InvestorsHub Logo

Antti

05/13/19 9:11 AM

#10533 RE: heldnova #10531

Does anyone know how the p value would change if we tripled the number of patients in each group and the SIB score difference remained the same. I'm thinking it would be substantially lower.



I think p value would be p<0.0001 with the current number of patients and with the same dropout percentage. And we already know that dropout percentage seems to be lower than in the previous trial (in 20 micro gram + placebo non menantine groups).

It may seem like a surprisingly low p value but the reality is that the treatment effect was so big that it resulted p<0.012 even with just a handful of patients. So around 100 patients will result very low p value with the similar results. And the other way around, we will get highly statistically significant results even with much lower treatment effect. E.g if SIB difference between the groups would be 4 it would still mean well below p<0.01.