Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Analysis of longitudinal binary data with missing data due to dropouts.

Mirza W Ali1, Enayet Talukder

  • 1Department of Biostatistics, Otsuka Maryland Research Institute, Rockville, Maryland 20850, USA.

Journal of Biopharmaceutical Statistics
|November 11, 2005
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Analysis of Response Data for Assessing Treatment Effects in Comparative Clinical Studies.

Annals of internal medicine·2020
Same author

Quantitative decision-making in randomized Phase II studies with a time-to-event endpoint.

Journal of biopharmaceutical statistics·2018
Same author

Evaluating Treatment Effect Based on Duration of Response for a Comparative Oncology Study.

JAMA oncology·2018
Same author

Novel dose-finding designs and considerations on practical implementations in oncology clinical trials.

Journal of biopharmaceutical statistics·2016
Same author

A multiple imputation method for sensitivity analyses of time-to-event data with possibly informative censoring.

Journal of biopharmaceutical statistics·2014
Same author

Switching patients to aripiprazole from other antipsychotic agents: a multicenter randomized study.

Psychopharmacology·2003

Handling missing data in longitudinal clinical trials is crucial. This study compares methods like Generalized Linear Mixture Models and Generalized Estimating Equations to address bias from missing observations, especially for NMAR dropouts.

Area of Science:

  • Biostatistics
  • Clinical Trial Methodology
  • Longitudinal Data Analysis

Background:

  • Longitudinal binary data in clinical trials often have missing observations, impacting analysis.
  • Traditional methods like Last Observation Carry Forward (LOCF) and Observed Cases (OC) can introduce bias.
  • Missing data mechanisms, particularly Not Missing At Random (NMAR) dropouts, pose significant challenges.

Purpose of the Study:

  • To evaluate and compare various statistical methods for handling missing data in longitudinal binary trial data.
  • To assess the sensitivity of results to different missing data assumptions and imputation methods.
  • To demonstrate the application of advanced techniques like Generalized Linear Mixture Models (GLMM) and Weighted Generalized Estimating Equations (GEE).

Main Methods:

Related Experiment Videos

  • Application of Generalized Linear Mixture Models (GLMM) for NMAR dropouts.
  • Utilization of Weighted Generalized Estimating Equations (GEE) for Missing At Random (MAR) dropouts.
  • Comparison with standard methods: Generalized Estimating Equations (GEE) for Missing Completely At Random (MCAR) dropouts, logistic regression on LOCF and OC data.

Main Results:

  • Different methods yield varying results, highlighting the impact of missing data mechanisms.
  • GLMM and Weighted GEE demonstrate potential for more robust analysis in the presence of MAR/NMAR data.
  • Sensitivity analyses are essential to understand the influence of chosen imputation strategies.

Conclusions:

  • No single method universally handles all missing data scenarios in longitudinal binary trials.
  • Employing multiple methods, including those addressing NMAR dropouts, is recommended for robust conclusions.
  • Advanced techniques like GLMM and Weighted GEE offer improved approaches for complex missing data patterns.