Generalized mean residual life models for case-cohort and nested case-control studies
- PDF / 947,790 Bytes
- 31 Pages / 439.37 x 666.142 pts Page_size
- 25 Downloads / 148 Views
Generalized mean residual life models for case-cohort and nested case-control studies Peng Jin1 · Anne Zeleniuch-Jacquotte1,2 · Mengling Liu1,2 Received: 29 September 2019 / Accepted: 25 May 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Mean residual life (MRL) is the remaining life expectancy of a subject who has survived to a certain time point and can be used as an alternative to hazard function for characterizing the distribution of a time-to-event variable. Inference and application of MRL models have primarily focused on full-cohort studies. In practice, case-cohort and nested case-control designs have been commonly used within large cohorts that have long follow-up and study rare diseases, particularly when studying costly molecular biomarkers. They enable prospective inference as the full-cohort design with significant cost-saving benefits. In this paper, we study the modeling and inference of a family of generalized MRL models under case-cohort and nested casecontrol designs. Built upon the idea of inverse selection probability, the weighted estimating equations are constructed to estimate regression parameters and baseline MRL function. Asymptotic properties of the proposed estimators are established and finite-sample performance is evaluated by extensive numerical simulations. An application to the New York University Women’s Health Study is presented to illustrate the proposed models and demonstrate a model diagnostic method to guide practical implementation. Keywords Counting process · Estimating equations · Inverse probability weighting · Model checking · Martingale residuals
B
Mengling Liu [email protected] Peng Jin [email protected] Anne Zeleniuch-Jacquotte [email protected]
1
Department of Population Health, New York University School of Medicine, New York, NY 10016, USA
2
Department of Environmental Health, New York University School of Medicine, New York, NY 10016, USA
123
P. Jin et al.
1 Introduction The mean residual life (MRL) is the remaining life expectancy given that a subject has survived to a certain time point. For a non-negative survival time T with finite expectation, the MRL at time t is m(t) = E(T − t|T > t). As alternatives to models based on the hazard function, those based on the MRL have a more intuitive explanation often leading to easier communication with patients. For instance, providing patients who have been on a treatment for a year with their remaining life expectancy is likely to be more informative than providing them with instantaneous hazards. Various statistical models have been proposed to characterize the MRL function given covariates. Specifically, the proportional MRL model (Oakes and Dasu 1990; Maguluri and Zhang 1994) is,
m(t|Z) = m 0 (t) exp(β Z),
(1)
where β is a vector of regression parameters characterizing the multiplicative effects of covariates on the MRL function, and m 0 (t) is an unknown baseline MRL function. The estimation and inference for model (1) have been developed
Data Loading...
