全民健康保險自民國八十四年實施以來，其財務狀況一直是各方關注的焦點。而在民國八十七年醫療支出大於當年度的保費收入進而衍生短絀50億，且持續至八十九年皆呈現入不敷出，收支不平衡的狀況。本研究嘗試以總體的角度，歸納所有可能影響「總體全民健康保險醫療支出」的因素，並將所有的影響因素，經統計分析，證實影響總體全民健康保險醫療費用的主因，並將影響因素擬定成一個合理的公式，進而推估總體全民健康保險醫療費用的支出。 本研究觀察期間為民國八十五年至民國八十九年，共五年的資料，以月為分析單位。而可能影響「平均每人每月實質全民健康保險醫療費用」的因素為「平均每月每人實質國內生產毛額」、「六十五歲以上的老人人口比例」、「每百萬人口昂貴或高危險醫療儀器的數目的成長率」、「每萬人口醫師數」。這些變項經由時序性檢定、共線性檢定後才開始作分析，並由單根檢定證實為一階穩定的序列，因此不適用一般的迴歸，需要經一階差分後才能加以分析。因而採用經過差分後的變項作複迴歸分析，並以ARMA修正殘差。 實證結果，發現第前一月至第前三月的「平均每人每月實質國內生產毛額的改變率的上一年度同月份的變動率」、當月「每月六十五歲以上的老人人口比例的改變比例」及第前四個月的「每月每萬人口醫師數的變動率」對醫療費用有顯著的影響。證實了Getzen（1991）的論點，當期的醫療費用受到前幾期的所得所影響。而台灣老人人人口比例上升速度高居世界第二（行政院經濟建設委員會,2000），所以在考量影響總體的醫療費用時，仍需考量人口結構中的「六十五歲以上的老人人口」。在供給面因素中的「每萬人口醫師數」對醫療費用有顯著的影響，這或許證實了在短期來說醫師數的增加會有供給誘導需求的現象產生。而將上述因子納入總額預算公式時，仍需考量物價指數波動對醫療費用的影響的影響，所以額外納入「國內生產毛額平減指數」來推估全民健康保險醫療費用的變動率。 而將上述諸多因子納入全民健康保險醫療費用變動率的推估公式時，仍需考量的協商因素約佔「平均每月每人實質全民健康保險醫療費用的變動率」再加減3.4752%。; The national health insurance (NHI) scheme was inaugurated in 1995 in Taiwan. Since then the expanding health care expenditure has become a major political issue. This study aims to explore the determinants of health care expenditure of NHI. It is also hopeful that the results of this study can imply the policy on the national cap of health care in the NHI scheme. Time-series regression is used, including multiple regression and ARMA residuals, to analyze the monthly data of health care expenditure over the period from 1996 to 2000. The explanatory variables are GDP per capita, proportion of the aged, hospital bed and doctor-population ratios. These variables are analyzed by Granger cause test, variance inflation factors test and unit root test prior to the time-series analysis. They have to be differenced in order to achieve stationarity. The percentage changes of real health care expenditure per capita are explained by the following variables: the percentage changes of GDP per capita between current and the prior first to fourth months, and between current and the prior 12th month, the ratio changed of the proportion of elderly, and the percentage change of doctor-population ratio. The percentage change of GDP per capita between current and the prior 12th month is the most important determinant. The results have implications on the national cap policy of health care. The GDP can and should be considered one of the most important components in the cap formula, together with GDP deflator for dealing with inflation. Beside the two components, it remains 3.48% of the health care expenditure to be explained or determined by other factors considered in this study.
