The Study of Quality Adjusted Survival Time under First-Order Autoregression Process

• Main Authors: Wan-I Lin, 林宛儀
• Other Authors: Cheng-Hwang Perng
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/27592166670954293082

碩士 === 真理大學 === 數理科學研究所 === 97 === In the past clinical trials, to evaluate the effect of different treatments, it is common to depend on the treatment time or the survival time. In the last few years, quality of life is another criterion. Both the patients and doctors pay much attention to quality of life during/ after treatment. To extend survival time and improve quality of life are the reaching targets. However, these two indicators are usually trade off. How to combine survival time with quality of life into a composite criterion is an important subject. Quality adjusted survival time (QAST) is one of the composite estimate indicators. For the QAST research, the analyzed result of the former paper appears to multiply with quality of life and survival time. The result of expected quality adjusted survival time will be less than expected original survival time. Cheng- Hwang Perng (2001) proposed the concept of QAST that will be larger/less than or even equal to original survival time. It relates to patient’s quality of life that relies on the change of the time closely. This thesis is based on the conception and estimation of Cheng-Hwang Perng (2001). It assumes survival time follow exponential distribution and quality of life initial value follow beta distribution. In order to correlate present quality of life with former quality of life, each quality of life were evaluations generated from AR(1) model. In addition, we set parameters of AR(1) model between 0 and 1, calculates expected QAST from different parameter combinations. To compare the difference of expected QAST with expected original survival time and how expected QAST is going to change as the different parameter settings of the model.