Estimation of multi-state models with missing covariate values based on observed data likelihood

Wenjie Lou, Erin L. Abner, Lijie Wan, David W. Fardo, Richard B. Lipton, Mindy Joy Katz, Richard J. Kryscio

Research output: Contribution to journalArticle

Abstract

Continuous-time multi-state models are commonly used to study diseases with multiple stages. Potential risk factors associated with the disease are added to the transition intensities of the model as covariates, but missing covariate measurements arise frequently in practice. We propose a likelihood-based method that deals efficiently with a missing covariate in these models. Our simulation study showed that the method performs well for both “missing completely at random” and “missing at random” mechanisms. We also applied our method to a real dataset, the Einstein Aging Study.

Original languageEnglish (US)
JournalCommunications in Statistics - Theory and Methods
DOIs
StateAccepted/In press - Jan 1 2018

Fingerprint

Multi-state Model
Missing Covariates
Likelihood
Missing Completely at Random
Missing at Random
Continuous-time Model
Risk Factors
Albert Einstein
Covariates
Simulation Study
Model

Keywords

  • Longitudinal data
  • MAR
  • MCAR
  • missing covariate
  • multi-state model

ASJC Scopus subject areas

  • Statistics and Probability

Cite this

Estimation of multi-state models with missing covariate values based on observed data likelihood. / Lou, Wenjie; Abner, Erin L.; Wan, Lijie; Fardo, David W.; Lipton, Richard B.; Katz, Mindy Joy; Kryscio, Richard J.

In: Communications in Statistics - Theory and Methods, 01.01.2018.

Research output: Contribution to journalArticle

@article{d14840f792284d2da453a97c558b37d5,
title = "Estimation of multi-state models with missing covariate values based on observed data likelihood",
abstract = "Continuous-time multi-state models are commonly used to study diseases with multiple stages. Potential risk factors associated with the disease are added to the transition intensities of the model as covariates, but missing covariate measurements arise frequently in practice. We propose a likelihood-based method that deals efficiently with a missing covariate in these models. Our simulation study showed that the method performs well for both “missing completely at random” and “missing at random” mechanisms. We also applied our method to a real dataset, the Einstein Aging Study.",
keywords = "Longitudinal data, MAR, MCAR, missing covariate, multi-state model",
author = "Wenjie Lou and Abner, {Erin L.} and Lijie Wan and Fardo, {David W.} and Lipton, {Richard B.} and Katz, {Mindy Joy} and Kryscio, {Richard J.}",
year = "2018",
month = "1",
day = "1",
doi = "10.1080/03610926.2018.1520884",
language = "English (US)",
journal = "Communications in Statistics - Theory and Methods",
issn = "0361-0926",
publisher = "Taylor and Francis Ltd.",

}

TY - JOUR

T1 - Estimation of multi-state models with missing covariate values based on observed data likelihood

AU - Lou, Wenjie

AU - Abner, Erin L.

AU - Wan, Lijie

AU - Fardo, David W.

AU - Lipton, Richard B.

AU - Katz, Mindy Joy

AU - Kryscio, Richard J.

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Continuous-time multi-state models are commonly used to study diseases with multiple stages. Potential risk factors associated with the disease are added to the transition intensities of the model as covariates, but missing covariate measurements arise frequently in practice. We propose a likelihood-based method that deals efficiently with a missing covariate in these models. Our simulation study showed that the method performs well for both “missing completely at random” and “missing at random” mechanisms. We also applied our method to a real dataset, the Einstein Aging Study.

AB - Continuous-time multi-state models are commonly used to study diseases with multiple stages. Potential risk factors associated with the disease are added to the transition intensities of the model as covariates, but missing covariate measurements arise frequently in practice. We propose a likelihood-based method that deals efficiently with a missing covariate in these models. Our simulation study showed that the method performs well for both “missing completely at random” and “missing at random” mechanisms. We also applied our method to a real dataset, the Einstein Aging Study.

KW - Longitudinal data

KW - MAR

KW - MCAR

KW - missing covariate

KW - multi-state model

UR - http://www.scopus.com/inward/record.url?scp=85059297150&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059297150&partnerID=8YFLogxK

U2 - 10.1080/03610926.2018.1520884

DO - 10.1080/03610926.2018.1520884

M3 - Article

JO - Communications in Statistics - Theory and Methods

JF - Communications in Statistics - Theory and Methods

SN - 0361-0926

ER -