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

Pseudo cluster randomization: a treatment allocation method to minimize contamination and selection bias.

George F Borm1, René J F Melis, Steven Teerenstra

  • 1Department of Epidemiology and Biostatistics, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands. g.borm@epib.umcn.nl

Statistics in Medicine
|July 12, 2005
PubMed
Summary

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

Patient-led, home-based follow-up for colorectal cancer: the DISTANCE multicentre stepped-wedge cluster-randomised trial.

The British journal of surgery·2026
Same author

Interpreting higher-order dependence in multimorbidity using cohort data: A partial information decomposition approach.

PLoS computational biology·2026
Same author

Trajectories of social health, cognitive, and daily functioning in community-dwelling older adults.

Innovation in aging·2026
Same author

"You can't live for yourself, can you? That's just boring": perspectives on social health as a relational concept as described by community-dwelling older adults in a large Dutch city.

The Gerontologist·2026
Same author

How does social health impact cognitive function and brain reserve? Findings from the SHARED Consortium.

Ageing research reviews·2026
Same author

Understanding hand grip measures in geriatric inpatients: associations with frailty, daily functioning and fatigue.

Aging clinical and experimental research·2025

Pseudo cluster randomization offers a solution for clinical trials where patient-level randomization is difficult. This method improves efficiency when treatment contamination is a concern, outperforming traditional patient or cluster randomization.

Area of Science:

  • Clinical Trials Methodology
  • Biostatistics
  • Public Health Research

Background:

  • Patient-level randomization is not always feasible in clinical trials.
  • Contamination of treatment effects can occur when patients interact, biasing results.
  • Cluster randomization (group-level) can lead to reduced statistical efficiency and recruitment bias.

Purpose of the Study:

  • To introduce and describe pseudo cluster randomization as an alternative method.
  • To evaluate the statistical properties and efficiency of pseudo cluster randomization.
  • To compare pseudo cluster randomization with patient-level and cluster-level randomization.

Main Methods:

  • Pseudo cluster randomization assigns treatments to clusters, with a majority of subjects in each cluster receiving one treatment and a minority receiving another.

Related Experiment Videos

  • Two groups of clusters are formed, with reversed treatment proportions.
  • Statistical properties of this novel randomization method are analyzed.

    • Main Results:

    • Pseudo cluster randomization mitigates contamination bias.
    • The method demonstrates improved statistical efficiency compared to patient-level or cluster-level randomization when contamination is present.
    • It addresses recruitment bias by introducing some unpredictability in treatment assignment within clusters.

    Conclusions:

    • Pseudo cluster randomization is a viable and efficient alternative for clinical trials facing contamination issues.
    • This method enhances the integrity and statistical power of studies where group-level randomization is necessary.
    • It offers a practical solution for complex trial designs.