Ann: Paradigm Reports Very Encouraging Real World Data, page-18

That's why trial design and screening is important. From their phase 2b and SAS data, they've learnt that Day 53 (week 8) is optimal for the chance of success. The placebo effect is not too bad after say 2 months looking at that graph. Under 20%?

Similarly, they've worked out that the gap is biggest for patients with NRS pain 4-6. It's not that PPS is less effective on patients with higher pain. But instead it's because the placebo effect is much higher on them, thus reducing the probability of demonstrating statistical significance. I would assume that statistical significance is almost certain when the sample size is as big as n=750 and n=400, considering they had no trouble for n=112.