A1360
Title: Additive shared mixed frailty model for bivariate gap time data
Authors: Ivo Sousa-Ferreira - FCiencias.ID (Portugal) [presenting]
Ana Maria Abreu - UMa and CIMA (Portugal)
Cristina Rocha - CEAUL (Portugal)
Abstract: Recurrent event data frequently arise in biomedical studies, where within-subject dependence and the presence of zero-recurrence subjects challenge classical survival models. A new additive shared frailty model for gap times between recurrent events is introduced, assuming that the recurrence process follows a non-homogeneous Poisson process with a Weibull rate function. The frailty acts additively on the rate function and has a non-central chi-squared distribution with zero degrees of freedom, a mixed distribution with a probability mass at zero and a continuous positive component. This formulation simultaneously accommodates within-subject correlation and zero-recurrence subjects. The resulting multivariate marginal survival function admits a competing risks interpretation with two independent causes. Since the model is fully specified, the maximum likelihood method is applied for parameter estimation using a marginal likelihood function, with particular attention to the bivariate recurrent event setting. A brief simulation study illustrates the asymptotic properties of the estimators under different censoring percentages. An application to a well-known hospital readmission data set is also presented to elucidate the practical contribution of the proposed model.
