Personalized Psychotherapy: Effective Methods for Tailoring Therapy to Individual Client with Dr. James Boswell Part 2

Special Guest:
Dr. James Boswell

Society for Psychotherapy Research

Keywords

low motivation, readiness to change, diminished outcome expectations, non-directive approach, motivational interviewing, humanistic principles, ambivalence, transdiagnostic cognitive behavioral therapy, tailor therapy, therapeutic process, client-therapist relationship, matching, therapist selection, outcomes, preferences, routine outcome monitoring, therapist effectiveness, strengths and weaknesses

Takeaways

• Low motivation and readiness to change are related concepts in therapy.
• Therapists should have discussions with clients about their concerns and difficulties with change.
• Motivational interviewing principles and humanistic principles are important in working with clients who are low on motivation.
• Therapists should explore ambivalence about change and therapy, and work on building motivation and hope.
• Transdiagnostic cognitive behavioral therapy combines different disorder-specific treatments into one approach, providing a more efficient and evidence-based therapy.
• Therapists should tailor their approach based on the client's primary problem and explore their expectations and concerns.
• Bringing the therapeutic process into the room and using the client-therapist relationship can be effective in therapy, even when using a cognitive behavioral approach. 
• Matching clients with therapists based on characteristics like race or gender doesn't consistently lead to better outcomes.
• Matching black clients with a strong preference for black therapists can have a positive impact.
• Routine outcome monitoring feedback can be used to identify patterns of therapist effectiveness.
• Matching clients based on therapist outcomes can lead to more effective therapy.
• Therapists have concerns about how outcome data will be used and the potential consequences.
• There is a need for resources and training to help therapists improve in specific areas.
• The use of technology and AI could enhance process research and therapist training.
• The field needs to explore how to make use of session-level data to understand therapist effectiveness.
• Therapists and clients generally support the idea of using data to guide therapist selection.
• There is pushback from therapists who question the validity of outcome data and concerns about being pigeonholed.
• Resources for further exploration include articles by James Boswell and Michael Constantino on context responsive integration and the book 'Responsiveness in Psychotherapy' edited by Watson and Wiseman.

🎞️ Video version of the show@PsychotherapyAppliedPsychology on YouTube
🛜 Check out the website: Listen to every episode on your podcast player of choice

Connect with Dan
☏ Leave a voice message on Speakpipe
🔗 LinkedIn
🐥 @TheAPPod on twitter
📬 TheAppliedPsychologyPodcast@gmail.com
📸 Instagram

