Cluster Randomised Controlled Trial.
A group of subjects (as opposed to an individual subject) are randomised, with clusters being allocated to intervention arms and to control arms of a trial (Bland 2004). Clusters can include households, schools, villages and compounds.