A NOTE ON SEMI-MARKOV MODELS FOR PARTIALLY CENSORED-DATA
成果类型：
Article
署名作者：
DINSE, GE; LARSON, MG
署名单位：
Harvard University; Harvard T.H. Chan School of Public Health
刊物名称：
BIOMETRIKA
ISSN/ISSBN：
0006-3444
发表日期：
1986
页码：
379386
关键词：
摘要：
A recharacterization of the semi-Markov model described by Lagakos, Sommer and Zelen (1978) permits the nonparametric maximum likelihood estimators to be expressed in terms of familiar and easily computable quantities, such as event-specific hazard estimators and Kaplan and Meier (1958) survival estimators. In addition to simplifying the calculation and clarifying the interpretation of the previously derived estimators, this reformulation also facilitates the estimation of covariance terms and suggests a simple test of equality among the outcome-specific distributions of sojourn times in a particular state.