An Illness-Death Process with Time-Dependent Covariates

A general model for the illness-death stochastic process with covariates has been developed for the analysis of survival data. This model incorporates important baseline and time-dependent covariates in order to make an appropriate adjustment for the transition and survival probabilities. The follow...

Full description

Saved in:
Bibliographic Details
Published in:Biometrics 1989-06, Vol.45 (2), p.669-681
Main Authors: Chiang, Yu-Kun, Hardy, Robert J., Hawkins, C. Morton, Kapadia, Asha S.
Format: Article
Language:English
Subjects:
Approximation
Biological and medical sciences
Biometrics
Biometry - methods
Clinical Trials as Topic
Computerized, statistical medical data processing and models in biomedicine
Diseases
Health hazards
Humans
Maximum likelihood estimation
Medical sciences
Medical statistics
Modeling
Models, Statistical
Morbidity
Mortality
Musical intervals
Myocardial infarction
Myocardial Infarction - epidemiology
Myocardial Infarction - mortality
Myocardial Infarction - prevention & control
Placebos
Propranolol - therapeutic use
Stochastic Processes
The Consultant's Forum
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A general model for the illness-death stochastic process with covariates has been developed for the analysis of survival data. This model incorporates important baseline and time-dependent covariates in order to make an appropriate adjustment for the transition and survival probabilities. The follow-up period is subdivided into small intervals and a constant hazard is assumed for each interval. An approximation formula is derived to estimate the transition parameters when the exact transition time is unknown. The method developed is illustrated with data from a study on the prevention of the recurrence of a myocardial infarction and subsequent mortality, the Beta-Blocker Heart Attack Trial (BHAT). This method provides an analytical approach with which the effectiveness of the treatment can be compared between the placebo and propranolol treatment groups with respect to fatal and nonfatal events simultaneously.
ISSN:0006-341X
1541-0420
DOI:10.2307/2531509