Statistical Analysis for Average Cost-Effectiveness Ratios

• Main Authors: Hsiao-Ting Wu, 吳筱婷
• Other Authors: Tsai-Yu Lin
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/m5z2g6

碩士 === 逢甲大學 === 應用數學學系 === 103 === Cost-effectiveness analysis has been frequently-used in economic, biomedical, social sciences, public health and other fields. The purpose is to compare the cost-effectiveness of two treatments to determine whether they are significant difference. In the past, the most widely-used indices of the cost-effectiveness analysis is the incremental cost-effectiveness ratio (ICER). In practice, comparing cost-effectiveness for multiple treatments is often faced. Based on the ICER, new treatments can be compared with standard treatment to decide which treatment is most cost effective. It is rarely mentioned the comparison of multiple treatments in previous references. In this thesis, we provide generalized p-value (GPV) methods to test the equality of the cost-effectiveness for multiple treatments by using the concept of average cost-effectiveness ratio (ACER), including the procedures of GPV^M,GPV^"ln" (M) and GPV^(〖"tan" 〗^"-1" (M) ). Since the cost and effectiveness usually be positively skewed, we assume the cost and effectiveness are bivariate log-normal distribution. Further, we conduct a simulation study to evaluate the performance of three GPV procedures based on their empirical size and the power. The simulation results show that the proposed GPV^(〖"tan" 〗^"-1" (M) ) is recommended. Finally, the proposed methods can be used practical applications such as depression and back pain.