Imputation-based strategies for clinical trial longitudinal data with nonignorable missing values

Xiaowei Yang, Jinhui Li, Steven Shoptaw

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

Biomedical research is plagued with problems of missing data, especially in clinical trials of medical and behavioral therapies adopting longitudinal design. After a literature review on modeling incomplete longitudinal data based on full-likelihood functions, this paper proposes a set of imputation-based strategies for implementing selection, pattern-mixture, and shared-parameter models for handling intermittent missing values and dropouts that are potentially nonignorable according to various criteria. Within the framework of multiple partial imputation, intermittent missing values are first imputed several times; then, each partially imputed data set is analyzed to deal with dropouts with or without further imputation. Depending on the choice of imputation model or measurement model, there exist various strategies that can be jointly applied to the same set of data to study the effect of treatment or intervention from multi-faceted perspectives. For illustration, the strategies were applied to a data set with continuous repeated measures from a smoking cessation clinical trial.

Original languageEnglish (US)
Pages (from-to)2826-2849
Number of pages24
JournalStatistics in Medicine
Volume27
Issue number15
DOIs
StatePublished - Jul 10 2008

Fingerprint

Missing Values
Imputation
Longitudinal Data
Clinical Trials
Drop out
Likelihood Functions
Smoking Cessation
Repeated Measures
Biomedical Research
Smoking
Literature Review
Incomplete Data
Likelihood Function
Missing Data
Therapy
Model
Partial
Strategy
Datasets
Modeling

Keywords

  • Intermittent missing values
  • Markov transition model
  • Multiple partial imputation
  • Nonignorable dropout
  • Pattern-mixture model
  • Selection model

ASJC Scopus subject areas

  • Epidemiology

Cite this

Imputation-based strategies for clinical trial longitudinal data with nonignorable missing values. / Yang, Xiaowei; Li, Jinhui; Shoptaw, Steven.

In: Statistics in Medicine, Vol. 27, No. 15, 10.07.2008, p. 2826-2849.

Research output: Contribution to journalArticle

Yang, Xiaowei ; Li, Jinhui ; Shoptaw, Steven. / Imputation-based strategies for clinical trial longitudinal data with nonignorable missing values. In: Statistics in Medicine. 2008 ; Vol. 27, No. 15. pp. 2826-2849.
@article{2531b515e64c40c39ef84e722db23ebe,
title = "Imputation-based strategies for clinical trial longitudinal data with nonignorable missing values",
abstract = "Biomedical research is plagued with problems of missing data, especially in clinical trials of medical and behavioral therapies adopting longitudinal design. After a literature review on modeling incomplete longitudinal data based on full-likelihood functions, this paper proposes a set of imputation-based strategies for implementing selection, pattern-mixture, and shared-parameter models for handling intermittent missing values and dropouts that are potentially nonignorable according to various criteria. Within the framework of multiple partial imputation, intermittent missing values are first imputed several times; then, each partially imputed data set is analyzed to deal with dropouts with or without further imputation. Depending on the choice of imputation model or measurement model, there exist various strategies that can be jointly applied to the same set of data to study the effect of treatment or intervention from multi-faceted perspectives. For illustration, the strategies were applied to a data set with continuous repeated measures from a smoking cessation clinical trial.",
keywords = "Intermittent missing values, Markov transition model, Multiple partial imputation, Nonignorable dropout, Pattern-mixture model, Selection model",
author = "Xiaowei Yang and Jinhui Li and Steven Shoptaw",
year = "2008",
month = "7",
day = "10",
doi = "10.1002/sim.3111",
language = "English (US)",
volume = "27",
pages = "2826--2849",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "15",

}

TY - JOUR

T1 - Imputation-based strategies for clinical trial longitudinal data with nonignorable missing values

AU - Yang, Xiaowei

AU - Li, Jinhui

AU - Shoptaw, Steven

PY - 2008/7/10

Y1 - 2008/7/10

N2 - Biomedical research is plagued with problems of missing data, especially in clinical trials of medical and behavioral therapies adopting longitudinal design. After a literature review on modeling incomplete longitudinal data based on full-likelihood functions, this paper proposes a set of imputation-based strategies for implementing selection, pattern-mixture, and shared-parameter models for handling intermittent missing values and dropouts that are potentially nonignorable according to various criteria. Within the framework of multiple partial imputation, intermittent missing values are first imputed several times; then, each partially imputed data set is analyzed to deal with dropouts with or without further imputation. Depending on the choice of imputation model or measurement model, there exist various strategies that can be jointly applied to the same set of data to study the effect of treatment or intervention from multi-faceted perspectives. For illustration, the strategies were applied to a data set with continuous repeated measures from a smoking cessation clinical trial.

AB - Biomedical research is plagued with problems of missing data, especially in clinical trials of medical and behavioral therapies adopting longitudinal design. After a literature review on modeling incomplete longitudinal data based on full-likelihood functions, this paper proposes a set of imputation-based strategies for implementing selection, pattern-mixture, and shared-parameter models for handling intermittent missing values and dropouts that are potentially nonignorable according to various criteria. Within the framework of multiple partial imputation, intermittent missing values are first imputed several times; then, each partially imputed data set is analyzed to deal with dropouts with or without further imputation. Depending on the choice of imputation model or measurement model, there exist various strategies that can be jointly applied to the same set of data to study the effect of treatment or intervention from multi-faceted perspectives. For illustration, the strategies were applied to a data set with continuous repeated measures from a smoking cessation clinical trial.

KW - Intermittent missing values

KW - Markov transition model

KW - Multiple partial imputation

KW - Nonignorable dropout

KW - Pattern-mixture model

KW - Selection model

UR - http://www.scopus.com/inward/record.url?scp=47249109100&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=47249109100&partnerID=8YFLogxK

U2 - 10.1002/sim.3111

DO - 10.1002/sim.3111

M3 - Article

C2 - 18205247

AN - SCOPUS:47249109100

VL - 27

SP - 2826

EP - 2849

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 15

ER -