(Total Views: 632)
Posted On: 09/30/2020 1:07:12 AM
Post# of 145890
I don't have any knowledge to doubt the PharmD's comments. The math is called sequential analysis and there once was a time I knew the basics.
The OncoImmune efficacy stoppage PR said "P=.005" though it didn't say whether that was the limit or actual.
Speaking first principles, if the sponsor gets to choose the most promising of several endpoints, you have an order statistic to compute: P(best of 3 endpoints < .05) is much greater than .05, so you'd need a big adjustment. If it's just increasing N, I see much less benefit. If the sponsor has no action they can take, why would there be a hit?
I don't see why the FDA wouldn't promote enrollment in a promising trial.
So you can see I'm clearly not knowledgeable about how the FDA actually works here.
The OncoImmune efficacy stoppage PR said "P=.005" though it didn't say whether that was the limit or actual.
Speaking first principles, if the sponsor gets to choose the most promising of several endpoints, you have an order statistic to compute: P(best of 3 endpoints < .05) is much greater than .05, so you'd need a big adjustment. If it's just increasing N, I see much less benefit. If the sponsor has no action they can take, why would there be a hit?
I don't see why the FDA wouldn't promote enrollment in a promising trial.
So you can see I'm clearly not knowledgeable about how the FDA actually works here.
![](/m/images/thumb-up.png)
![](/m/images/thumb-down.png)