It may come as no surprise to patients and therapists to hear that the therapist matters when it comes to the quality of psychotherapy. This contrasts with the perspective of health systems that often view clinicians as interchangeable; or, to researchers who often view therapist differences as a statistical nuisance for which one must control. Findings from psychotherapy outcome research have “caught up” to anecdote and patient and therapist intuition. Accumulating research convincingly demonstrates that differences among therapists matter for understanding and predicting patient outcomes (Boswell et al., 2015).
The term therapist effects simply describes the empirical observation that different therapists obtain different average outcomes with their patients. In other words, there is a systematic effect of the therapist on patient outcomes beyond what is explained by differences among patients or the type of treatment delivered. Far more than a statistical artifact, therapist-level outcome differences have real-world implications for psychotherapy quality and the likelihood that a given patient will experience a benefit (Johns et al., 2019).
What Do We Think We Know about Therapist Differences?
Soberingly, research demonstrates that approximately 4-5% of therapists have patients who have reliably negative treatment outcomes (Kraus et al., 2011). However, when patient reported outcomes are assessed multidimensionally (e.g., with an outcome monitoring tool that assesses symptoms and functioning across multiple domains; e.g., Kraus et al., 2005), therapist-level patterns of outcome strengths and weaknesses with different types of patients emerge for the vast majority of therapists. Within a given therapist’s caseload, we might find that a therapist is particularly effective at reducing depression symptoms, while not particularly effective at addressing substance use problems. A different therapist may possess a different pattern of domain-specific strengths and weaknesses than another (Kraus et al., 2011; 2016).
A therapist’s pattern of relative outcome strengths and weaknesses is also reasonably stable (Kraus et al., 2016; Constantino et al., 2021). Research has demonstrated that a prior track record of effectiveness in a particular outcome domain is a valid predictor of subsequent effectiveness within that domain; similarly, ineffectiveness in a particular outcome domain is a valid predictor of subsequent ineffectiveness within that domain.
Potential Implication of Therapist Outcome Differences as a Function of the Patient
Research findings related to therapist outcome differences do not imply that “some therapists are good while other therapists are bad.” Rather, certain therapists appear to be particularly effective with certain types of patients and less effective with others. However, stakeholders are generally unaware of therapist outcome track records and therapists tend to be poor judges of their own effectiveness (Walfish et al., 2012). This leaves it up to chance that a patient will end up working with an empirically well-suited therapist. Conversely, there may be promise in intentionally matching patients to therapists based on empirically determined historical outcome profiles, or at the very least, steering patients away from therapists with a track record of ineffectiveness.
In fact, we have now tested this in a Patient Centered Outcomes Research Institute (PCORI) funded randomized controlled trial (Constantino et al., 2021). After a baseline phase of collecting therapist outcome data and “teaching” the outcome-match algorithm, new routinely presenting patients were randomized to an empirical match condition or to a condition with case assignment as usual. Notably, we did not manipulate the actual therapy provided; therapists were asked to simply do what they normally do. As expected, we observed a statistically significant and large in magnitude difference in the rate of patient improvement that favored the match condition, as well as a statistically significant and large in magnitude difference in post-treatment severity that favored the match condition. Results supported the efficacy of an empirical outcome-based match relative to usual practices.
Outcome-Based Matching Matters More for Some than Others
Outcome-based matching is one pathway of actualizing personalized psychotherapy that translates to better patient outcomes, yet there is the potential for enhanced precision. For example, the importance of matching may be relative to a given patient. The magnitude of the therapist effect appears to be larger with more severe patients – that is, the therapist matters more when the patient is more severely impaired (Johns et al., 2019). Therefore, it stands to reason that the match effect will be more pronounced for more severe patients. Conversely, for patients who present with mild severity, they could end up seeing just about any minimally skilled therapist and experience a benefit. There is also evidence that some therapists appear to be particularly effective with racial/ethnic minority patients (Hayes et al., 2015). Consequently, an empirical match may matter more for racial/ethnic minority patients.
We recently tested these “tailoring” factors in the context of the matching-outcome effects observed in our PCORI trial (Boswell et al., 2021). We found that high severity patients experienced greater impairment reduction and lower posttreatment impairment in the outcome-match vs. case assignment as usual condition, whereas low severity patients had comparable reductions in impairment across both conditions. In addition, the match effect on symptom change was more than two timesstronger for racial/ethnic minority vs. white patients. Racial/ethnic minority patients who were treated by matched therapists experienced the greatest amount of improvement, whereas racial/ethnic minority patients who were assigned to a therapist as usual experienced the least amount of improvement. It is worth noting that the main effects revealed that the beneficial match effects on both impairment reduction and posttreatment impairment were still significant and moderately sized for white participants; this interactive effect suggests that although the match effect is present for both racial/ethnic groups, it is significantly stronger for racial/ethnic minority patients.
