Standard Article

Accelerated Failure-time Models

  1. Ian James

Published Online: 15 JUL 2005

DOI: 10.1002/0470011815.b2a11003

Encyclopedia of Biostatistics

Encyclopedia of Biostatistics

How to Cite

James, I. 2005. Accelerated Failure-time Models. Encyclopedia of Biostatistics. 1.

Author Information

  1. Murdoch University, Perth, Western Australia, Australia

Publication History

  1. Published Online: 15 JUL 2005


The accelerated failure-time model assumes a survival function of the form S(t) = S0t), where S0 is an underlying survival function and θ may depend on a number of covariates. This is equivalent to a location-shift model for the log failure time, and in particular to a loglinear regression model when θ is loglinear. Parametric and semiparametric approaches to analyses based on these models with possibly censored failure-time responses are reviewed. The accelerated failure-time assumption is compared and contrasted with that of proportional hazards.


  • accelerated failure time;
  • survival;
  • hazard rate;
  • censored data;
  • loglinear;
  • EM algorithm;
  • semiparametric