@BmfReports But please, hold on to that short position. It'll make a lot of money for those of us intelligent enough to be long....based on actual data, not FUD
@BmfReports SELLAS described 0.636 as the threshold for statistical significance โ not as the powered effect size. That is precisely the number I computed two turns ago as the alpha-spending-adjusted critical HR at 80 events: 0.637. The company was accurate in 2022.
@BmfReports The sentence that settles the 90%-power argument
Also verbatim from the same filing:
"Statistical significance would be achieved by an estimated hazard ratio (HR) for OS of 0.636, corresponding to an OS of 12.6 months versus eight months for GPS versus BAT, respectively."