“And if we tell everybody in the world that Michelle had the active treatment, what possible difference can it make to George or his parents or his clinician even if George has only just started? None again that I can see. George and his team know what Michelle had, but it tells them nothing about what George has.”
No TT ... wrong.
Perhaps George is ¾ of the way through. His parents hear on the news of the tremendously positive preliminary results.
But little George has made no progress. Bugger it his parents think – we must have the placebo. So they give up. And guess what .. the probability of their guess most likely is better than chance.
Blinding is not binary ... its about subjects not being better than chance at guessing.
There are three reasons why this was exceptionally stupid for Carrick and NTI to have done this.
First as soon as you do something like this the onus changes. You have to prove why something isn’t a problem instead of someone having to prove there is a problem.
Second in effect Carrick has reduced the power of the study. He is having two cracks at statistically testing the difference. And so you incur a statistical penalty for the second test – in effect the P reduces. A tougher P ... less power. Carrick probably doesn’t even know this ... but the FDA does.
Third everyone hates interim – preliminary results being made public. And will discount the credibility of your study simply because of it. This includes the FDA and the people who will review and critique your study if you want to publish it in a proper journal - not a vanity journal where NTI have previously published. In order to build credibility and legitimacy when you are starting from a very low bar you have to be completely squeaky clean.
So its a complete rookie mistake that only a group inexperienced in running trials to a serious methodological standard will make.
NTI Price at posting:
21.0¢ Sentiment: Sell Disclosure: Not Held
