A Markov regression random-effects model for remission of functional disability in patients following a first stroke

A Bayesian approach

Shin Liang Pan, Hui Min Wu, Amy Ming Fang Yen, Tony Hsiu Hsi Chen

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

Few attempts have been made to model the dynamics of stroke-related disability. It is possible though, using panel data and multi-state Markov regression models that incorporate measured covariates and latent variables (random effects). This study aimed to model a series of functional transitions (following a first stroke) using a three-state Markov model with or without considering random effects. Several proportional hazards parameterizations were considered. A Bayesian approach that utilizes the Markov Chain Monte Carlo (MCMC) and Gibbs sampling functionality of WinBUGS (a Windows-based Bayesian software package) was developed to generate the marginal posterior distributions of the various transition parameters (e.g. the transition rates and transition probabilities). Model building and comparisons was guided by reference to the deviance information criteria (DIC). Of the four proportional hazards models considered, exponential regression was preferred because it led to the smallest deviances. Adding random effects further improved the model fit. Of the covariates considered, only age, infarct size, and baseline functional status were significant. By using our final model we were able to make individual predictions about functional recovery in stroke patients.

Original languageEnglish
Pages (from-to)5335-5353
Number of pages19
JournalStatistics in Medicine
Volume26
Issue number29
DOIs
Publication statusPublished - Dec 20 2007
Externally publishedYes

Fingerprint

Regression Effects
Bayes Theorem
Random Effects Model
Disability
Stroke
Bayesian Approach
Regression Model
Random Effects
Markov Model
Covariates
Markov Chains
Infarct Size
WinBUGS
Deviance Information Criterion
Deviance
Proportional Hazards Models
Proportional Hazards
Model Comparison
Monte Carlo Sampling
Multi-state

Keywords

  • Activities of daily living
  • Bayes theorem
  • Cerebrovascular accident
  • Markov chains
  • Monte Carlo method
  • Stochastic processes

ASJC Scopus subject areas

  • Epidemiology

Cite this

A Markov regression random-effects model for remission of functional disability in patients following a first stroke : A Bayesian approach. / Pan, Shin Liang; Wu, Hui Min; Yen, Amy Ming Fang; Chen, Tony Hsiu Hsi.

In: Statistics in Medicine, Vol. 26, No. 29, 20.12.2007, p. 5335-5353.

Research output: Contribution to journalArticle

@article{083775ef12cb4269a82426cca56e1b92,
title = "A Markov regression random-effects model for remission of functional disability in patients following a first stroke: A Bayesian approach",
abstract = "Few attempts have been made to model the dynamics of stroke-related disability. It is possible though, using panel data and multi-state Markov regression models that incorporate measured covariates and latent variables (random effects). This study aimed to model a series of functional transitions (following a first stroke) using a three-state Markov model with or without considering random effects. Several proportional hazards parameterizations were considered. A Bayesian approach that utilizes the Markov Chain Monte Carlo (MCMC) and Gibbs sampling functionality of WinBUGS (a Windows-based Bayesian software package) was developed to generate the marginal posterior distributions of the various transition parameters (e.g. the transition rates and transition probabilities). Model building and comparisons was guided by reference to the deviance information criteria (DIC). Of the four proportional hazards models considered, exponential regression was preferred because it led to the smallest deviances. Adding random effects further improved the model fit. Of the covariates considered, only age, infarct size, and baseline functional status were significant. By using our final model we were able to make individual predictions about functional recovery in stroke patients.",
keywords = "Activities of daily living, Bayes theorem, Cerebrovascular accident, Markov chains, Monte Carlo method, Stochastic processes",
author = "Pan, {Shin Liang} and Wu, {Hui Min} and Yen, {Amy Ming Fang} and Chen, {Tony Hsiu Hsi}",
year = "2007",
month = "12",
day = "20",
doi = "10.1002/sim.2999",
language = "English",
volume = "26",
pages = "5335--5353",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "29",

}

TY - JOUR

T1 - A Markov regression random-effects model for remission of functional disability in patients following a first stroke

T2 - A Bayesian approach

AU - Pan, Shin Liang

AU - Wu, Hui Min

AU - Yen, Amy Ming Fang

AU - Chen, Tony Hsiu Hsi

PY - 2007/12/20

Y1 - 2007/12/20

N2 - Few attempts have been made to model the dynamics of stroke-related disability. It is possible though, using panel data and multi-state Markov regression models that incorporate measured covariates and latent variables (random effects). This study aimed to model a series of functional transitions (following a first stroke) using a three-state Markov model with or without considering random effects. Several proportional hazards parameterizations were considered. A Bayesian approach that utilizes the Markov Chain Monte Carlo (MCMC) and Gibbs sampling functionality of WinBUGS (a Windows-based Bayesian software package) was developed to generate the marginal posterior distributions of the various transition parameters (e.g. the transition rates and transition probabilities). Model building and comparisons was guided by reference to the deviance information criteria (DIC). Of the four proportional hazards models considered, exponential regression was preferred because it led to the smallest deviances. Adding random effects further improved the model fit. Of the covariates considered, only age, infarct size, and baseline functional status were significant. By using our final model we were able to make individual predictions about functional recovery in stroke patients.

AB - Few attempts have been made to model the dynamics of stroke-related disability. It is possible though, using panel data and multi-state Markov regression models that incorporate measured covariates and latent variables (random effects). This study aimed to model a series of functional transitions (following a first stroke) using a three-state Markov model with or without considering random effects. Several proportional hazards parameterizations were considered. A Bayesian approach that utilizes the Markov Chain Monte Carlo (MCMC) and Gibbs sampling functionality of WinBUGS (a Windows-based Bayesian software package) was developed to generate the marginal posterior distributions of the various transition parameters (e.g. the transition rates and transition probabilities). Model building and comparisons was guided by reference to the deviance information criteria (DIC). Of the four proportional hazards models considered, exponential regression was preferred because it led to the smallest deviances. Adding random effects further improved the model fit. Of the covariates considered, only age, infarct size, and baseline functional status were significant. By using our final model we were able to make individual predictions about functional recovery in stroke patients.

KW - Activities of daily living

KW - Bayes theorem

KW - Cerebrovascular accident

KW - Markov chains

KW - Monte Carlo method

KW - Stochastic processes

UR - http://www.scopus.com/inward/record.url?scp=37349040368&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=37349040368&partnerID=8YFLogxK

U2 - 10.1002/sim.2999

DO - 10.1002/sim.2999

M3 - Article

VL - 26

SP - 5335

EP - 5353

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 29

ER -