교차계획 구간절단 생존자료의 비례위험모형을 이용한 분석 = Analysis of Interval-censored Survival Data from Crossover Trials with Proportional Hazards Model

국문 초록 (Abstract)

협심증 치료의 신약에 대한 교차계획 임상시험(crossover clinical trials)에서 신약의 효능을 알아보는 운동테스트(treadmill exercise test) 결과는 중도절단 생존시간(censored survival times)으로 측정된다. 이 논문에서는 교차계획에서 수집된 중도절단 생존자료의 여러 가지 분석법에 대해 설명한다. 중도절단을 감안한 비모수적 방법들과 층화 Cox 비례위험모형 (stratified Cox proportional hazards model)에 근거한 분석법이 제시되었다. 한편, 교차계획의 두 시기에 걸쳐 수집된 생존시간의 차(difference)로부터 구간절단자료(interval censored data)가 생성되며 이에 근거한 분석법으로서 이 논문에서는 구간절단자료에 대한 Cox 비례위험모형 (proportional hazards model)의 가능성을 알아보며, 예제 자료로써 여러 방법들의 결과를 비교해 본다.

다국어 초록 (Multilingual Abstract)

Crossover trials of new drugs in the treatment of angina pectoris, which frequently use treadmill exercise test for the assessment of its efficacy, produce censored survival times. In this paper we consider analysis approaches for censored survival times from crossover trials. Previously, a stratified Cox model for paired observation and nonparametric methods have been presented as possible analysis methods. On the other hand, the differences of two survival times would produce interval-censored survival times and we propose a Cox model for interval-censored data as n alternative analysis method. Example data is analyzed in order to compare these different methods.

참고문헌 (Reference)

1 Fine, J., "Why bother with nonparametric estimators? In the Section of \A Practicum on Interval Censoring" 2006

2 Freeman, "The performance of the two-stage analysis of two-treatment Statistics in Medicine" 1989

3 Turnbull, "The empirical distribution function with arbitrarily grouped Journal of the Royal Statistical Society" 1976

4 France, "The analysis of failure time data in crossover studies Statistics in Medicine" 1991

5 Hoel, "Statistical analysis of survival experiment Journal of the National Cancer Institute" 1972

6 Turnbull, "Nonparametric estimation of a survivorship function with doubly censored data Journal of the American Statistical Association" 1974

7 Kim, J. H, "Logrank-type tests for comparing sur-vival curves with interval-censored data" 50 : 3165-1378, 2006

8 Feingold, M, "Cross-over trials with censored data" 15 : 953-967, 1996

9 Peto, "Conservatism of the approximation P 2=E in the logrank test for survival data or tumor incidence data" 1973

10 Finkelstein, "A semiparametric model for regression analysis of interval-censored failure time data" 1985

11 Finkelstein, "A proportional hazards model for interval-censored failure time data" 1986

12 Lee, "A Monte-Carlo study of the power of some two-sample tests" 1975