Randomization unit is not individual subject but group of subjects
Design Effect: 1+(m-1)*ICC
ICC: intra-cluster correlation coefficent
#Advantage
- optimal design for evaluating qualit improvement strategies in health care intervention and education program studies
- account for contamination between patients within a cluster
- easier to administrate the randomization and centers
# Disadvantages
- larger sample size is required than for a simple randomized trial
- the patients and clinicians may recognize whether they are in the active or placebo arm
- clinicians might transfer information between clusters
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment