I think it would be fair to say it depends on the drugs being tested, and what they are being tested for, and where those drugs are in the testing process.
At some point, all* new drugs will be tested versus placebo for safety reasons. Later research may be conducted against other drugs to determine efficacy of one versus another for something specific. Perhaps placebo-arm trials are rarely published due to a variety of reasons, such as earliness in the testing cycle or relative unimportance of the drug being tested. Something “hot” or new might be worth publication, versus something with OK safety results but equivocal efficacy results. It might depend also on how many effective treatments already exist in the area being investigated. In the end, what is worth publishing may be that new drug “X” is better than older drug “y”, hence the trials that use those two drugs in a double-blind study.
TL;DR I think we may be discussing a difference in what is published, versus the protocols.
*Almost all, with exceptions as noted above in the thread. Some drugs can’t effectively ever be tested in this manner. The FDA does allow the step to be skipped in some cases, following very specific guidelines, etc.
I still have this on my to-do this, but would caution against its utility. It’s more likely to be “interesting” but considered non-scientific at this point in time. Depends on what I find, of course. I’ll keep you posted.
Ran out of time - I did want to acknowledge this specific point. Yes, you are correct. It’s a case of placebo studies all being double-blind, but not all double-blind studies necessarily being placebo. (Let’s not even get into multi-arm. ;))
Well let’s say that the study is done and shows a significant effect. That could lead to an argument that it is very important for physicians to forcefully emphasize to their patients just how effective the drug they’re given is (even exaggerate the effectiveness).
^Melbourne indicated otherwise and I was just trying to figure out what was going on. In my line of work, I tend to be dealing with P3 data, so that’s what I read. Looks like Melbourne works in an area where active-control is much more common so we’re coming at it from different directions.