Dismaying as it may seem, a success rate of 60% is actually pretty amazing when it comes to treating any kind of human ailment. If a new drug were successful in treating 60% of patients or if it ameliorated symptoms by an average of 60% it would be hailed as a major breakthrough. With respect to harm, a number needed to harm of 4-6 is not that unusual in any type of treatment.

The other question is whether those 15-24% who are worse post-therapy are being compared to a no-treatment control group. (ie were they harmed by therapy or is the natural progression of the condition that a subgroup of its sufferers worsens? What if the no treatment group is 30-40% worse after the same period elapsed?)

As you no doubt know, I fully believe that therapy has the potential to harm. My experience is that there are many people of very questionable competence practicing therapy. I also question, as was brought up in recent thread, whether therapists have a “first do no harm” ethic. As a profession, I don’t think they truly grasp the tremendous potential of a normal, average therapist to cause harm in discharging their duties (like leaving out therapists who are overtly unethical or abusive).

Still, as a stats/epi nerd, I’d argue that the study you cite actually supports psychotherapy as a reasonable treatment modality.