Thanks TechGuru. Good post as usual. One poi
Post# of 154722
(Total Views: 685)
Posted On: 10/24/2020 8:37:57 PM
Thanks TechGuru. Good post as usual.
One point though:
You Wrote:
This is precisely what the DSMC communicated:
Not only did they NOT indicate that “we can probably meet statistical significance” -- but the exact wording of what they DID say actually leaves open the possibility that the DSMC might ask for a larger sample size at 75% interim -- beyond the full enrollment protocol of 390.
I spend quite a bit of time documenting what is -- and what is not said during these calls. We are all human, but I think we get into trouble when we imprint and overlay our own wishes and desires on facts that are presented to us.
One point though:
You Wrote:
Quote:I have listened to the October 20 call and recent Proactive video several times (as I always do) and can find absolutely no reference to ANYONE saying that the DSMC statistician considered that with 75% of patients we can probably meet statistical significance.
What is very encouraging is that the statistician of the DSMC considered that with 75% of patients we can probably meet statistical significance (p + power). This would indicate that we are now somewhere in the range 0.0065 - 0.016 meaning we could possible have a well-powered result with 293 patients.
This is precisely what the DSMC communicated:
Quote:
Continue the study as planned with the protocol, defined sample size and power to achieve the primary endpoint. The DSMC recommendation was to continue the study as planned but conduct another interim analysis when 75% of the planned subjects have been randomized and have been on the study for 42 days or withdrawn, whichever occurs first. Sample size reassessment will be considered at that interim analysis.
Not only did they NOT indicate that “we can probably meet statistical significance” -- but the exact wording of what they DID say actually leaves open the possibility that the DSMC might ask for a larger sample size at 75% interim -- beyond the full enrollment protocol of 390.
I spend quite a bit of time documenting what is -- and what is not said during these calls. We are all human, but I think we get into trouble when we imprint and overlay our own wishes and desires on facts that are presented to us.
(7)
(0)