Objective outcome measures, though limited, are indeed available for psychopsychiatric interventions, even on purely psychiatric disorders. I suspect that the difficulty involved, and the likely outcome such as showing their interventions are without value, is why they do not happen. For example, you could use actometers, document increased hours able to work, document prescription drug usage, doctors hours and other medical interventions such as the number of days hospitalised. This would be about getting a broad objective framework. Or could it be they know the result, and that much of what they do results in no documentable improvements, or even documentable worsening?
Instead what we often see is a deliberately laxity in experimental design. Psychiatric research is often given special privileges as a result. While this is permitted, tolerated or excused, how can we trust any of their research? I am sure there is some good research there, but perhaps I am expecting too much of them to expect that most of it should be good quality research.
The brain is about the hardest thing scientists have ever studied. That does not mean they can be excused for low quality science.