Client feedback counteracts therapist over-confidence
The skill of obtaining feedback from clients about their session is experiencing a resurgence of attention and popularity, thanks to common factors advocates like Scott Miller and Daryl Chow. However, obtaining feedback has always been fundamental to cognitive behaviour therapy practice – so much so that it was one of only 11 areas of skill rated for adherence in the original cognitive therapy rating scale. I regret that I was both fortunate to have had a supervisor (Brian Johnston) who modelled seeking feedback at the end of sessions and yet rapidly neglected to emulate this once I began working unsupervised. I am grateful to have rediscovered feedback, with renewed appreciation for its importance, later in my career. I doubt I am the only clinician who has struggled to maintain soliciting client feedback on sessions. Whether because of defensive avoidance when the alliance and outcomes aren’t solid, or boredom and laziness when our clients continue to be seemingly satisfied with us, asking for feedback is an easy habit to get out of. So, this week’s study might motivate those of us who qualify as filthy backsliders, to get back on the wagon.
In a study from the Netherlands, Janse and colleagues (2023) examined data from 38 therapists and 843 adult outpatient clients who had taken part in a randomized controlled trial where cognitive behaviour therapists either (1) provided feedback using Miller’s Session Rating Scale (SRS), Outcome Rating Scale (ORS) and their Feedback Informed Therapy (FIT-Outcomes) software or (2) did not. Previous analyses of their data showed that overall, adding the FIT feedback to CBT did not improve client outcomes but did reduce dropouts. In the current paper though, the focus was on whether certain therapists might benefit more or less from using feedback. They considered therapist personality (Big-5 traits), propensity to rely on their own self-evaluation or external feedback, self-efficacy and attitude toward feedback.
Neither therapist age, nor years of experience, nor gender showed any association with therapy outcomes. The Big-5 personality trait of Openness to experience was associated with better client outcomes. Self-efficacy was associated with worse outcomes and dropout rates. Readers of mhy blogs will no doubt already be familiar with the fact that psychotherapists fall prey to the Dunning-Kruger effect; in this study, no therapist thought they were below average and there was no correlation between therapists’ assessment of their own effectiveness and their caseload’s improvement rates. To illustrate the degree of overconfidence we’re talking about, therapists with high self-efficacy estimated that 84% of their clients recovered after treatment, yet in reality only 52% had. This study supports a common finding – one I tell my trembling students in their first placements all the time – that your confidence has no bearing on your effectiveness. If anything, low confidence can be a sign of appropriate humility.
Arguably the most interesting finding was the interaction between therapist self-efficacy and the use of feedback, on client outcomes; Therapists low in self-efficacy had better client outcomes without the FIT feedback system, whereas therapists with high self-efficacy had better client outcomes with the FIT feedback system. The proportion who achieved clinically significant change was not significantly different between therapist types; those with low self-efficacy achieved a 55% recovery rate compared to 51.9% of those with high self-efficacy.
The pernicious effects of overconfidence were evident in dropout rates too: those with high self-efficacy had almost twice the dropout rate as those with low self-efficacy (14% v 8%) and this time, the use of FIT feedback - throughout the whole sample - made no difference. It should be noted that more experienced therapists had less drop-out than less experienced therapists and the analysis of the effect of FIT feedback adjusted for years of experience. Interestingly, a posthoc analysis looking at just the group of therapist-client dyads where FIT feedback was used on more than 90% of sessions showed a significant association between feedback condition and dropout, suggesting that feedback might influence dropout if used consistently and investigated with a large enough sample.
The study was limited by a low sample size to detect some of these effects so we can be more confident in the relationships that were found than we can be confident that the absence of relationships in this study means these relationships do not exist at the population level.
The authors presented two possible explanations for the main finding (interaction between therapist self-efficacy and feedback): 1) therapists with low self-efficacy might not be able to handle a high frequency of feedback and so do better when this is not collected, or; 2) feedback helps correct the unrealistically optimistic biases of therapists with (too) high self-efficacy. The design of this specific study couldn’t conclusively support one over the other. In my arguably irrelevant anecdotal experience, I don’t find therapists who lack confidence (pretty much everyone I work with at the start) are too fragile to handle frequent client feedback. If anything, seeking client feedback helps therapists rapidly correct misunderstandings, assuage client fears, and brings conversations forward that might otherwise fester and lead to dropout or apparent client resistance behaviours. Conversely, I have met several overconfident therapists who protect themselves from finding out what their clients think of their sessions. So, I tend to favour the authors’ second explanation for why FIT feedback might enhance the outcomes of people with high self-efficacy.
Clinician implications
Therapists – especially confident ones - should adopt a systematic practice of obtaining feedback on the client’s outcomes and satisfaction with sessions, every session.
Feedback can be sought at the start of the session as a bridge from the previous session, in the final 10 minutes of the session, and/or whenever needed throughout the session to check for alliance ruptures.
