Obviously, NP claims he doesn't know the results.

Obviously, NP claims he doesn't know the results. It's possible that protocols keeps those at Amarex. Or not.

The primary runs were done immediately at Amarex. The people there are now looking at outlier data points and redoing the main results manually, looking for any mismatch they can possibly tease out.

I note the DSMB committee meeting was described by NP as a periodic meeting - not driven by either a data event or a request. There should be looks at safety, efficacy and futility - what more can you ask. The standards are high (low p-value) because otherwise multiple looks at the data add up to too great a chance of false significance.

There was a SAP (Stat Anal Pkg) filed at the start of the trial that lays out the (pre-decided) "decision" boundaries. Asking for a change in trial size based on (blinded) combined event rates (deaths) is explicitly OK.

NP has never been exactly clear what other leverage he holds in calling for either a partial or full analysis at any other point..