抽菸習慣對呼吸相關疾病醫療使用之影響

The Impact of Smoking Habits on Respiratory-Related Diseases Medical Utilization

賴永翰(Yung-Han Lai)

指導教授 ： 劉彩卿

國立臺北大學/公共事務學院/財政學系/碩士(2013年)

摘要

全民健康保險醫療給付，從84年1千零60億點，上升至100年是3千750億點，達兩倍之多。因抽菸習慣有礙健康，政府於1997年制定菸害防制法，於2002年開徵菸品健康福利捐，每包菸課徵5元，並於2006年調漲為10元，於2009年調漲為20元，藉以抑制菸品消費及補助困窘之健保財務。由於抽菸行為可能直接呼吸相關器官耗損，因此本研究欲探討抽菸習慣是否導致呼吸相關疾病的醫療使用與費用上升，造成目前台灣全民健康保險醫療給付上升的原因之一。
本研究採用2005年「國民健康訪問調查」和「全民健康保險資料庫」，將資料樣本以年齡: 18歲至30歲「青少年」、31歲至60歲「壯年」以及61歲以上「老年」，以及居住區域:北部、南部、中部以及東部。探討不同年齡層以及居住區域，其抽菸習慣以及其他可能因素對呼吸相關疾病醫療使用之影響。本文使用兩部模型法(Two-part model)，首先使用邏輯斯迴歸討論影響醫療使用機率的因素。其次，利用負二項分配及最小平方法迴歸模型，探討影響醫療使用次數以及費用之因素。
實證結果發現，31-60歲群組，抽菸者較沒有抽菸者醫療使用機率意外減少。可能因抽菸者較易至疾病嚴重方行就醫，導致後續醫療費用的發生劇增。然而，61歲以上，抽菸者較沒有抽菸者醫療使用次數及費用，則明顯增加；可能因為「菸齡」較高，對健康危害較深，相關醫療服務使用增加。就居住區域而言，醫療使用機率四個區域皆達負的顯著水準，意味抽菸者較沒有抽菸者醫療使用機率低。然而，醫療使用次數及費用僅北部地區達正的顯著水準，抽菸者較沒有抽菸者有較高醫療使用次數及費用。如前所述，顯示抽菸者可能至疾病嚴重時才使用醫療，尤其是北部地區情況更加明顯。欲改善國人健康，對於抽菸族群，加強宣導預防勝於治療，尤其31-60歲群組應定期健康檢查。

關鍵字

抽菸習慣 ； 呼吸相關疾病 ； 醫療使用 ； 兩部模型法

並列摘要

Medical expenditures have rapidly increased since the implementation of National Health Insurance (NHI), from 106 billion points in 1995 to 375 billion points in 2011. Smoking behavior has been found causing the problem of health and leads to a considerable medical care use. Therefore, government set up the tobacco hazards prevention act in 1997 in order to further improve people health and indirectly relief the financial pressure of NHI. Since the respiratory organs is highly related to smoking habits. The objective of this study is to investigate whether smoking habits increase the probability of medical care use related to respiratory diseases, and the sequential the number of medical care visits and the medical expenditures
In this study, data is selected from 2005 NHIRD (National Health Insurance Research Data) and NHIS (National Health Interview Survey). To explore the impact of smoking habit and other factors on respiratory-related diseases medical utilization, the sample is classified by three age groups:18-30, 31-60, 61~ and four region areas: North, South, Center, and East. This paper use Two-part model .In the first part, we estimate the probability of medical care use by a logistic regression. In the second part, we discuss visit times of medical care use by a negative binomial regression, as well as the medical expenditures by ordinary least square regression.
Empirical results show that the odds ratio of smoke in the 31-60 age groups is less than one, suggesting that people aged 31-60, smokers are less likely to use medical care than non-smokers. Conversely, the coefficient of 61~ is shown to be positive, implying that elderly aged above 60 tended to have more sequential medical care visits and medical expenditures occurred. This may be due to elder smokers with longer-term of smoking period severely impact their health. For the four regional areas, all of them have odds ratios less than one, indicating that smokers have lower probability of medical care use than their counterparts in four areas. However, in the second part, only in the north area, the coefficient of smoker is found significant and positive. This finding suggests that northern smokers have higher number of sequential medical care visits and more sequential medical expenditures occurred than northern non-smokers. This reason as indicated above, smokers might not visit doctors until illness was severe, and cause higher medical expenditures later on. The health authority may need to pay more attentions on those groups of people, particularly for the age 31-60, to further improve their health, such as the promotion of health examination use to early detect the health problem.

並列關鍵字

smoking habits ； respiratory diseases, medical care utilization ； two-part model

參考文獻

行政院衛生署國民健康局，http://www.bhp.doh.gov.tw。

行政院衛生署，http://www.doh.gov.tw。

江東亮、文羽萍、黃旭明 ，(2012) 台灣醫療保健支出成長率的分析：醫療通膨，質量與公平性， 台灣衛誌， 31(1)， 1-10。

江東亮 ，(2002) 台灣醫療保健支出之趨勢分析， 台灣衛誌， 21(3)， 157-163。

邱雅苓、陳世能 ，(2003) 醫療保健支出成長因素之探討- 時間序列分析與門檻模型的應用， 經濟研究， 39(2)， 197-240。