Title: Multi-state models for multi-type recurrent events and terminal events with feedback in longitudinal covariates
Authors: Chuoxin Ma - University of Cambridge (United Kingdom) [presenting]
Jianxin Pan - The University of Manchester (United Kingdom)
Abstract: In cardiovascular disease study, one of the main interest is to investigate the association between risk factors and a series of multi-type recurrent events and terminal events, such as myocardial infarction, stroke and cardiovascular death. When the trajectories of some biomarkers contain past event feedbacks, existing approaches handling time-dependent covariates in event history analysis can be problematic. We propose a class of multi-state models for the analysis of multi-type recurrent events and terminal events when biomarkers contain past event feedback and are intermittently observed and subject to measurement errors. The competing risk structure and the progressive nature of the multiple events can be well captured by state-specific intensity functions. Both time-varying and constant coefficients can be accommodated. Estimation procedure based on polynomial splines approximation and an extension to the corrected score approach is developed. The consistency and asymptotic normality of the proposed estimators are provided. Simulations show that the naive estimators which either ignore the past event feedback or the measurement errors are biased. Our method achieves better coverage probability of the time-varying/constant coefficients, compared to the naive methods. An application to the data set from the Atherosclerosis Risk in Communities Study is presented.