Jalaluddin, you may have missed my previous post on the subject,...

Jalaluddin, you may have missed my previous post on the subject, but I have rejected my earlier 'swap hypothesis'. I'll try to explain my reasoning in a bit more detail.

Statisticians will tell you that as you increase the sample size, the standard error will reduce (by the inverse square root of sample size) but the standard deviation should remain the same.

We don't have the standard deviation numbers in the phase 2 results presentation, as they are rarely provided. However we have the standard error, and a while ago now, Davisite kindly fleshed that out a bit more for us.

Taking the standard error results and the sample sizes, you can derive the standard deviation for the US and Aussie results. I didn't have the time to calculate exact numbers, but my back-of-the-envelope workings indicated that the standard deviations of the US results were up to 50% higher than the Aussie results.

Neither the 'swap hypothesis' nor the 'null hypothesis' can explain the high standard deviation of the US results. But there are three possibilities (that I can think of) to explain this high standard deviation:

1. Differences in cohort selection criteria & processes
2. Differences in clinical application & monitoring processes
3. Variation in the product tested

The company has categorically ruled out #1 & #2, which leaves us with #3 as the only logical explanation.