I was hoping someone might be able to offer information about calculating sample size when cluster randomisation is used and the cluster sizes vary considerably.
In the case of unequal cluster sizes Donner & Klar (2000) suggest that 'cluster size' (used in the sample size calculation) is replaced by an estimate of the average cluster size. They state that "this approximation will tend to slightly underestimate the actual required sample size, but the underestimation will be negligible provided the variation in cluster size is not substantial". Does anyone have any suggestions for the case when the variation in cluster sizes is substantial?
Any help is much appreciated.
Paul
Donner, A. & Klar N. (2000) Design and Analysis of Cluster Randomisation Trials in Health Research. Arnold, London.
_________________________________________
Paul Brown MSc
Medical Statistician
NYCTRU
17 Springfield Mount
Leeds
LS2 9NG
UK
Tel: 0113 233 1474
Fax: 0113 233 1471
Email: [log in to unmask]
Web site: www.nyctru.co.uk
Confidential information may be contained in this message. If you are not the addressee indicated in this message, you may not copy or deliver this message to anyone. In such cases, you should destroy this message, and notify us immediately.
|