Figure 4From: Sample size calculations for cluster randomised controlled trials with a fixed number of clusters Minimum detectable difference ( π 2 ) at 80% power various different baseline proportions ( π 1 ): limiting values as the cluster size approaches infinity.Back to article page