decision. they have overwritten advisory committees that have kind of been close but this is very different from that. there's another problem with this study, and the other problem is that in the analysis that was presented to us, the original study as designed failed completely. then they tried to go back and retro actively pick out a sub-group that got better. that type of approach, statistically, is very biased, and typically does not permit -- predict what an outcome would be if one did a forward-thinking or what we call a prospective study. >> okay. >> that was really a failure. there's a lot of pressure from the alzheimer's groups and others to bring forward a treatment. alzheimer's disease is a terrible disease. it's incurable. so the public pressure for this is great. but i have a major concern. and that major concern is if we approve something that truly is not effective, and has