Manfred, I have post in the past very well calculated numbers similar to the ones you presented, about two years ago.
My quick example of an "N" of to 10,000 has to deal with the actual lower mortality ranges, and the actual decrease from 9.2% to 7.5% (correct numbers?), one definitely need very high numbers to prove this.
For amputations 7.5% to 3.2% the N will be lower but still probably like 800 to 2000.
In many cardiac studies this are the kind of numbers that are use.
Manfred there was a logistical problem, if they do not take all the lower mortality patients the study as the recruitment was going it was going to last 3 to 4 more years at best.
There is no way around the actual low incidence of meningococci disease. It was the St Mary's Imperial College in London that was able to recruited most of the patients, if not it will have been worse. Except if one goes to an epidemic like three years ago in West Africa, then the problem is not time but resources and quality of the data, and the difference in care with the developed world.
And were you said "beneficial effect HAD been proven" I am reading "WOULD HAVE or WOULD HAD" correct?
Anyway, to assume that patients in more dire needs will benefit the most is higly speculative now that one see the difference in mortality in the actual "stronger" patients if they would have reduced mortality 70% in this actual group of subjects it will probably have been statistically significant, but there was some 22% difference at best (quick guess, got to get to those numbers).
The protocol that we have in the thread in the past do not give the expected statistical expectations (We tried to guess it on the thread over the years), only a presentation of the protocol according to the modified Glasgow scores, exclusion criteria, arms division and so on. |
