Table 4 Effect of study length on the power of the study
Scenario | Estimated power | Threshold of statistical significance | HCEQ Standard Deviation | Minimum detectable difference (Cohen’s effect size of 0.6) | Intra-cluster correlation | Adjusted standard deviation | Number of clusters (clinics) | Totalsample size | Study length (months) | Mean number of patients per clinic |
|---|---|---|---|---|---|---|---|---|---|---|
1 | 0.519 | 0,05 | 2.46 | 1.476 | 0,02 | 2.65 | 6 | 54 | 1 | 9 |
2 | 0.76 | 0,05 | 2.46 | 1.476 | 0,02 | 2.85 | 6 | 108 | 2 | 18 |
3 | 0.88 | 0,05 | 2.46 | 1.476 | 0,02 | 2.98 | 6 | 162 | 3 | 27 |
4 | 0.92 | 0,05 | 2.46 | 1.476 | 0,02 | 3.21 | 6 | 216 | 4 | 36 |
5 | 0.948 | 0,05 | 2.46 | 1.476 | 0,02 | 3.37 | 6 | 270 | 5 | 45 |
6 | 0.963 | 0,05 | 2.46 | 1.476 | 0,02 | 3.53 | 6 | 324 | 6 | 54 |