Bayesian inference in time-varying additive hazards models with applications to disease mapping

A. Chernoukhov, A. Hussein, S. Nkurunziza, D. Bandyopadhyay

Research output: Contribution to journalArticlepeer-review


Environmental health and disease mapping studies are often concerned with the evaluation of the combined effect of various sociodemographic and behavioral factors, and environmental exposures, on time-to-event outcomes of interest, such as death of individuals, organisms, or plants. In such studies, estimation of the hazard function is often of interest. In addition to the known explanatory variables, the hazard function may be subject to spatial/geographical variations, such that proximally located regions may experience hazards similar to those of regions that are distantly located. A popular approach for handling this type of spatially correlated time-to-event data is the Cox proportional hazards regression model with spatial frailties. However, the proportional hazards assumption poses a major practical challenge, as it entails that the effects of the various explanatory variables remain constant over time. This assumption is often unrealistic, for instance, in studies with long follow-ups where the effects of some exposures on the hazard may vary drastically over time. Our goal in this paper is to offer a flexible semiparametric additive hazards model with spatial frailties. Our proposed model allows both the frailties and the regression coefficients to be time varying, thus relaxing the proportionality assumption. Our estimation framework is Bayesian, powered by carefully tailored posterior sampling strategies via Markov chain Monte Carlo techniques. We apply the model to a data set on prostate cancer survival from the U.S. state of Louisiana to illustrate its advantages.

Original languageEnglish (US)
Article numbere2478
Issue number5-6
StatePublished - Aug 1 2018

Bibliographical note

Funding Information:
Natural Sciences and Engineering Research Council of Canada (NSERC); VCU Massey Cancer Center and The National Institutes of Health, Grant/Award Number: R03DE023372 and R01DE024984

Publisher Copyright:
Copyright © 2017 John Wiley & Sons, Ltd.


  • Bayesian
  • Metropolis–Hastings
  • additive hazards
  • conditionally autoregressive prior
  • proposal density
  • prostate cancer
  • spatial


Dive into the research topics of 'Bayesian inference in time-varying additive hazards models with applications to disease mapping'. Together they form a unique fingerprint.

Cite this