
Sample Size Determination for Clustered Outcomes 53
TABLE 2.3
Stepped wedge design, where C represents control and I represents interven-
tion
Cluster Number Intervention
1 C I I I I I I
2 C C I I I I I
3 C C C I I I I
4 C C C C I I I
5 C C C C C I I
6 C C C C C C I
clusters cross over in one direction only, for example from control to interven-
tion. Table 2.3 shows the intervention sequence of the stepped wedge design.
Woertman et al. [61] derived the design effect (also called sample size
correction factor) from the formulas provided by Hussey and Hughes [60]
assuming that cluster size (m) is equal across all clusters. The design effect
can be used to estimate ...