Using generalized estimating equations and extensions in randomized trials with missing longitudinal patient reported outcome data

Melanie L Bell, Nicholas J. Horton, Haryana M. Dhillon, Victoria J. Bray, Janette Vardy

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Objective: Patient reported outcomes (PROs) are important in oncology research; however, missing data can pose a threat to the validity of results. Psycho-oncology researchers should be aware of the statistical options for handling missing data robustly. One rarely used set of methods, which includes extensions for handling missing data, is generalized estimating equations (GEEs). Our objective was to demonstrate use of GEEs to analyze PROs with missing data in randomized trials with assessments at fixed time points. Methods: We introduce GEEs and show, with a worked example, how to use GEEs that account for missing data: inverse probability weighted GEEs and multiple imputation with GEE. We use data from an RCT evaluating a web-based brain training for cancer survivors reporting cognitive symptoms after chemotherapy treatment. The primary outcome for this demonstration is the binary outcome of cognitive impairment. Several methods are used, and results are compared. Results: We demonstrate that estimates can vary depending on the choice of analytical approach, with odds ratios for no cognitive impairment ranging from 2.04 to 5.74. While most of these estimates were statistically significant (P < 0.05), a few were not. Conclusions: Researchers using PROs should use statistical methods that handle missing data in a way as to result in unbiased estimates. GEE extensions are analytic options for handling dropouts in longitudinal RCTs, particularly if the outcome is not continuous.

Original languageEnglish (US)
Pages (from-to)2125-2131
Number of pages7
JournalPsycho-Oncology
Volume27
Issue number9
DOIs
StatePublished - Sep 1 2018

Fingerprint

Research Personnel
Neurobehavioral Manifestations
Reproducibility of Results
Brain Neoplasms
Survivors
Odds Ratio
Drug Therapy
Patient Reported Outcome Measures
Research
Cognitive Dysfunction
Therapeutics

Keywords

  • cancer
  • dropout
  • inverse probability weighting
  • missing data
  • multiple imputation
  • oncology
  • statistical methods

ASJC Scopus subject areas

  • Experimental and Cognitive Psychology
  • Oncology
  • Psychiatry and Mental health

Cite this

Using generalized estimating equations and extensions in randomized trials with missing longitudinal patient reported outcome data. / Bell, Melanie L; Horton, Nicholas J.; Dhillon, Haryana M.; Bray, Victoria J.; Vardy, Janette.

In: Psycho-Oncology, Vol. 27, No. 9, 01.09.2018, p. 2125-2131.

Research output: Contribution to journalArticle

Bell, Melanie L ; Horton, Nicholas J. ; Dhillon, Haryana M. ; Bray, Victoria J. ; Vardy, Janette. / Using generalized estimating equations and extensions in randomized trials with missing longitudinal patient reported outcome data. In: Psycho-Oncology. 2018 ; Vol. 27, No. 9. pp. 2125-2131.
@article{b64694ed2f954fca8f5a9895228e63e8,
title = "Using generalized estimating equations and extensions in randomized trials with missing longitudinal patient reported outcome data",
abstract = "Objective: Patient reported outcomes (PROs) are important in oncology research; however, missing data can pose a threat to the validity of results. Psycho-oncology researchers should be aware of the statistical options for handling missing data robustly. One rarely used set of methods, which includes extensions for handling missing data, is generalized estimating equations (GEEs). Our objective was to demonstrate use of GEEs to analyze PROs with missing data in randomized trials with assessments at fixed time points. Methods: We introduce GEEs and show, with a worked example, how to use GEEs that account for missing data: inverse probability weighted GEEs and multiple imputation with GEE. We use data from an RCT evaluating a web-based brain training for cancer survivors reporting cognitive symptoms after chemotherapy treatment. The primary outcome for this demonstration is the binary outcome of cognitive impairment. Several methods are used, and results are compared. Results: We demonstrate that estimates can vary depending on the choice of analytical approach, with odds ratios for no cognitive impairment ranging from 2.04 to 5.74. While most of these estimates were statistically significant (P < 0.05), a few were not. Conclusions: Researchers using PROs should use statistical methods that handle missing data in a way as to result in unbiased estimates. GEE extensions are analytic options for handling dropouts in longitudinal RCTs, particularly if the outcome is not continuous.",
keywords = "cancer, dropout, inverse probability weighting, missing data, multiple imputation, oncology, statistical methods",
author = "Bell, {Melanie L} and Horton, {Nicholas J.} and Dhillon, {Haryana M.} and Bray, {Victoria J.} and Janette Vardy",
year = "2018",
month = "9",
day = "1",
doi = "10.1002/pon.4777",
language = "English (US)",
volume = "27",
pages = "2125--2131",
journal = "Psycho-Oncology",
issn = "1057-9249",
publisher = "John Wiley and Sons Ltd",
number = "9",

