๐Ÿ“ Statistical Power & Sample Size

Determine the minimum number of vessels required to achieve a reliable f2 result before running a dissolution study. Uses a Monte Carlo bootstrap simulation (up to 10,000 reps per n) to estimate empirical power โ€” the probability that f2 โ‰ฅ 50 is correctly detected when the true profiles are similar.

1 โ€” Profile Parameters

5 %

Average point-wise difference between test and reference. 0 = identical.

5

Sampling points (e.g. 15, 30, 45, 60 min = 4 points).

8 %

Vessel-to-vessel variability at each time point.

0.6

How correlated adjacent time points are for the same vessel.

Type-I error rate. FDA/EMA practice uses 0.05 or 0.10.

More replications = more accurate but slower. 5,000 is sufficient for most purposes.

2 โ€” Reference Profile Shape

Reference mean % dissolved at each time point, comma-separated.