有序雙重事件時間分析使用與時間相關的共變數－邊際方法的比較

Ordered Bivariate Survival Time with Time Dependent Covariate -Comparison of Marginal Method

陳安朋(An-peng Chen)

指導教授 ： 曾議寬

國立中央大學/理學院/統計研究所/碩士(2007年)

摘要

台灣地區的衛生署自1997年4月開始免費提供「雞尾酒療法」之治療藥物給全國各愛滋病指定醫院的病患使用，至今已達十年之久。我們感興趣的是雞尾酒療法對於愛滋病發病前後的療效差異，故本篇使用台灣地區136個愛滋病患的資料，焦點放在多維存活時間的邊際方法，如AG(Andersen and Gill, 1982)模型、WLW(Wei, Lin, and Wiessfeld, 1989)模型和PWP(Prentice, Williams and Petersen, 1981)模型之比較，並探討使用雞尾酒療法，對於兩段存活時間，愛滋病毒檢驗呈現陽性到發病的時間和愛滋病發病到死亡的時間，療效之差異及CD4細胞數量之影響。

關鍵字

none

並列摘要

The Department of Health in Taiwan began to freely provide the treatment of HAART (highly active antiretroviral therapy) for the AIDS patients in the appointed hospitals all over the country ever since April, 1997 and up to now, reach a decade period. We are interested in the different effects of HAART to the 136 AIDS patients before and after the onset of AIDS. To investigate this research problem, we focus on three marginal approaches, the AG (Andersen and Gill, 1982) model, WLW (Wei, Lin, and Wiessfeld, 1989) model and PWP (Prentice, Williams and Petersen, 1981) model. In addition to compare the performance of the three approaches, we also study the effect of CD4 count to both survival times.

並列關鍵字

none

參考文獻

1. Andersen, P. K. and Gill, R. D.(1982). Cox''s regression model for counting processes: A large sample study. Annals of Statistics, 10:1100-1120.

2. Cai, J. and Prentice, R. L.(1995). Estimating equations for hazard ratio parameters based on correlated failure time data. Biometrika, 82:151–164.

3. Clayton, D. G.(1978). A model for association in bivariate life tables and its application in epidemiological studies of chronic disease incidence. Biometrika, 65:141–151.

4. Clayton, D. G. and Cuzick, J.(1985). Multivariate generalisations of the proportional hazards model. Journal of the Royal Statistical Society, Series A, 148:82–117.

5. Clegg, L. X., Cai, J. and Sen, P. K.(1999). A marginal mixed baseline hazards model for multivariate failure time data. Biometrics, 55: 805–812.