Not so fast... There are different boundaries, e.g. Haybittle-Peto, O'Brien-Flemming, and in my opinion, consistent with other COVID-19 studies (like DISCOVERY), a more conservative one will be used (0.005 or 0.001), especially when the sample size isn't that large and two events can produce a big variability...

https://en.m.wikipedia.org/wiki/Haybittle%E2%...o_boundary

https://www.google.com/url?sa=t&source=we...t8ms70w46o