There are really two datasets at play. One is the safety dataset and the other is an efficacy dataset.
The safety dataset comprises all the 66 subjects. The efficacy dataset comprises only subjects who completed the treatments per protocol. It is very rare for 100% of subjects to complete treatments as per protocol.
If you try and reverse engineer to derive what the actual numbers were in the efficacy dataset by group my hunch is the numbers in the control group were quite a bit lower than in the treatment group. Suggesting differential attrition.
The lower numbers also means less power to test whether the differences were statistically significant. But no formal power calculation was performed in the design of the study. And so it is unclear whether the lack of p values in the announcement is due to a lack of statistical testing full stop or a lack of statistical significance which has simply not been reported.
So you pull a thread … the whole jumper unravels.
BOT Price at posting:
14.0¢ Sentiment: None Disclosure: Not Held
(20min delay)