Regression models for interval censored data using parametric pseudo-observations.

摘要: Time-to-event data that is subject to interval censoring common in the practice of medical research and versatile statistical methods for estimating associations such settings have been limited. For right censored data, non-parametric pseudo-observations proposed as a basis regression modeling with possibility use different association measures. In this article, we propose method calculating data. We develop an extension recently developed set parametric based on spline-based flexible estimator. The inherent competing risk issue event interest necessitates illness-death model, formulate our within framework. To evaluate empirical properties method, perform simulation study calculate well alternative approaches. also present analysis real dataset patients implantable cardioverter-defibrillators who are monitored occurrence particular type device failures by routine follow-up examinations. dataset, information exact times so can compare analyses those obtained using Our simulations show provides unbiased estimates cumulative incidence function exposure variables appropriate coverage probabilities. suggests which agreement from approach solution specific challenges arise This allows range