As someone who has both designed and run clinical trials as my job, this assertion is likely* nonsense. First, the division you submit to will reject your recruitment criteria if they feel it does not accurately reflect the treatment population within the United States. They also reflect the specialty of the particular division (e.g. Oncology is allegedly far less worried about most side effects than, say, the Dermatology division, due to the kinds of indications they deal with).

Second of all this kind of info is on the label (prescribing information) even if it doesn’t make it into the short summary (package insert, typically only a dozen pages or so) given to patients. Doctors do read those, you know, and within their specialities know what kinds of things to look for.

There have been some notorious cases, but by and large I’ve found the people I worked with at the agency to be professional and solid. I’m no longer in that business and have no reason to say anything I don’t believe.

* I didn’t bother to look up the label for this drug but they are all public info on the FDA web site and In the USP.