Constantino, M. J., Boswell, J. F., Coyne, A. E., Muir, H. J., Gaines, A. N., & Kraus, D. R. (2023). Therapist perceptions of their own measurement-based, problem-specific effectiveness. Journal of Consulting and Clinical Psychology, 91(8), 474–484. https://doi.org/10.1037/ccp0000813
Previous research showed that therapists may be over-confident about their effectiveness. In one study, 91% of therapists felt that they were more effective than 75% of their peers, and 100% of therapists felt that they were more effective than 50% of their peers. These therapist perceptions relative to their peers are statistically impossible. What is the impact of this over-confidence on patient outcomes? In this study, Constantino and colleagues evaluated the perceptions and outcomes of 50 therapists who treated 1,363 patients (about 27 patients per therapist). Most therapists practiced CBT, integrative, or interpersonal psychotherapy. Patients had a variety of problems including depression, anxiety, sleep, social functioning, etc. Patients self-report of symptoms in 12 problem domains (depression, anxiety, etc.) from pre to post treatment. Therapists were asked to report on their perception of their effectiveness by rating from “mostly effective” to “mostly ineffective” for each patient problem domain. Then therapists’ ratings of their effectiveness were compared to their patients’ actual measurement-based outcomes from pre- to post-treatment. In this manner, therapists were classified as accurately estimating their outcomes, or as underestimating their outcomes, or as overestimating their outcomes compared to their patients’ outcome data. The researchers also controlled for case mix (complexity and size of the therapists’ caseload) and patient symptom severity at pre-treatment. For 7 of 12 patient problem domains, at least half of the therapists overestimated their own effectiveness. For the remaining 5 domains, most therapists accurately estimated their effectiveness. There were no domains for which most therapists underestimated their effectiveness. In 11 of the 12 patient problems, therapists were no better than chance at predicting their own effectiveness compared to patient rating. However, therapists who underestimation their effectiveness had patients with better outcomes at post-treatment, and this effect was stronger for patients with higher pre-treatment symptom severity. Conversely, therapists who overestimated their outcomes had patients with worse outcomes, and this effect was stronger for patients with higher pre-treatment symptom severity.
Relative to patient self-reported measurement-based outcomes, therapists tend to be over-confident about their effectiveness. This over-confidence has implications for patients’ outcomes. Patients of therapists who show more humility about their effectiveness (tend to be neutral or under-estimate their effectiveness) have better outcomes. This is particularly true for patients with more severe symptoms. One of the few therapist factors known to predict better outcomes is professional self-doubt - the willingness to critically evaluate one’s practices and to continually improve one’s skills. Self-doubt may allow therapists to be more alert to potential signals that the therapy is off course, that they might have committed a micro-aggression, or that there was a therapeutic alliance rupture. Humility might allow the therapist to flexibly respond, change course, or repair the therapeutic relationship when necessary.