(Total Views: 630)
Posted On: 03/07/2021 10:27:59 AM
Post# of 154718
From last PR we now know that there were a total of 80 deaths and a total of 384 patients. This allow us to elaborate with some more precision some scenarios. Please find below one of them:
The assumptions are:
Total Number of Patients: 384 (patients are in violet cells with maroon numbers)
Total Critical patients: 62 (in lower table, 43 Vx/19 SOC)
Total Severe patients: 332 (215 Vx/ 107 SOC)
Total number of deaths: 80 (34 in Severe and 46 in Critical)
To note: Most probable scenario were 14 survivals in Vx against 2 in SOC (for critical). Since we did not reach statistical significance in the Severe group I surmise we had more than 21 deaths (need 17 or less for p<0.05).
It is apparent how difficult is to reach significance. Also, if the placebo group had one more survival, the p-value jumps dramatically.
Now, this table is just “food for tough” as many assumptions have been made. I do suspect that there were 3 deaths in the SOC group, however, this would bring the deaths reduction to 17.2% and this is not what we reported.
Will be happy to run stats in any other scenario that is proposed provided it conforms to reported numbers.
The assumptions are:
Total Number of Patients: 384 (patients are in violet cells with maroon numbers)
Total Critical patients: 62 (in lower table, 43 Vx/19 SOC)
Total Severe patients: 332 (215 Vx/ 107 SOC)
Total number of deaths: 80 (34 in Severe and 46 in Critical)
To note: Most probable scenario were 14 survivals in Vx against 2 in SOC (for critical). Since we did not reach statistical significance in the Severe group I surmise we had more than 21 deaths (need 17 or less for p<0.05).
It is apparent how difficult is to reach significance. Also, if the placebo group had one more survival, the p-value jumps dramatically.
Now, this table is just “food for tough” as many assumptions have been made. I do suspect that there were 3 deaths in the SOC group, however, this would bring the deaths reduction to 17.2% and this is not what we reported.
Will be happy to run stats in any other scenario that is proposed provided it conforms to reported numbers.
(10)
(0)
