If you're new to scientific papers, the heuristic to use is: Read the Abstract, Introduction, and the Discussion/Conclusion. Methods & Results are boring and numbery, and you can dive into those later.
I highly recommend reading the whole General Discussion section. It's four pages. The main idea is that:
- There are previous well-controlled studies that show the benefits of therapy, but
- When we are simply observing a cohort of people, the people who have some form of therapy experience a decline in positive indicators and a rise in negative indicators. (Indicators like self-esteem, depression, etc)
This might be because:
1. The therapy did help, relatively, but because of the strong force from a patient's trauma, they were still generally doing poorly. But it could have been worse.
2. Because of the broad definition of "therapy experience" and the limited data in the survey, maybe this is mostly capturing people who only received a few sessions, and didn't continue long enough to receive the benefits, or in the case of the US cohort, only received drugs.
3. Maybe everyday therapists are kinda shitty? The therapists selected for controlled studies are always up to date with modern techniques. The paper said it more eloquently:
> The third explanation concerns therapists who deliver therapy. Again, in a controlled intervention, treatment integrity is often one of the highest priorities of the research protocol. In contrast, implementation of empirically supported therapy in real-world settings may not happen with as much structure or rigor.
I think these are all good hypotheses, but the 3rd explanation is pretty scandalous.