}

TY - JOUR

T1 - Using generalized estimating equations and extensions in randomized trials with missing longitudinal patient reported outcome data

AU - Bell, Melanie L

AU - Horton, Nicholas J.

AU - Dhillon, Haryana M.

AU - Bray, Victoria J.

AU - Vardy, Janette

PY - 2018/9/1

Y1 - 2018/9/1

N2 - Objective: Patient reported outcomes (PROs) are important in oncology research; however, missing data can pose a threat to the validity of results. Psycho-oncology researchers should be aware of the statistical options for handling missing data robustly. One rarely used set of methods, which includes extensions for handling missing data, is generalized estimating equations (GEEs). Our objective was to demonstrate use of GEEs to analyze PROs with missing data in randomized trials with assessments at fixed time points. Methods: We introduce GEEs and show, with a worked example, how to use GEEs that account for missing data: inverse probability weighted GEEs and multiple imputation with GEE. We use data from an RCT evaluating a web-based brain training for cancer survivors reporting cognitive symptoms after chemotherapy treatment. The primary outcome for this demonstration is the binary outcome of cognitive impairment. Several methods are used, and results are compared. Results: We demonstrate that estimates can vary depending on the choice of analytical approach, with odds ratios for no cognitive impairment ranging from 2.04 to 5.74. While most of these estimates were statistically significant (P < 0.05), a few were not. Conclusions: Researchers using PROs should use statistical methods that handle missing data in a way as to result in unbiased estimates. GEE extensions are analytic options for handling dropouts in longitudinal RCTs, particularly if the outcome is not continuous.

AB - Objective: Patient reported outcomes (PROs) are important in oncology research; however, missing data can pose a threat to the validity of results. Psycho-oncology researchers should be aware of the statistical options for handling missing data robustly. One rarely used set of methods, which includes extensions for handling missing data, is generalized estimating equations (GEEs). Our objective was to demonstrate use of GEEs to analyze PROs with missing data in randomized trials with assessments at fixed time points. Methods: We introduce GEEs and show, with a worked example, how to use GEEs that account for missing data: inverse probability weighted GEEs and multiple imputation with GEE. We use data from an RCT evaluating a web-based brain training for cancer survivors reporting cognitive symptoms after chemotherapy treatment. The primary outcome for this demonstration is the binary outcome of cognitive impairment. Several methods are used, and results are compared. Results: We demonstrate that estimates can vary depending on the choice of analytical approach, with odds ratios for no cognitive impairment ranging from 2.04 to 5.74. While most of these estimates were statistically significant (P < 0.05), a few were not. Conclusions: Researchers using PROs should use statistical methods that handle missing data in a way as to result in unbiased estimates. GEE extensions are analytic options for handling dropouts in longitudinal RCTs, particularly if the outcome is not continuous.

KW - cancer

KW - dropout

KW - inverse probability weighting

KW - missing data

KW - multiple imputation

KW - oncology

KW - statistical methods

UR - http://www.scopus.com/inward/record.url?scp=85053902972&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85053902972&partnerID=8YFLogxK

U2 - 10.1002/pon.4777

DO - 10.1002/pon.4777

M3 - Article

VL - 27

SP - 2125

EP - 2131

JO - Psycho-Oncology

JF - Psycho-Oncology

SN - 1057-9249

IS - 9

ER -