This must be the most hilariously trolly study I’ve seen in a while.
Understanding the role that therapists play in psychotherapy outcome, and the contribution to outcome made by individual therapist differences has implications for service delivery and training of therapists. In this study we used a novel approach to estimate the magnitude of the therapist contribution overall and the effect of individual therapist differences. We conducted a meta-analysis of studies in which participants were randomised to receive the same treatment either through self-help or through a therapist. We identified a total of 15 studies (commencement N = 910; completion N = 723) meeting inclusion criteria. We found no difference in treatment completion rate and broad equivalence of treatment outcomes for participants treated through self-help and participants treated through a therapist. Also, contrary to our expectations, we found that the variability of outcomes was broadly equivalent, suggesting that differences in efficacy of individual therapists were not sufficient to make therapy outcomes more variable when a therapist was involved. Overall, the findings suggest that self-help, with minimal therapist input, has considerable potential as a first-line intervention. The findings did not suggest that individual differences between therapists play a major role in psychotherapy outcome.
Authors: “Hey psychotherapy people! Yeah you! You don’t seem to do anything of value. Sincerely, science.”, “PS. Maybe get a real job.”
This is on top of loads of data telling us that these people can’t make better clinical predictions than simple algorithms. In fact, algorithms usually beat them.
Of course, the sample size here is not impressively large, and so they might simply have missed a small but real effect. Still, one would expect publication bias to produce them something positive to find, but apparently that was not the case.