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

The analysis of correlated panel data using a continuous-time Markov model

E W Lee1, M Y Kim

  • 1Epidemiology and Biostatistics Program, Institute of Environmental Medicine, New York University Medical Center, New York 10010, USA.

Biometrics
|January 12, 1999
PubMed
Summary

This study introduces a new statistical method for analyzing correlated health data over time when exact event timings are unknown. The approach enables robust comparison of treatment effects in clinical trials.

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

Immunogenicity and protective potential of a mucosal protein-only vaccine candidate for tuberculosis.

Frontiers in immunology·2026
Same author

A Case of Injury of the Spine-Enormous Accumulation of Intestinal Gas-Evacuation with Aspiration-Recovery.

The Chicago medical journal and examiner·2023
Same author

On the Best Means of Promoting Union by First Intention.

The Chicago medical journal and examiner·2023
Same author

Report of Seven Cases of Tracheotomy.

The Chicago medical journal and examiner·2023
Same author

Tracheotomy in Croup and Diphtheria, with Cases.

The Chicago medical journal and examiner·2023
Same author

Select Topics of Modern Surgery, Illustrated by Cases from Hospital Service and Private Practice.

The Chicago medical journal and examiner·2023

Area of Science:

  • Biostatistics
  • Clinical Trials
  • Epidemiology

Background:

  • Analyzing correlated multistate processes in panel data is challenging, especially when exact transition times are unknown.
  • Existing methods may not adequately handle the complexity of multiple, interdependent health outcomes observed periodically.

Purpose of the Study:

  • To develop and validate a statistical procedure for analyzing correlated panel data with unknown transition times.
  • To enable marginal modeling of multiple Markov processes incorporating covariates.
  • To facilitate simultaneous inference for comparing treatment effects.

Main Methods:

  • Utilized time-homogeneous Markov models for marginal analysis of each correlated process.
  • Developed estimators for asymptotic joint normality and consistent covariance matrix estimation.

Related Experiment Videos

  • Applied the methods to analyze toxic effects in an AIDS clinical trial.

    • Main Results:

    • The proposed estimators are asymptotically jointly normal.
    • A consistent method for estimating the covariance matrix was established.
    • Simultaneous inference procedures were successfully developed and applied.

    Conclusions:

    • The new statistical procedure provides a robust framework for analyzing complex correlated panel data.
    • This method is effective for comparing treatment effects on multiple health outcomes, as demonstrated in an AIDS clinical trial setting.
    • The approach allows for covariate adjustment and simultaneous inference, enhancing the analysis of longitudinal health data.