Extended Hazard Regression for Censored Survival Data with Covariates: A Spline Approximation for the Baseline Hazard Function

A regression model for censored survival data with covariates is introduced and termed extended hazard regression (EHR). EHR includes the proportional hazards (PH) and the accelerated failure time (AFT) models as special cases. By approximating the baseline hazard function with quadratic splines, we...

Full description

Saved in:
Bibliographic Details
Published in:Biometrics 1987-03, Vol.43 (1), p.181-192
Main Authors: Etezadi-Amoli, Jamshid, Ciampi, Antonio
Format: Article
Language:English
Subjects:
Approximation
Biological and medical sciences
Biometrics
Censored data
Computerized, statistical medical data processing and models in biomedicine
Datasets
Inference
Mathematical independent variables
Maximum likelihood estimation
Maximum likelihood estimators
Medical sciences
Medical statistics
Polynomials
Regression analysis
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 regression model for censored survival data with covariates is introduced and termed extended hazard regression (EHR). EHR includes the proportional hazards (PH) and the accelerated failure time (AFT) models as special cases. By approximating the baseline hazard function with quadratic splines, we develop a maximum likelihood estimation procedure to provide a simultaneous estimate of both the hazard function and the regression coefficients. The AFT and PH assumptions can then be compared by likelihood ratio tests within the more general EHR model. Several examples based on artificial and real data are discussed.
ISSN:0006-341X
1541-0420
DOI:10.2307/2531958