Premium
Censored regression with the multistate accelerated sojourn times model
Author(s) -
Huang Yijian
Publication year - 2002
Publication title -
journal of the royal statistical society: series b (statistical methodology)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.523
H-Index - 137
eISSN - 1467-9868
pISSN - 1369-7412
DOI - 10.1111/1467-9868.00322
Subject(s) - covariate , inference , goodness of fit , regression , regression analysis , statistics , logarithm , computer science , term (time) , econometrics , mathematics , artificial intelligence , mathematical analysis , physics , quantum mechanics
Many disease processes are characterized by two or more successive health states, and it is often of interest and importance to assess state‐specific covariate effects. However, with incomplete follow‐up data such inference has not been satisfactorily addressed in the literature. We model the logarithm‐transformed sojourn time in each state as linearly related to the covariates; however, neither the distributional form of the error term nor the dependence structure of the states needs to be specified. We propose a regression procedure to accommodate incomplete follow‐up data. Asymptotic theory is presented, along with some tools for goodness‐of‐fit diagnostics. Simulation studies show that the proposal is reliable for practical use. We illustrate it by application to a cancer clinical trial.