Abstract

Recurrent event data frequently occur in longitudinal studies, and it is often of interest to estimate the effects of covariates on the recurrent event rate. This paper considers a flexible semi-parametric additive-multiplicative transformation mean model for recurrent event data, which includes the multiplicative model and additive transformation model as special cases. The new model is flexible in that they allow for both additive and multiplicative covariates effects, and additive effects are allowed to be time-varying. The estimation of regression parameters in the model is given by using the idea of estimating equations, and the asymptotic properties of the resulting estimators are established. Numerical studies under different settings were conducted for assessing the proposed methodology and an application to a bladder cancer study is illustrated. The results suggest that they work well.

摘要

复发事件数据经常出现在纵向研究中，估计协变量对复发事件率的影响通常很有趣。本文考虑了一种灵活的半参数加法-乘法变换均值模型用于循环事件数据，其中包括乘法模型和加法变换模型作为特例。新模型是灵活的，因为它们允许加法和乘法协变量效应，并且允许加法效应随时间变化。利用估计方程的思想给出了模型中回归参数的估计，并建立了所得估计量的渐近性质。进行了不同环境下的数值研究以评估所提出的方法，并说明了在膀胱癌研究中的应用。结果表明它们运作良好。