Survival analysis with time-dependent covariates subject to missing data or measurement error: Multiple Imputation for Joint Modeling (MIJM)

Margarita Moreno-Betancur, John B. Carlin, Samuel L. Brilleman, Stephanie K. Tanamas, Anna Peeters, Rory Wolfe

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Modern epidemiological studies collect data on time-varying individual-specific characteristics, such as body mass index and blood pressure. Incorporation of such time-dependent covariates in time-to-event models is of great interest, but raises some challenges. Of specific concern are measurement error, and the non-synchronous updating of covariates across individuals, due for example to missing data. It is well known that in the presence of either of these issues the last observation carried forward (LOCF) approach traditionally used leads to bias. Joint models of longitudinal and time-to-event outcomes, developed recently, address these complexities by specifying a model for the joint distribution of all processes and are commonly fitted by maximum likelihood or Bayesian approaches. However, the adequate specification of the full joint distribution can be a challenging modeling task, especially with multiple longitudinal markers. In fact, most available software packages are unable to handle more than one marker and offer a restricted choice of survival models. We propose a two-stage approach, Multiple Imputation for Joint Modeling (MIJM), to incorporate multiple time-dependent continuous covariates in the semi-parametric Cox and additive hazard models.Assuming a primary focus on the time-to-event model, theMIJMapproach handles the joint distribution of the markers using multiple imputation by chained equations, a computationally convenient procedure that is widely available in mainstream statistical software. We developed an Rpackage survtd that allows MIJM and other approaches in this manuscript to be applied easily, with just one call to its main function.A simulation study showed that MIJM performs well across a wide range of scenarios in terms of bias and coverage probability, particularly compared with LOCF, simpler two-stage approaches, and a Bayesian joint model. The Framingham Heart Study is used to illustrate the approach.

Original languageEnglish
Pages (from-to)479-496
Number of pages18
JournalBiostatistics
Volume19
Issue number4
DOIs
Publication statusPublished - 1 Oct 2018

Keywords

  • Additive hazard model
  • Cox model
  • Joint model
  • Measurement error
  • Missing data
  • Multiple imputation
  • Multiple overimputation
  • Regression calibration
  • Time-dependent covariates.

Cite this

@article{0a9e0b405dfc460b83ce2f9817313898,
title = "Survival analysis with time-dependent covariates subject to missing data or measurement error: Multiple Imputation for Joint Modeling (MIJM)",
abstract = "Modern epidemiological studies collect data on time-varying individual-specific characteristics, such as body mass index and blood pressure. Incorporation of such time-dependent covariates in time-to-event models is of great interest, but raises some challenges. Of specific concern are measurement error, and the non-synchronous updating of covariates across individuals, due for example to missing data. It is well known that in the presence of either of these issues the last observation carried forward (LOCF) approach traditionally used leads to bias. Joint models of longitudinal and time-to-event outcomes, developed recently, address these complexities by specifying a model for the joint distribution of all processes and are commonly fitted by maximum likelihood or Bayesian approaches. However, the adequate specification of the full joint distribution can be a challenging modeling task, especially with multiple longitudinal markers. In fact, most available software packages are unable to handle more than one marker and offer a restricted choice of survival models. We propose a two-stage approach, Multiple Imputation for Joint Modeling (MIJM), to incorporate multiple time-dependent continuous covariates in the semi-parametric Cox and additive hazard models.Assuming a primary focus on the time-to-event model, theMIJMapproach handles the joint distribution of the markers using multiple imputation by chained equations, a computationally convenient procedure that is widely available in mainstream statistical software. We developed an Rpackage survtd that allows MIJM and other approaches in this manuscript to be applied easily, with just one call to its main function.A simulation study showed that MIJM performs well across a wide range of scenarios in terms of bias and coverage probability, particularly compared with LOCF, simpler two-stage approaches, and a Bayesian joint model. The Framingham Heart Study is used to illustrate the approach.",
keywords = "Additive hazard model, Cox model, Joint model, Measurement error, Missing data, Multiple imputation, Multiple overimputation, Regression calibration, Time-dependent covariates.",
author = "Margarita Moreno-Betancur and Carlin, {John B.} and Brilleman, {Samuel L.} and Tanamas, {Stephanie K.} and Anna Peeters and Rory Wolfe",
year = "2018",
month = "10",
day = "1",
doi = "10.1093/biostatistics/kxx046",
language = "English",
volume = "19",
pages = "479--496",
journal = "Biostatistics",
issn = "1465-4644",
publisher = "Oxford University Press",
number = "4",

}

Survival analysis with time-dependent covariates subject to missing data or measurement error : Multiple Imputation for Joint Modeling (MIJM). / Moreno-Betancur, Margarita; Carlin, John B.; Brilleman, Samuel L.; Tanamas, Stephanie K.; Peeters, Anna; Wolfe, Rory.