In closing, therapist effects exist and can translate to actionable strategies in mental health care. Personalization can go beyond treatment matching to also involve therapist matching, informed by more nuanced findings regarding for whom a good matching therapist matters most. Future work should focus on how such evidence-based matching can be implemented in more routine and diverse community mental health settings. To that end, with additional funding from PCORI, our team is about to embark on a large-scale match implementation project that will involve mental health clinics across the United States.
Target Article
Boswell, J.F., Constantino, M.J., Coyne, A.E., & Kraus, D.R. (2021). For whom does a match matter most? Patient-level moderators of evidence-based patient-therapist matching. Journal of Consulting and Clinical Psychology, 90, 61-74. http://dx.doi.org/10.1037/ccp0000644
Discussion Questions
- As a clinician, how do you go about determining your own areas of relative practice strengths and weaknesses?
- Are there particular patients with whom you believe you tend to me more effective, and particular patients with whom you believe that you tend to be less effective?
- What factors do you take into consideration when determining if a new patient will be a good fit for you and your practice?
- Would you ever consider posting or communicating your own outcome track records (based on validated routine outcome measurement tools) to patients and/or referral sources?
About the Author
James F. Boswell, PhD is an Associate Professor of Clinical Psychology at the University at Albany, State University of New York. He is the head of the Practice-Oriented Research Lab and is President of the North American Society for Psychotherapy Research. His research focuses on psychotherapy process and outcomes, measurement-based care, and evidence-based practice implementation.References Cited
Boswell, J.F., Constantino, M.J., Coyne, A.E., & Kraus, D.R. (2021). For whom does a match matter most? Patient-level moderators of evidence-based patient-therapist matching. Journal of Consulting and Clinical Psychology, 90, 61-74. http://dx.doi.org/10.1037/ccp0000644
Boswell, J.F., Constantino, M.J., Kraus, D.R., Bugatti, M., & Oswald, J. (2015). The expanding relevance of routinely collected outcome data for mental health care decision making. Administration and Policy in Mental Health and Mental Health Services Research, 43, 482-491. 10.1007/s10488-015-0649-6
Constantino, M.J., Boswell, J.F., Coyne, A.E., Swales, T.P., & Kraus, D.R. (2021). Enhancing mental health care by matching patients to providers empirically derived strengths: A randomized clinical trial. JAMA Psychiatry, 78, 960-969. doi:10.1001/jamapsychiatry.2021.1221
Hayes, J. A., Owen, J., & Bieschke, K. J. (2015). Therapist differences in symptom change with racial/ethnic minority clients. Psychotherapy, 52, 308–314. https://doi.org/10.1037/a0037957
Johns, R.G., Barkham, M., Kellett, S., & Saxon, D. (2019). A systematic review of therapist effects: A critical narrative update and refinement to Baldwin and Imel’s (2013) review. Clinical Psychology Review, 67, 78-93. 10.1016/j.cpr.2018.08.004.
Kraus, D.R., Bentley, J.H., Alexander, P., Boswell, J.F., Constantino, M.J., Baxter, E.E., & Castonguay, L.G. (2016). Predicting therapist effectiveness from their own practice-based evidence. Journal of Consulting and Clinical Psychology, 84, 473-483. http://dx.doi.org/10.1037/ccp0000083
Kraus, D.R., Castonguay, L.G., Boswell, J.F., Nordberg, S.S., & Hayes, J.A. (2011). Therapist effectiveness: Implications for accountability and patient care. Psychotherapy Research, 21, 267-276.
Kraus, D.R., Seligman, D., Jordan, J.R., (2005). Validation of a behavioral health treatment outcome and assessment tool designed for naturalistic settings: The Treatment Outcome Package. Journal of Clinical Psychology, 61, 285-314. 10.1002/jclp.20084
Walfish, S., McAlister, B., O’Donnell, P., & Lambert, M. J. (2012). An investigation of self-assessment bias in mental health providers. Psychological Reports, 110, 639-644. 10.2466/02.07.17.