Sold out a year ago, page-35

The trial data was promising, they cocked up the trial design though by:
1) not monitoring the real time general death rates of patients with covid (which dropped dramatically as standard of care improved, for example learning that patients on their stomach survived much better) and re-powering the trial accordingly

2) not updating their exclusion criteria as it became clear that there were certain age and weight cohorts that died much more frequently

3) being too aggressive on their targeted improvement for patients in the Bayesian trial design

Basically, if they were a bit more careful and cautious with their approach they could be approved by now which would have made the company. While all indications are that the product probably worked for a sub cohort of patients, we simply don't have enough data on that sub cohort and collected in a double blind trial to know for sure, and that could have been avoided and mitigated with a better approach to the trial design.

TLDR: it wasn't taking on covid that was the problem, it was probably the mistakes they made and the overconfidence in the early trial numbers that meant the trial was a failure.