Replies to post #150246 on Cytodyn Inc (CYDY)

[djjazzyjeff]

because if you were to flip a coin 100 times twice, and you got 40 heads the first time and 30 heads the second time, would it surprise you? No. And 10/40 is 25% That's why it's not significant, because there's simply no way to know if that 24% is by chance. I mean, if you flipped a coin 10 times and got all heads would you be bringing it to trade shows to sell as a magic coin? No.

[stockorus]

We have ~ 60 critical patients.

Assume we get a 1:2 distribution between SOC and LL arms. (I recall Nader saying that they had this distribution for the whole trial, not sure; but we can hope that the same holds for critical subpopulation.)

Then 20 in SOC and 40 in LL.

Suppose 40% death rate in SOC and 30% in LL. (That's 25% reduction.) Looks high but it is Critical group; anyway if lower, it will not be bad for the point I am making below.

Then we have 8 died in SOC and 12 in LL.

This may be too small sample-size to make a general statement on efficacy or statistical significance. Our calculations will not stand out beyond the margins of error. So Nader mentions the % reduction but cannot say more.

But then on the reverse side, we have 12 survivors in SOC and 28 in LL.

For them, the data says: the average hospital stay for the 12 in SOC is 6 days more than the average stay for the 28 in LL. Now we have a relatively bigger sample size and the number 6 is so sharp that we can do statistical analysis and determine significance for this type of question (as opposed to mortality).

If the data came out somewhat in above manner, I think this time-to-discharge result can make a strong enough impression on FDA (given that we know they took it seriously for Remdesivir), that LL is showing clear benefit in the critical survivor population at least.