Infinity, patient numbers were in as ITT=
Well, silly or not, I, at least, need some comparable measures of benefit, even in small trials. Fractions, like 6/18, are probably better that percentages, like 33.3%, because the expression includes the population. And it is easy to compare 6/18 to, say, 4/23. One spots the difference instantly.
Of course it dangerous to compare results from 50 head trial to 500 head trial - only an ingrate would assume that the comparison is level and fair.
Unfortunately, one needs some kind of historical decision landscape and history is what history is.