The steps involved in calculating the sample size of the study
| Step | Hypothesis | Total number of patients to be included | Total number of patients to be included in the intervention group | Total number of patients to be included in the control group |
|---|---|---|---|---|
| #1 | To detect a difference of at least 15 points in the DQ (SD=25) with a two-sided 5% significance and a power of 90% with a ratio 1:2 (intervention:control) | 132 | 44 | 88 |
| #2 | A 10% cross-over rate in the control group involve a 20% dilution rate | 132+27=159 | 44+9=53 | 88+18=106 |
| #3 | A 10% drop-out rate over the 24-month follow-up | 159+21=180 | 53+7=60 | 106+14=120 |
After the 1st step, a total of 132 patients (44:88) are required to detect a difference between groups under assumptions.
After the 2nd step, a total of 159 patients (53:106) are required to correct the 20% dilution bias due to the cross-over effect. To maintain the 1:2 ratio, we rounded the sample size up to the next whole number (27 patients rather than 26). After randomisation, all patients will be then followed up during 24 months.
After the 3rd step, a total of 180 patients (60:120) are to be included to account for ∼10% drop-out rate during the 24-month follow-up. Again, we rounded the sample size up to the next whole number to keep the 1:2 ratio between groups.
DQ, developmental quotient.