Atlas,
You give me far to much credit. All i did was assume 1-100 placebo events because it was clear that not all placebo patients
had an event. if you are given the number of placebo events and a hazard ratio then mathematically you must have the number of gevo events. that is all the chart is plain and simple.
kaplan-meir, i know this trial will be analysed with KM, but i
can't answer your questions at all.
Yes, I'd like to know a few things too.
What is the target event number?
"in the early months of the study the exacerbation rate was running at Servier's expected rate"
What does expected rate mean? up until this trial they were only checking efficacy and now suddenly they are exacerbation endpoint experts? the data was blinded. they assumed 0.3. how did they know any 'expected rate' if they didn't know who was placebo and who was gevo? How did they come up with 0.3?
As far as your assumption as to a "different observed hazard ratio" i can easily agree. Assuming ~5 or less additional [is
that a handful? ] patients are needed to exacerbate out of the additional 20 they signed you can get some good numbers in the "0" (ie only placebo exacerbations) and "0.05" (1 to a possible round up to 2 gevo exacerbations) which could mean from 1 to 6 placebo's never exacerbating.
all this of course assuming target events=30
Market Data 
Markets 
AI
