Replies to post #45118 on Biotech Values

[Jonathan Robinson]

Now, assuming(and I think it is a bad assumption) that 6/9 reflects true SVR20 proportion, the real SVR20 that would playout with larger study numbers could be assumed to be 0.666*0.799 = 0.532 ie 53.2%



You are still missing the drop-out effect however.
Most if not all drop-outs came much later. The better measure is the 70% SVR ITT measure at 12 weeks for the whole trial. Using 2/3 you 47%. Using 15% (feels more right to me), you get 60%. But I still think 12 + 24 SOC (12 being concurrently and 12 after stopping) will win out ultimately.

Jon

[gofishmarko]

>>> the real SVR20 that would playout with larger study numbers could be assumed to be 0.666*0.799 = 0.532 ie 53.2% <<<

I don't think you can look at it that way. The only people who would stop tx. at 12 wks would be those neg at both 4 and 10 (or 12 , for our purposes here ) wks. Assume that the 70% who were neg at 12 wks were also neg at 4 , and use the 66.6 % SVR rate from group D , and you get .666 x .7 = 46.6%. Of course , that doesn't include SVRs from patients who would continue tx. past wk 12 after failing the cutoff criteria. That is academic though , since no SOC would stop tx. at 12 wks. with only a 66% chance of SVR , though many individual patients would , I'm sure. :)

EDIT :
JR beat me to the punch this time.