Table 2. The original data and adjusted data for dichotomous data of the cluster-randomised trial (Luoto 2011).
Outcomes | Intervention (Original data) | Control (Original data) | M1 | ICC2 | Design effect3 | Intervention (Adjusted data)4 | Control(Adjusted data)4 | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Cluster number | Total number | n | Cluster number | Total number | n | total number | n | total number | n | ||||
Preeclampsia | 7 | 216 | 14 | 7 | 179 | 10 | 28.21 | 0.12 | 4.27 | 50.64 | 3.28 | 41.96 | 2.34 |
Birth-weight > 4000 grams | 7 | 216 | 37 | 7 | 179 | 36 | 28.21 | 0.12 | 4.27 | 50.64 | 8.67 | 41.96 | 8.44 |
Infant birth-weight > 90th centile | 7 | 216 | 26 | 7 | 179 | 34 | 28.21 | 0.12 | 4.27 | 50.64 | 6.10 | 41.96 | 7.97 |
Infant birth-weight < the 10th centile | 7 | 216 | 10 | 7 | 179 | 5 | 28.21 | 0.12 | 4.27 | 50.64 | 2.34 | 41.96 | 1.17 |
M = average cluster size = (total number of intervention + total number of control)/(cluster number of intervention + cluster number of control)
ICC = intraclass correlation; obtained from the reliable external source (Luoto 2010).
Design effect = 1 + (M-1)/ICC
Adjusted data = n / design effect