Minimum detectable effect size for main outcomes (accounting for sample
design and clustering)
Assuming a target population of 1,100 participants and 40% non-response rate, we expect an effective sample of 660 individuals. We conducted power calculations for an individually randomized design with a binary outcome (Y), using a two-sided test with significance level α = 0.05 and power 1 − β = 0.80. Given the assumed control-group baseline outcome rate (specified below), the implied Minimum Detectable Effect (MDE) is:
MDE= 9.5 pp if E(Y)=0.20, for two treatment groups (C=1/3 vs T1=1/3 or T2=1/3)
MDE= 8.4 pp if E(Y)=0.20, for one pooled treatment group (C=1/3 vs pooled T=2/3)
MDE= 11.4 pp if E(Y)=0.30, for two treatment groups (C=1/3 vs T1=1/3 or T2=1/3)
MDE= 10.0 pp if E(Y)=0.30, for one pooled treatment group (C=1/3 vs pooled T= 2/3)