REGRESSION-ANALYSIS OF MULTIVARIATE GROUPED SURVIVAL-DATA

Multivariate failure time data arise when each study subject may exper ience several types of event or when there are clusterings of observat ional units such that failure times within the same cluster are correl ated. The failure times are often subject to interval grouping or have truly discrete measurements. In this paper, the marginal distribution for each discrete failure time variable is formulated by a grouped-da ta version of the proportional hazards model while the dependence stru cture is unspecified. Generalized estimating equations in the spirit o f Liang and Zeger (1986, Biometrika 73, 13-22) are proposed to estimat e the regression parameters and survival probabilities. The resulting estimators are consistent and asymptotically normal. Robust estimators for the limiting covariance matrices are constructed. Simulation stud ies demonstrate that the asymptotic approximations are adequate for pr actical use and that ignoring the intracluster dependence in the varia nce-covariance estimation would lead to invalid statistical inference. A psychological experiment is provided for illustration.