> so the results were dismissed due to not being obviously bad. Not sure if 'dismissed' is the correct phrasing. The results were statistically analyzed as it'

> so the results were dismissed due to not being obviously bad. Not sure if 'dismissed' is the correct phrasing. The results were statistically analyzed as it's always done with empirical studies. > They also lost track of a huge number of participants (like 200-400?), and never figured out why. It's entirely possible they died and the safety profile world be completely different if they were included. Sounds like that were the drop outs which is normal for empirical studies. <1% drop out sounds quite small (don't know for this type of study in particular, though). I wouldn't be surprised if the very question you are asking has been subject to research in the last decades as every medication goes through a very similar process. In general, that phrasing sounds to me that somebody tries to frame something that's super normal (drop outs) to something questionable and mystical. I don't mean you, but the source you used back then.