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

Randomizing patients by family practice: sample size estimation, intracluster correlation and data analysis.

Roxanne H Cosby1, Michelle Howard, Janusz Kaczorowski

  • 1Department of Family Medicine, McMaster University and Centre for the Evaluation of Medicines, St. Joseph's Healthcare, Hamilton, Ontario, Canada. cosbyr@mcmaster.ca

Family Practice
|January 2, 2003
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

Development of a Canadian emergency department checklist to improve care for people who use opioids: a modified Delphi study.

CJEM·2026
Same author

Physician perspectives on reducing harm and supporting emergency department patients who use drugs.

PloS one·2025
Same author

Causal mediation analysis: what is it and how can it be used to inform practice and policy?

Family practice·2025
Same author

Removable Boot vs Casting of Toddler's Fractures: A Randomized Clinical Trial.

JAMA pediatrics·2025
Same author

Patient preferences for key organizational features of primary cardiovascular care in Quebec: a discrete choice experiment.

BMC primary care·2025
Same author

Changes in local access to mifepristone dispensed by community pharmacies for medication abortion in Ontario: a population-based repeated cross-sectional study.

CMAJ : Canadian Medical Association journal = journal de l'Association medicale canadienne·2025

Cluster randomization in health interventions significantly increases required sample sizes compared to individual randomization. Accurate intraclass correlation coefficients (ICCs) are crucial for proper study design and power calculations.

Area of Science:

  • Biostatistics
  • Health Services Research
  • Clinical Trials

Background:

  • Cluster randomized controlled trials (CRCTs) are increasingly used for health interventions.
  • Patients within clusters (e.g., practices, hospitals) exhibit similarities, necessitating specialized statistical methods for sample size and analysis.
  • Understanding intraclass correlation coefficients (ICCs) is vital for accurate CRCT design.

Purpose of the Study:

  • To illustrate sample size calculations and data analysis methods for CRCTs.
  • To provide ICC estimates for key variables from the Seniors Medication Assessment Research Trial (SMART).
  • To inform the design of community-based trials optimizing drug therapy in older patients.

Main Methods:

  • A paired cluster randomized trial design was employed, with family physician practices as clusters.

Related Experiment Videos

  • Sample size calculation was based on a 15% medication reduction, 80% power, 0.05 alpha, and an ICC of 0.08.
  • Intraclass correlation coefficients (ICCs) were estimated for various outcomes; analyses used a random effects model.

    • Main Results:

    • The design effect due to clustering was 2.12, inflating the required sample size from 340 to 720 patients.
    • Individual randomization required 340 patients, while practice randomization (15 patients/48 practices) needed 720.
    • ICCs for medication use, healthcare utilization, and general health were below 0.1; ICC for systolic blood pressure was 0.199.

    Conclusions:

    • Cluster randomization can substantially increase sample size requirements compared to individual randomization.
    • Varying ICCs across outcome variables highlight the importance of accurate estimation for robust study design.
    • Valid ICC estimates are essential for ensuring adequate statistical power in CRCTs.