ikulo wrote:Even if the trial is set for a long duration, it may be ended early if the placebo group gets much worse. I've read about such results before.
fogdweller wrote:In most cases if nat all there will be an intermediate point, and this would be another consideration to the 6 CeCe listed. If they check after 6 months or after 2 years it would make a difference to deciding if you want to volunteer.
fogdweller wrote:This is true, but in most studies there is an intermediate analysis point where the statisticians look at the data and tell the researchers one of three things...(1) either to stop the trial early and offer the treatment to everyone, the end effect has already been established, or (2) keep on with the study, it may or may not find statistically significant results, or (3) stop the study, no matter how good the final group is, it will not be enough to establish statistical significance.
AMcG wrote:Almost all of them were imperfectly blinded by substantial amounts. I think one reported 80% of the treatment group knew they were on treatment. If I remember right none had controls matched for sex or age. One even had 70% females but no information about how they were distributed between the groups. I had expected these studies to be whiter than white since so many neuros bang on about the importance of placebo effects and controls.
Like you, I am beginning to wonder whether it is only important to include controls because you will be criticised if you don’t but that they don’t really think it is important.
I really wish I could trust these people more.
Users browsing this forum: No registered users