The FIT system offers one option for outcome tracking: https://www.fit-outcomes.com/
Any outcome measure appropriate to the clients you work with that has meaningful information about reliable change and how the client is functioning compared to healthy and clinically severe populations can be useful (e.g., DASS, PHQ-9 etc).
The SRS offers standardized questions for evaluating the major parameters of the working alliance: agreement on tasks, goals, and bond; however, any questions that help the client explore potential misunderstandings, misgivings, talking at cross-purposes or other concerns, will probably achieve the same positive effect (See Box)
For the original article, go to: https://onlinelibrary.wiley.com/doi/10.1002/cpp.2828
REFERENCES
Janse, P. D., Veerkamp, C., de Jong, K., van Dijk, M. K., Hutschemaekers, G. J. M., & Verbraak, M. J. P. M. (2023). Exploring therapist characteristics as potential moderators of the effects of client feedback on treatment outcome. Clinical Psychology & Psychotherapy, 30(3), 690– 701. https://doi.org/10.1002/cpp.2828
BOX: Benefits of routinely exploring feedback
Socialises clients to working collaboratively. Clients are more likely to do homework and make lasting changes if they and the therapist share treatment goals and agree on ways to reach them. Having the client reflect on and express how relevant and productive the session was in helping them achieve their goals is an act of collaboration. Inviting their feedback socialises clients to take an active (rather than passive) role in therapy; it facilitates them “buying in” and “taking ownership” of their therapy so they believe in the rationale and treatment plan and want to put it into action.
Helps identify ruptures early. Sometimes therapists get the sense the session didn’t go as well as it could, and they’re right. Raising this helps correct course before the client leaves. For example, T: “When I asked you to do the thought record, I thought I noticed a shift in how you felt about the session. Am I reading too much into that, or was I picking up on something?” C: “Well, I’ve tried writing down my thoughts before, but it’s never helped. I thought ‘is this the only thing we can do in therapy?’ T: “It’s so good you told me that. Of course, you would be feeling despondent about working with me if you thought I was just going to do something that hasn’t helped before! We probably need more time than we have now for me to show you how I think this could help…Could we agree to leave the thought record this week and then next session make sure we both agree on what seems to be causing your anxiety and therefore what would be good to do about it? Are you willing to come back so we can have that discussion?” C: “Sure”.
Invites gentle opportunities for empathic confrontation when needed. Suppose you have felt frustrated that a client has spent a great deal of the session insisting on talking on tangents. Then, at the end of the session, you ask, “To what extent did we cover the issues you hoped to talk about today?” And suppose the client says, “I was hoping to talk more about my anxiety” which they avoided by insisting they describe the state of an unrelated family drama. This provides an ideal opportunity to invite the client to take responsibility for the use of session time. For example, you could say, “I’m glad you said that – I was a little worried about that too! What could we do next session to make sure we don’t get off track?” or “I noticed that too. Can we agree that next session, we will start by assessing your anxiety and make sure we only discuss other topics if we have time?” or any number of other instructions or experiments to have a more satisfactory session. Client complaints about not having time to talk about key issues also provides a compelling rationale for homework such as keeping diaries – T: “Time does go very fast in here so the more work you can do on your own before you get here, the better the chance we can get to the points you really want us to discuss together. What if you kept a record of the key details about the situations that bother you so I can quickly read them before we start getting into deciding what to do about them?”
Prevent therapist anxiety and second-guessing. Sometimes therapists get the sense the session didn’t go as well as it could, and they’re wrong. Instead, they use emotional reasoning and let their own anxiety levels inform their impression of how well the session went. The feedback portion of the session is an ideal time for the therapist to test their hypotheses about whether they had offended, confused, disappointed, or otherwise had a negative impact on the client. Most of the time, the client will indicate that the therapist’s fears are unfounded, e.g., T: “I wondered whether you felt I was too controlling when I asked you to tell me exactly when you would find time to practice guitar this week”?” C: “No. I probably would’ve ended up not doing it if you hadn’t pushed me to think through exactly when I could fit it in”. In these cases, the therapist can save themselves a world of stress and ensure they don’t abandon an effective treatment plan unnecessarily. However, even if the therapist was right, raising it at the end of the session allows the therapist to repair any rupture and improve the chance of the client returning, e.g., T: “I wondered whether you felt I was too controlling when I asked you to tell me exactly when you would find time to practice guitar this week”?” C: “Well…maybe a bit. My mum used to nag me to do things all the time and I ended up hating it”. T: “I appreciate you telling me that and I’m sorry if I was getting too pushy. I’ll try to work on that next time. If I check in with you from time to time to see if I’m coming across as too controlling, will you let me know?” Instead of worrying about what they had done, that might have put the client offside, the therapist can find out and stick with an effective approach or make informed decisions about adjusting course.