保健調查之發現。中華衛誌 2000；19：381-8。