RE: What's the Hold Up?19 Feb 2022 19:48
“Either the top lines results are a bit sketchy and they are really having to work the numbers to get them looking half decent.”
I see this written a lot but there are a few reasons why this can’t happen on the topline primary endpoint data.
1. Firstly the data once cleaned it is locked for analysis. This dataset can be used by anyone (regulators/peer reviewers/SNG etc). So this dataset can’t be played with to get the result you want as it will get found out at a later date either by regulators/peer review.
2. The locked dataset has the Statistical Analysis Plan (SAP) applied to it. This document has already been submitted to the regulators, prior to completion of the trial or any unblinding of the data. It is very specific and outlines exactly what analysis is applied to each primary and secondary endpoint. The main analyses to be applied to the primary and secondary endpoints are outlined in the protocol for anyone that wants to see what they are doing.
3. Lastly the success of the trial, as outlined in the protocol, is based on very simple analysis of the two primary endpoints. One of these primary endpoints needs to be statistically significant for the trial to be considered successful.
So any suggestions that they are working the numbers don’t relate to reality. The way the numbers are ‘worked’ are documented prior to trial completion and both regulators and peer reviewers have access to these documents.