Replies to post #176082 on Biotech Values

[hirogen]

EXEL

? What needs reconciliation? Everything reconciles to the sig figs given.

E.g. the trial was probably designed with assumed HR=0.75, 90% power (math vs patient planned for final trigger is almost perfect match). As per David's post placebo arm got 7 months - which implies that treated arm would be about 9.3 months. But in a conf call or conversation I would actually expect them to round - i.e. call the 9.3 months "9 month"

I thought where David wrote 9.0 and 7.0 for the trial arm median survivals the .0 was significant, e.g. if taken from a co presentation. Taken literally and under PH that would mean a material reduction to stated power. So to get to 90% with 578 events the .75 HR would have been some weighted average or otherwise summary of time dependent hazards that would tie out in standard powering calcs. Don't see NPH in trial design too often, the possibility piqued my interest.