本研究探討了有關臺灣地區十大死因同時聯合去除某特定百分比時對平均壽命之影響， 吾等發現若將十大死因同時完全去除 (去除百分比 100 ％ )，刖自出生起男、女性之平均餘命將分別增加 16.29 年及 13.97 年而達 87.83 歲及 90.51 歲。在 0 歲，40 歲及 65 歲時十大死因同時被去除 10 ％，20 ％，...，100 ％時，平均餘合命之獲得和去除百分比呈二次曲線上升； 0 歲和 40 歲時男性平均餘命之獲得皆高於女性，在 65 歲時則女性略高於男性。且在十大死因聯合去除 10 ％，50 ％或 100 ％時，男、女平均餘命之獲得因年齡之增長而呈二次曲線下降。 0 ～ 60 歲間男性平均餘命之獲得均高於女性，惟 60歲以後差異減少，女性甚而較男性為高。 此外，由於工作年齡群 (15-65 歲 ) 對國家經濟發展影響至鉅，故乃針對此一特定年齡群，算出十大死因聯合部分法除對平均餘命影響之迴歸模式，在控制年齡、性別等因素下，求出其十大死因去除不同百分比迴歸係數所佔之百分比， 此可視為十大死因間競爭性風險對平均餘命之影響程度， 其依序為腦血管疾病 (21.3％ )、惡性腫瘤 (20.5 ％ )、心臟病 (20 ％ )、意外災害 (8.6 ％ )、肺炎 (5.8 ％ )、糖尿病 (5.8 ％ )、 高血壓 (5.2 ％ )、支氣管疾瘤 (5.2 ％ )、腎臟病 (3.9 ％ ) 及慢性肝病 (3.6 ％ )。此當有助於政府相關部門預防十大死因預算之合理分配。

　　　　　This paper studies the effect of jointlypartial elimination of the ten leading causes of death on life expectancy inTaiwan. We find that when all the ten causes of death are simultaneouslyeliminated, the life expectancy of male and female are incresed by 16.23 yearsand 13.97 years respectively. At age 0,40 and 65, when the ten causcs aresimultaneously eliminated at 10%, 20%, ..., and 100%, the life expectancyincreases almost quadratically. At age 0 and 40 the gain in life expectncy ofmale is greater than that of female, but at age 65 the gain of female is greaterthan that of male; also at the eliminating rates of 10%, 50% and 100% the gainin life expectancy of male is greater than that of female between age 0 and age60 while the gain of life expectancy of female might be greater than that ofmale after age 60. In addition, we investigate the gain of years of working agegroup. Through adjustments of age and sex, we contruct a regression function ofthe gain of life expectancy on the percentage of jointly partial elimination ofthe ten causes. By using the partial regression coefficients (β 's), we canfigure out each corresponding percentage of the ten leading causes of death.This can be regarded as the effect of competing risks among the ten leadingcuases of death on life expectancy. The higher the corresponding percentage, themore impact. of the cause of death on the life expectancy. By controlling ageand sex the order of corresponding percentage of regression coefficients of theten leading causes of death are Cerebrovascular disease (21.3%), Malignantneoplasm (20.5%), Heart disease (20%), Accident (8.6%), Pneumonia (5.8%),Diabetes (5.8%), Hypertension (5.2%), Bronchitis (5.2%), Nephritis (3.9%) andChronic Liver Disease (3.6%). This shows the relative effect on life expectancyamong the ten leading causes of death at working age group. Based on thecost-benefit standpoint, this result would help allocating the budget forsetting an effective preventive programs of the government.