Transcript

Broadcasting from the most beautiful city in the world, I'm your host, Dr. Dan Cox, a professor in counseling psychology at the University of British Columbia. Welcome to episode number 14 of Psychotherapy and Applied Psychology, where we dive deep with the world's leading applied psychology researchers to uncover practical insights, pull back the curtain, and hopefully have some fun along the way in an attempt to bring new ideas to practitioners, those training to be practitioners, and the applied psychology curious. I'd love to connect with you.
Send me a text or leave me a voice message. You can find links in the episode show notes to do both of these. And if you're willing, I'd love to know your name and where you're from.
Today, I am thrilled to present part two of my conversation on tailoring psychotherapy to the individual client. My guest is a faculty member and clinical psychology program director at the University at Albany State University of New York and past president of the North American Chapter of the Society of Psychotherapy Research. This is part two of my conversation on tailoring psychotherapy to the individual.
In this conversation, we discuss addressing low motivation and diminished outcome expectations, flexibility and responsiveness, incorporating different therapeutic approaches, and client-therapist matching. We begin our conversation with my guest answering my question about how to tailor therapy when a client has low motivation for change. So without further ado, it's my pleasure to welcome Dr. James Boswell.
You could quibble about whether or not they're the same thing, but thinking about in terms of readiness to change, you know, as one, as one, sort of, another way of framing it. You know, and so, you know, in that regard, you know, it's having a, you know, taking a less directive approach and having a discussion with the client about the concerns that they might be having about the change, or what's making it difficult to sort of consider or get excited about changing, or thinking, or sort of reaching a more hopeful stance instead of expectations about change, you know, and so whoever is, you know, consistent with the broader motivational interviewing principles, or you can even argue just broader humanistic principles. It's, you know, anybody who's low on motivation or low on readiness, you're probably not going to get very far if you try to plow ahead with being more problem focused and action oriented.
You'll need to take a bit more time talking with the client to try to better understand what might be explaining or behind that that lack of readiness could be ambivalence about change, could be ambivalence about therapy. Like, I think this could be helpful, but there's another part of me that thinks that, you know, I've tried a million other things that they haven't worked. It's just going to be another thing that I tried that doesn't work.
You know, every client is going to be a little bit different. And so you have to explore those things. And part of it, too, is if you're taking a non-directive approach, and you're trying to explore this and make the ambivalence more explicit, and sort of put it out there and talk about it together.
But I think also the role of the therapist is to also then, in the context of that discussion, also try to build motivation through trying to enhance expectancies and provide a sense of hope that something could be different, that maybe this is worth a shot, and really try to identify and align with and expand on that part of the client who's the part of them that is willing to give it a try, or that thinks that it could be useful. So without getting on too much of a tangent, I've done some work in transdiagnostic cognitive behavioral therapy, and one of the manuals that I've used and done some research with, it's a modular-based approach. And so the modules include a lot of the standard cognitive behavioral therapy interventions and principles that you could use with anything.
Before you go on, what is transdiagnostic cognitive behavioral therapy?
So for many years, the cognitive behavioral therapies that have been developed and tested, they've focused primarily on single disorders or single diagnosis categories. And so the idea is that we have all these different disorders within the DSM, diagnostic and statistical manual, and that we should be developing specific interventions that would fit or target that particular disorder. So you'd have cognitive behavioral therapy as a broader theoretical paradigm or system, but then you would develop a particular form of cognitive behavioral therapy, manualize and codify it in a way that would be targeting specifically something like generalized anxiety disorder or panic disorder or social anxiety disorder or obsessive-compulsive disorder.
So you have individual treatments that are, let's say, CBT-oriented, but they're targeting these particular disorders. And for decades, most of the research in cognitive behavioral therapy focused on more, these are treatments that are geared toward very specific problem domains. Not to say that this is the controversy for many years in the field, the inclusion-exclusion criteria in randomized clinical trials, trying to enhance internal validity, and are you excluding clients who maybe do have the primary disorder that you're interested in targeting in the study, but they have other problems too?
And to what degree are those secondary or co-primary or whatever? And carefully selecting people to really more narrowly fit this particular diagnostic classification. And so a lot of research done focusing on just those single disorder treatments showing that they're on average effective, or at least many of them are.
But with the thin slicing of the diagnostic manual, and you have, as you mentioned before, that even if you look at cognitive behavioral therapy for a panic disorder, you might have half a dozen different manuals or treatment guidebooks describing what should be done following a cognitive behavioral therapy model for people with primary panic disorder.
If you actually look at the manuals, sort of look at what am I being asked to do as a cognitive behavioral therapist sitting in front of somebody who has primary panic disorder versus sitting in front of somebody who has a primary generalized anxiety disorder. The principles, negative reinforcement, positive reinforcement, conditioning, they're all consistent across... It's all cognitive behavioral therapy, so the theoretical foundation is identical.
