在健康經濟學、公共衛生與健康科學的領域裡,成本效果分析經常在臨床 試驗中用以評估不同治療方式或不同介入方式的成本效果,而成本效果增 量比、增量淨效益、增量淨健康效益、成本效果可接受曲線等指標都是一 般最常使用的分析工具。成本效果機率是另一種分析工具,它是以機率分 配比較不同治療方式的成本效果。在實務上,成本資料多屬偏斜分配,而 上述的分析工具都是以平均數為基礎進行推論,平均數卻對於極端值相當 敏感,但對成本效果機率卻影響有限。在此計畫中,我們以成本效果機率 來探討偏斜分配對不同評估成本效果工具的影響;並在臨床試驗樣本數有 限的情況下,提出廣義樞紐量估計法建立確切的有母數推論,並以數個不 同分配的臨床試驗成本效果資料比較不同評估工具分析結果的異同與原由。 In health economics, public health and health science, the cost-effectiveness analysis is a type of economic evaluation that examines the costs-effectiveness of two competing treatments or interventions. The cost-effectiveness data are usually collected from randomized clinical trial. Traditionally, the incremental cost-effectiveness ratio (ICER) and its derivative measures, incremental net benefit, incremental net health benefit, and cost-effectiveness acceptability curve, are used to be analytic indices for cost-effectiveness analysis. The probability of cost-effectiveness is another measure, which expresses the chance of gaining net benefit based on the probability distribution of cost and effectiveness. Unlike the ICER and derivative measures of ICER that are constructed by mean cost and mean effectiveness, the probability of cost-effectiveness is not sensitive to extreme value. In this project, we will discuss the influence of skewed distribution and symmetric distribution for ICER, derivative measures of ICER and probability of cost-effectiveness. Consider the limited sample size in clinical trials we will propose an exact parametric inference for probability of cost-effectiveness based on the concept of generalized pivotal quantities. Finally simulation studies and numerical studies will be conducted by several real data sets that collected from public clinical trails, and analyze the statistical results focus on the distribution of these data sets.