J
John Morris
If the "switcher" knows the identitity of the product being tested, it's not a DBT.
Most times, the trial is not unblinded until it is finished. Sometimes, though, as the trial progresses, you can guess which choice the subject is getting just by how they respond to the choice.
A while back, my wife was involved in a NIH trial and after awhile it wasn't unusual for us to get phone calls in the middle of the night from a frantic wife who was crying that her husband was unconscious and not breathing. In these cases, the blind needed to be broken immediately so that she could advise the paramedics if the patient was on study drug or placebo. Almost always, they were on study drug. Shortly thereafter, as more and more patients who were on drug died, the study was stopped and unblinded fully. It seemed that in this specific situation, the drugs commonly used to treat this problem was actually leading to increased mortality and morbidity, rather than helping. The only way that we could have known that this accepted treatment was harmful was to run this randonized, double blinded, placebo controlled trial.
Thank goodness that our use of these DBTs isn't so serious or important!
Oh and Larry, it was CAST.