Estimated statistical powers required for two period cross over trials for a 20% treatment effect using variable durations of follow up for the reference data set. The simulated duration of follow up was 3 months for each period and the tested response variable was the number of new enhancing lesions
| Sample size | Number of months used for simulations | ||
|---|---|---|---|
| 3 | 6 | 9 | |
| 2×10 | 27 | 25 | 23 |
| 2×20 | 67 | 47 | 41 |
| 2×30 | 77 | 71 | 56 |
| 2×40 | 89 | 73 | 64 |