A pattern-mixture model approach for handling missing continuous outcome data in longitudinal cluster randomized trials

Mallorie H. Fiero, Chiu Hsieh Hsu, Melanie L. Bell

Research output: Research - peer-reviewArticle

Abstract

We extend the pattern-mixture approach to handle missing continuous outcome data in longitudinal cluster randomized trials, which randomize groups of individuals to treatment arms, rather than the individuals themselves. Individuals who drop out at the same time point are grouped into the same dropout pattern. We approach extrapolation of the pattern-mixture model by applying multilevel multiple imputation, which imputes missing values while appropriately accounting for the hierarchical data structure found in cluster randomized trials. To assess parameters of interest under various missing data assumptions, imputed values are multiplied by a sensitivity parameter, k, which increases or decreases imputed values. Using simulated data, we show that estimates of parameters of interest can vary widely under differing missing data assumptions. We conduct a sensitivity analysis using real data from a cluster randomized trial by increasing k until the treatment effect inference changes. By performing a sensitivity analysis for missing data, researchers can assess whether certain missing data assumptions are reasonable for their cluster randomized trial.

LanguageEnglish (US)
Pages4094-4105
Number of pages12
JournalStatistics in Medicine
Volume36
Issue number26
DOIs
StatePublished - Nov 20 2017

Fingerprint

Pattern-mixture Model
Randomized Trial
Missing Data
Research Personnel
Drop out
Sensitivity Analysis
Hierarchical Data
Parameter Sensitivity
Multiple Imputation
Missing Values
Treatment Effects
Hierarchical Structure
Extrapolation
Data Structures
Vary
Decrease
Estimate

Keywords

  • cluster randomized trials
  • missing data
  • multiple imputation
  • pattern-mixture model

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Cite this

A pattern-mixture model approach for handling missing continuous outcome data in longitudinal cluster randomized trials. / Fiero, Mallorie H.; Hsu, Chiu Hsieh; Bell, Melanie L.

In: Statistics in Medicine, Vol. 36, No. 26, 20.11.2017, p. 4094-4105.

Research output: Research - peer-reviewArticle

@article{d0e4aaf14866431d846184e85a9eb3bc,
title = "A pattern-mixture model approach for handling missing continuous outcome data in longitudinal cluster randomized trials",
abstract = "We extend the pattern-mixture approach to handle missing continuous outcome data in longitudinal cluster randomized trials, which randomize groups of individuals to treatment arms, rather than the individuals themselves. Individuals who drop out at the same time point are grouped into the same dropout pattern. We approach extrapolation of the pattern-mixture model by applying multilevel multiple imputation, which imputes missing values while appropriately accounting for the hierarchical data structure found in cluster randomized trials. To assess parameters of interest under various missing data assumptions, imputed values are multiplied by a sensitivity parameter, k, which increases or decreases imputed values. Using simulated data, we show that estimates of parameters of interest can vary widely under differing missing data assumptions. We conduct a sensitivity analysis using real data from a cluster randomized trial by increasing k until the treatment effect inference changes. By performing a sensitivity analysis for missing data, researchers can assess whether certain missing data assumptions are reasonable for their cluster randomized trial.",
keywords = "cluster randomized trials, missing data, multiple imputation, pattern-mixture model",
author = "Fiero, {Mallorie H.} and Hsu, {Chiu Hsieh} and Bell, {Melanie L.}",
year = "2017",
month = "11",
doi = "10.1002/sim.7418",
volume = "36",
pages = "4094--4105",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "26",

}

TY - JOUR

T1 - A pattern-mixture model approach for handling missing continuous outcome data in longitudinal cluster randomized trials

AU - Fiero,Mallorie H.

AU - Hsu,Chiu Hsieh

AU - Bell,Melanie L.

PY - 2017/11/20

Y1 - 2017/11/20

N2 - We extend the pattern-mixture approach to handle missing continuous outcome data in longitudinal cluster randomized trials, which randomize groups of individuals to treatment arms, rather than the individuals themselves. Individuals who drop out at the same time point are grouped into the same dropout pattern. We approach extrapolation of the pattern-mixture model by applying multilevel multiple imputation, which imputes missing values while appropriately accounting for the hierarchical data structure found in cluster randomized trials. To assess parameters of interest under various missing data assumptions, imputed values are multiplied by a sensitivity parameter, k, which increases or decreases imputed values. Using simulated data, we show that estimates of parameters of interest can vary widely under differing missing data assumptions. We conduct a sensitivity analysis using real data from a cluster randomized trial by increasing k until the treatment effect inference changes. By performing a sensitivity analysis for missing data, researchers can assess whether certain missing data assumptions are reasonable for their cluster randomized trial.

AB - We extend the pattern-mixture approach to handle missing continuous outcome data in longitudinal cluster randomized trials, which randomize groups of individuals to treatment arms, rather than the individuals themselves. Individuals who drop out at the same time point are grouped into the same dropout pattern. We approach extrapolation of the pattern-mixture model by applying multilevel multiple imputation, which imputes missing values while appropriately accounting for the hierarchical data structure found in cluster randomized trials. To assess parameters of interest under various missing data assumptions, imputed values are multiplied by a sensitivity parameter, k, which increases or decreases imputed values. Using simulated data, we show that estimates of parameters of interest can vary widely under differing missing data assumptions. We conduct a sensitivity analysis using real data from a cluster randomized trial by increasing k until the treatment effect inference changes. By performing a sensitivity analysis for missing data, researchers can assess whether certain missing data assumptions are reasonable for their cluster randomized trial.

KW - cluster randomized trials

KW - missing data

KW - multiple imputation

KW - pattern-mixture model

UR - http://www.scopus.com/inward/record.url?scp=85030315485&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85030315485&partnerID=8YFLogxK

U2 - 10.1002/sim.7418

DO - 10.1002/sim.7418

M3 - Article

VL - 36

SP - 4094

EP - 4105

JO - Statistics in Medicine

T2 - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 26

ER -