Therapy is not a pill. We should not be testing it like one by relying so heavily on randomised trials. The way forward is to pick the right research tools and take them seriously.
The NHS Talking Therapies programme generates one of the
largest routine mental health outcomes datasets in the world. However, this observational data plays a much smaller role in Nice guideline development than evidence from randomised controlled trials, despite research suggesting that many schools of therapy perform
broadly on par with each other.
Nor do these institutions take
research with patients into account, meaning that groups that struggle with the sorts of therapy adapted for trials like CBT, continue to get inadequate treatment.