Replies to post #118222 on NorthWest Biotherapeutics Inc (NWBO)

[monentum2play]

Okay so who is right and what is correct? This could take another year and how is the company suppose to survive? Anybody know?

[exwannabe]

I think you are wrong here....If we already know 166th Patient OS data then median Overall survival (mOS often referred) data will not change ...

No, because not everybody enrolls at the same time.

The "166 event" refers to the real time calendar. In theory, it could have happened to a patient run over by a car on the day he was randomized. It is NOT number 166 in events ordered by time since randomization.

Senti is somewhat on the right tract, but can not really show the point w/o getting into a clear understanding of how K/M and censoring work. And. IMO, trying to explain that is near futile. Good luck to Senti if she wants to try :-)

It is true that IF the trial show a big fat tail, then some key stats (P and HR) will improve with more data.

The median though should tend to stay where the KM curve currently predicts it (wherever that might be). But that is really the least important number. Senti's "crayon" KM OS curve example would be very approvable despite zero median OS improvement.

[sentiment_stocks]

Median OS data will be when 166th patient in trail died.

Believe me I understand why you think that, because I used to get confused by this as well.

But "time" in terms of the trial is not the same as time as we know and live it.

For example, pretend the 166th OS event right now is positioned at 25 months. But remember, there are still over 98 patients alive in the trial. If one of them OSevents before 25 months, they will move the current 166th mOS out of that position, to the position of the former 165 OS event.

It can get worse but will not get better with remaining patients low OS data.

So actually, the data can only get better the longer out we go. But at some point, the patients may no longer pass, and so the median stays in place.