Early Severity Predicts Remission of Cognitive Behavior Therapy for Depression
This project is a very exciting one that was initiated by Cannon Thomas and relies on data from a database of anonymous patient progress monitoring data collecting during routine care. The database was collected by Jackie Persons and Cannon Thomas and others (especially our long-time research assistant Cory Pallatto Hughan) during Jackie’s years at the San Francisco Bay Area Center for Cognitive Therapy. This project has its origins in Cannon’s clinical observation and data produced by other investigators showing that most depressed patients who receive cognitive therapy show a marked reduction in symptoms very early in the treatment and that early response predicts outcome. Our study takes up two questions. First, is it early response that predicts outcome, or is it that the patient who has shown an early response is quite a bit less severely symptomatic after that early response, and so early severity predicts outcome. We also examine whether the effect is true in both naturalistic routine clinical care and in treatment provided in research studies. Our paper, which we have presented at several conferences and submitted for publication, shows that symptom severity at week 4 of CBT predicts remission from depression. Our data show that when patients have Beck Depression Inventory scores in the severe range after 4 weeks of treatment (that is, in session 5), they are very unlikely to show a complete remission of their depression at the end of treatment. The clinical implication of this finding is that when patients remain severely depressed early in treatment it is essential to move quickly and aggressively to make changes in the treatment plan in order to increase the patient’s chances of a good treatment outcome.
The Congruence Study
The congruence study flows out of a previous study that failed. A large group of us collected a lot of data to test the hypothesis that the degree of a patient’s engagement in and use of the feedback form that we use in therapy at the Oakland CBT Center, the CBT & Science Center, the San Francisco Bay Area Center for Cognitive Therapy, and the San Francisco Group for Evidence-Based Psychotherapy, predicts symptom change. Results were negative.
One of our research assistants, Alexandra Jensen, as she worked on the data, became interested in the question of whether congruence between what the patient reported was important to remember in the session and the content of the homework assignment might be important. She and Dr. Persons worked out the idea that congruence between these two things might predict homework compliance. Homework compliance is an important topic because it predicts outcome of psychotherapy, and this finding is especially strong for CBT for depression. As a result, this study has the very elegant quality that it contributes to science and also has direct clinical applicability. Allie Jensen did an outstanding job of moving this study forward, learning to use R and do multilevel modeling to analyze the data, and working with our research assistant Connie Fee to prepare this paper for presentation at the Association for Behavioral and Cognitive Therapies in San Diego in November 2017. The poster Allie produced was gorgeous, and she and Connie did a great job of presenting it at the conference, as this photo illustrates! Allie and Connie and Jackie are now working to write up this project for submission to a peer-reviewed journal.
Study of Skills Learning in CBT with Amy Sanchez, M.A.
Although there is a lot of evidence that CBT is quite effective in treating mood and anxiety disorders, surprisingly little is known about exactly how it achieves its benefits. One hypothesis is that in CBT patients learn skills to change behaviors and thoughts and manage their emotions, and by practicing the skills repeatedly, the person learns skills that help them reduce symptoms of anxiety and depression and achieve their other treatment goals. Other hypotheses include that the therapy helps by increasing a person’s confidence about having skills that he or she can call on when needed, or by changing schemas, or the beliefs that underpin and guide a person’s perceptions of his or her experience.
The beautiful thing about this research question is that it is a question that is of importance both to psychotherapy researchers and to the clinician who is working with patients every day. We all want to know where to focus our energy to provide the most help to our patients. (And of course, the answer might be different for different patients!) Jackie is beginning to collect data to test these hypotheses by designing the progress note she uses in her clinical work and the feedback form she uses in therapy to collect information about skills teaching in the session by the therapist, and about the patient’s skills practice and confidence. Dr. Persons is collecting those data in the course of her clinical work, and asking her patients for permission to place the data in an anonymous database that can be used for later research on skills learning and CBT.
An Idiographic Analysis of Change Processes During Routine Treatment with Cognitive Behavioral Therapy
This study is led by Casey L. Brown, with collaborators Hannah G. Bosley, Kuan-Hua Chen, Robert W. Levenson, and Jacqueline B. Persons
The gorgeous thing about these data is that they were collected (with the patient’s permission), during routine treatment with CBT. Adopting an idiographic approach, this study examined intra-individual processes of change during routine treatment using the case-formulation approach to CBT. The study relies on sophisticated data analytic methods contributed by Hannah Bosley, including dynamic factor modeling and time-lagged regression. The study analyzed routine progress monitoring data collected from a 56-year-old male client receiving CBT for depression, anxiety, and anger problems. The study examined a) whether symptoms declined over the course of treatment, b) whether targeted mechanisms declined over the course of treatment, c) whether mechanisms changed when the therapist intervened to change them during treatment, and d) whether changes in hypothesized mechanisms were associated with later change in symptoms. As predicted, we observed significant change in symptoms during treatment. Further, we observed change in specific mechanisms underlying those symptoms specifically when those mechanisms were targeted in treatment. We also found time-lagged relationships between mechanisms and symptoms over time (for example, reducing the client’s concern over mistakes predicted less anger at subsequent measurements). The findings demonstrate that idiographic research methods focusing on a single case can provide rich information regarding processes of change. Ultimately, better understanding processes of change during treatment allows for the development of more targeted, efficient, and effective interventions. We have submitted this paper for presentation at the Society for Psychotherapy conference in Amsterdam in June 2018 and will write it up for publication.