If you are just reading this "Statistical concerns - For the...

If you are just reading this "Statistical concerns - For the same power a trial that is randomised 2:1 needs 12% more patients," and adding a 12% of 60 (around 7 or 6point something) to 60 to get 67" - which I think IS your line of thought - you could have equally gone the other way keeping the statistical power constant.

i.e. you could have assumed the 60 already included an extra 12% already for the same statistical power and said (math is rough here - you don't get the same amounts adding and subtracting 12%) but 60 - (say) 7 is 53 or 54 for rounding.

Then you could have concluded it 53 versus 60 (for the same amount of statistical power) and say that that's not a huge difference.



Start with this AXIOM - "For the same power a trial that is randomised 2:1 needs 12% more patients".

We don't do trials to spend patients we do them to discover truth to try to help patients.

We (as a matter of public policy on doing clinical trials generally) don't want more patients then we need (because that involves taking longer than we need in a world where the unsatisfactory status quo (sickness and suffering) continues whilst we fiddle) - trials are supposed to be truth discovering tools - you get the truth and then you use the knowledge of what is true to try to change the status quo - which is the best way to alter world in which people are sick and suffering for the better sooner.