Sample Size Calculations for Cluster Randomised Trials
Author Information
Author(s): Karla Hemming, Alan J Girling, Alice J Sitch, Jennifer Marsh, Richard J Lilford
Primary Institution: Department of Public Health, Epidemiology and Biostatistics, University of Birmingham, UK
Hypothesis
How can sample size calculations be effectively conducted for cluster randomised controlled trials with a fixed number of clusters?
Conclusion
Designing a CRCT with a fixed number of clusters might lead to feasibility issues, necessitating a minimum detectable difference or maximum achievable power.
Supporting Evidence
- Cluster randomised controlled trials are often used in health service evaluations.
- Sample size calculations are crucial for ensuring the feasibility of trials with fixed clusters.
- Power and detectable differences are limited when the number of clusters is fixed.
Takeaway
When planning a study with groups, it's important to know how many groups you can use and how many people you can have in each group, or else you might not be able to see the results you want.
Methodology
The paper outlines sample size formulae for CRCTs with fixed clusters, including required randomisation units, detectable difference, and power.
Potential Biases
Small numbers of clusters can introduce bias and affect the validity of results.
Limitations
The study assumes equal sized arms and normality, which may not hold in small cluster trials.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website