In: Biostatistics, Vol. 19, No. 4, 01.10.2018, p. 479-496.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Survival analysis with time-dependent covariates subject to missing data or measurement error

T2 - Multiple Imputation for Joint Modeling (MIJM)

AU - Moreno-Betancur, Margarita

AU - Carlin, John B.

AU - Brilleman, Samuel L.

AU - Tanamas, Stephanie K.

AU - Peeters, Anna

AU - Wolfe, Rory

PY - 2018/10/1

Y1 - 2018/10/1

N2 - Modern epidemiological studies collect data on time-varying individual-specific characteristics, such as body mass index and blood pressure. Incorporation of such time-dependent covariates in time-to-event models is of great interest, but raises some challenges. Of specific concern are measurement error, and the non-synchronous updating of covariates across individuals, due for example to missing data. It is well known that in the presence of either of these issues the last observation carried forward (LOCF) approach traditionally used leads to bias. Joint models of longitudinal and time-to-event outcomes, developed recently, address these complexities by specifying a model for the joint distribution of all processes and are commonly fitted by maximum likelihood or Bayesian approaches. However, the adequate specification of the full joint distribution can be a challenging modeling task, especially with multiple longitudinal markers. In fact, most available software packages are unable to handle more than one marker and offer a restricted choice of survival models. We propose a two-stage approach, Multiple Imputation for Joint Modeling (MIJM), to incorporate multiple time-dependent continuous covariates in the semi-parametric Cox and additive hazard models.Assuming a primary focus on the time-to-event model, theMIJMapproach handles the joint distribution of the markers using multiple imputation by chained equations, a computationally convenient procedure that is widely available in mainstream statistical software. We developed an Rpackage survtd that allows MIJM and other approaches in this manuscript to be applied easily, with just one call to its main function.A simulation study showed that MIJM performs well across a wide range of scenarios in terms of bias and coverage probability, particularly compared with LOCF, simpler two-stage approaches, and a Bayesian joint model. The Framingham Heart Study is used to illustrate the approach.

AB - Modern epidemiological studies collect data on time-varying individual-specific characteristics, such as body mass index and blood pressure. Incorporation of such time-dependent covariates in time-to-event models is of great interest, but raises some challenges. Of specific concern are measurement error, and the non-synchronous updating of covariates across individuals, due for example to missing data. It is well known that in the presence of either of these issues the last observation carried forward (LOCF) approach traditionally used leads to bias. Joint models of longitudinal and time-to-event outcomes, developed recently, address these complexities by specifying a model for the joint distribution of all processes and are commonly fitted by maximum likelihood or Bayesian approaches. However, the adequate specification of the full joint distribution can be a challenging modeling task, especially with multiple longitudinal markers. In fact, most available software packages are unable to handle more than one marker and offer a restricted choice of survival models. We propose a two-stage approach, Multiple Imputation for Joint Modeling (MIJM), to incorporate multiple time-dependent continuous covariates in the semi-parametric Cox and additive hazard models.Assuming a primary focus on the time-to-event model, theMIJMapproach handles the joint distribution of the markers using multiple imputation by chained equations, a computationally convenient procedure that is widely available in mainstream statistical software. We developed an Rpackage survtd that allows MIJM and other approaches in this manuscript to be applied easily, with just one call to its main function.A simulation study showed that MIJM performs well across a wide range of scenarios in terms of bias and coverage probability, particularly compared with LOCF, simpler two-stage approaches, and a Bayesian joint model. The Framingham Heart Study is used to illustrate the approach.

KW - Additive hazard model

KW - Cox model

KW - Joint model

KW - Measurement error

KW - Missing data

KW - Multiple imputation

KW - Multiple overimputation

KW - Regression calibration

KW - Time-dependent covariates.

UR - http://www.scopus.com/inward/record.url?scp=85054753738&partnerID=8YFLogxK

U2 - 10.1093/biostatistics/kxx046

DO - 10.1093/biostatistics/kxx046

M3 - Article

VL - 19

SP - 479

EP - 496

JO - Biostatistics

JF - Biostatistics

SN - 1465-4644

IS - 4

ER -