Good question. Who's makes the "Censoring" decisions?
A censor occurs when there is no more data on the patient.
In ongoing trials there are many of these due to time. If the OS (or whatever) curve goes out 2 years, and a patient has only been in the trial for 1 year, then he gets censored at 1 year. This is simple, and not the point of AF article.
Some patients though are lost to follow up, i.e. they just vanished. This data (or lack there of) would of course have to be collected at the site. It is these type events that are being discussed here.
If there were more than a very small number, one has to wonder what was going on in the trial. And if they were unbalanced, the concern is worse.
I do not think anybody is accusing PPHM of actually doctoring the numbers, more like a "lack of institutional control" over the sites allowing this to happen. Matters little if it was the CRO or PPHM.
Market Data 
Markets 
AI
