雙向截切資料在加速失效時間模式的半母數分析

Semiparametric Analysis of Accelerated Failure Time Models With Doubly Truncated Data

吳性慧(Sing-Huei Wu)

指導教授 ： 張淑惠

國立臺灣大學/公共衛生學院/流行病學與預防醫學研究所/碩士(2013年)

https://doi.org/10.6342/NTU.2013.10100

摘要

雙向截切資料在臨床醫學研究中是一種常見病例個案研究之資料結構，其僅截取某一特定時間區間內發生感興趣事件的個體進行研究，例如：由醫院病歷檔、癌症登記檔或全民健保資料庫中，收集發生特定疾病的病患資料，此種限制抽樣的研究設計，導致由樣本在推估母群體時產生偏差。周(2009)針對半母數加速事件時間模式，提出了一種以U統計量為基礎的回歸參數估計方法，本研究將進一步地對其估計方法進行校正、推廣與討論。模擬結果顯示，研究所提方法未優於周(2009)之估計式，且對於母體分布設定相當敏感。

關鍵字

雙向截切 ； 半母數 ； 加速事件時間模型 ； U統計量

並列摘要

Doubly truncated data are common data structures of case-cohort studies in clinical practices. The researchers collect cases with a specific event of interest within a given time period. For example, we may collect cases information from medical records, cancer registry records and National Health Insurance Research Database. Such study designs accompanied with some sampling criterions result in biases when inferring population. Chou (2009) proposed a U-statistic-based estimating equation for parameter estimation in AFT model. In this study, we modify the estimating equation and further conduct more simulation scenarios. The simulation results show that the proposed methods are not superior to Chou (2009) and sensitive to population distributions.

並列關鍵字

doubly truncated data ； semi-parametric ； AFT model ； U statistic

參考文獻

Bhattacharya, P. K., Chernoff, H. and Yang, S. S. (1983) Nonparametric estimation of the slope of a truncated regression. The Annals of Statistics, 11, 505-514.

Chen, Y. Q. and Jewell, S. W. (2001) On a general class of semiparametric hazards regression models. Biometrika, 88, 687-702.

Chou, Tong-Yang and Chang, Shu-Hui * (2009). A semi-parametric accelerated failure time model for doubly-truncated data. Technique Report in Division of Biostatistics, School of Public Health, NTU,.

Cox, D. R. (1972) Regression models and life-tables. Journal of Royal Statistical Society Series B, 34, 187-220.

Finkelstein, D. M. (1986) A proportional hazards model for intervalcensored failure time data. Biometrics, 42, 845-854.