> RCTs are fine but the obsession with them is overwrought and counterproductive. My own drum to beat on this is regarding clinical trials for fatal diagnoses like cancer:
RCTs for mental health conditions are a completely different situation. The short-term placebo response rate for cancers is not high (obviously) though the influence of unblinded trial operators making subjective analyses can be a problem.
Many mental health conditions, on the other hand, have unbelievably high placebo response rates over the duration of a short trial. The magnitude of the placebo response is almost hard to believe in certain studies.
The placebo effect can be a problem for approving new drugs as some times the placebo group improved so much that there isn’t much room left for the active drug to improve beyond that. This is a problem of study design and rating systems that is difficult to solve.
Unfortunately, some study operators use this fact to their advantage by omitting placebo group. Without a placebo group, it’s not obvious that the drug is actually doing anything better than placebo, of course.
RCTs for mental health conditions are a completely different situation. The short-term placebo response rate for cancers is not high (obviously) though the influence of unblinded trial operators making subjective analyses can be a problem.
Many mental health conditions, on the other hand, have unbelievably high placebo response rates over the duration of a short trial. The magnitude of the placebo response is almost hard to believe in certain studies.
The placebo effect can be a problem for approving new drugs as some times the placebo group improved so much that there isn’t much room left for the active drug to improve beyond that. This is a problem of study design and rating systems that is difficult to solve.
Unfortunately, some study operators use this fact to their advantage by omitting placebo group. Without a placebo group, it’s not obvious that the drug is actually doing anything better than placebo, of course.