NorthWest Biotherapeutics Inc (NWBO)

[flipper44]

"With 108 survivors at 3/17, a 2% or 2 per month OS events is almost the same thing. -- meirluc"

I thought this was self evident, and the whole point of why I was trying to distinguish % from n.

Cliff notes: Dr. Bosch spoke in June 2017, when event % rate and n rate were exactly the same. 2% per month and n= 2 per month. From that time forward even if the event % rate stayed constant, the n rate would decrease because the pool becomes smaller.

That is why it's important to disingush. For instance, automatically, once you get down to 90, even if the event % rate stays constant, the number rate is already down to 1.8 per month.

Now look at the whole picture,

The reason it is much more likely factors compound to lessen number eventing is a combination of:

smaller numerical pool over time + phenomenon of event rate slowing, because the longer one likely survives in the future is based on the longer they already lived post surgery + the last 31 likely being randomized only to treatment + wild card of whether or not improving manufacturing.

Even without the last two factors the % rate would slow but the n rate would slow more dramatically. If one or both of the last two factors are added, the reduction in event rate n would be even more pronounced.

The fact is, if a statistician can't project from 36 months post last surgery in an eleven year old trial, then the field of statistics ought to pack up and go home.