And then if you actually look at the interventions that I'm expected to deliver as part of this broad treatment package, cognitive restructuring, behavioral activation, exposure, experience monitoring, et cetera, not a ton of difference between the two. And so there's other more basic psychopathology research to support this point too, but there's a general recognition that there's too much splitting, there should be more lumping, and that maybe it would be better in the long run and more consistent with the evidence if we actually tried to combine what we know the best available evidence for the cognitive behavioral therapy model and its application for these different anxiety and say, depressive disorders. And so that we have kind of one main source or guide for how to deliver cognitive behavioral therapy with a broader group of clients.
And I'll just sort of add that, and I could get into it if you're interested, but is that even if this more transdiagnostic, so this model that's intended to be applied to people with varying presenting problems or kind of following the same general scheme, even if it's no better than the existing single disorder protocols, at the very least it starts to get us past this problem of, okay, I'm a clinician who wants to keep up with the evidence base. I see a variety of people who come in with commonly presenting problems, various anxiety disorders, depressive disorders, looking at my bookshelf, and I have a different manual for each one of those individuals. And by the way, most of those people actually come in with multiple problems, multiple anxiety disorders or anxiety and depression.
So how do I become an expert in learning, delivering those individual treatment models? If there's a way to put all those things together into one guide, then it becomes much more efficient in terms of training and dissemination and implementation.
Yeah, when I first learned cognitive processing therapy for PTSD, I remember going in and being excited, I'm going to learn cognitive processing therapy, and I have this two-day training and all that, and about an hour into the training, I was like, I already know all this stuff. This is not a critique at all of the cognitive processing therapy people or the training or the manual, which I think is actually a very useful manual, but that was my clinical experience of like, oh, this is it? This is it?
Two days? That seems like a lot for this, because I had already had so much training in cognitive therapy that it was just like, oh, okay, I got it.
And I think that there are meaningful differences, I think, in some of these manuals, and certainly meaningful differences depending on the presenting problem. So it's not to say that, you know, again, getting back to Taylor, it's not to say that, let's say I have this transdiagnostic cognitive behavioral therapy model that I'm using with most of my clients, what I do, you know, is probably still going to vary somewhat, depending on whether or not the person sitting in front of me, their primary, explicit problem, the main reason they're here, the thing they want to work on is their panic attacks, you know, versus a client who has social anxiety or OCD. I think this is one of the critiques of the transdiagnostic approaches, is that it in some ways prevents you from tailoring.
And I think that it's, well, let's at least start with some of these core foundational principles that we know are broadly applicable, but we're not saying that it's inconsistent with any sort of personalization. You want to do that. And just as another quick example, we've done some work with intensive eating disorder treatment centers, and we've used this transdiagnostic cognitive behavioral model as a foundation for those.
But because it was written initially for anxiety and depressive disorders, everything's applicable to eating disorder patients, but the examples, the homework assignments, some of the literature review, the rationale, none of it was really tailored toward people with a primary eating disorder. And so, yeah, obviously, you want to adapt. You know, to a particular population, again, the core is the same, but it's going to be more suitable if you make certain adjustments.
And so, it's not safe. We can't tailor. But getting back to the original point was that the first module in this transdiagnostic treatment is focused on motivation enhancement.
And when we would do drawing largely from motivational interviewing research and theory. And there is that conundrum, like when we would give workshops, and we would introduce the MI to start this introductory workshop, but then clinicians understandably would ask, what if you never reach a point where the client's willing to do the therapy? And it's like, I'm still stumped by that question.
It's like, if you're trying to, if you have gathered, you know, based on various evidence that the client is not ready to change and is not motivated to be in therapy, the therapy could be trying to unpack that, you know, and hope that, you know, it might take us a little while, but eventually you'll get to a point where you're ready to change and we can shift what the goal is, or you might say, it sounds like maybe therapy is not, this isn't the best time for you, and maybe if you think about it a little bit more and change your mind, you can come back, give me a call, you know, I'll see you again, you know, maybe provide some guidance, help, help materials or something. But, you know, so I think like if you're a cognitive behavioral therapist, who's used to very brief treatments, sort of following a particular schedule, not that I believe that cognitive behavioral therapists in the quote unquote, while, you know, adhere so rigidly to treatment manuals, even if they're following them generally. But I think it's like, well, what if by session two, the client's not motivated, then what do I do?
Do I just move on to the next thing? It's like, well, unless you're in a randomized clinical trial where you are required to move on to the next thing and you're being rated for adherence, you probably don't move on to the next thing. You probably sit with the client and try to unpack the ambivalence.
And if you reach a point where motivation is not enhanced, so to speak, then you decide that maybe this isn't the right time.
So you talked about it a little bit, but what's the difference between low motivation and diminished outcome expectations?
Yeah, well, I think motivation is multi-dimensional on its own. So outcome expectations are being right that I'm entering into this or potentially entering into this activity, psychotherapy. The degree to which I feel like this is something that's going to be helpful to me, that we will do this and at the end I will experience the benefit I'm looking for.
So I expect this to be helpful.
And clients vary in their level of expectation, positive to negative. And presumably, right, that if you are ambivalent or concerned about the effectiveness of this activity, or part of you really has doubts that it's going to be useful, so your outcome expectations are low, then that would co-vary with being less motivated, because I'm not willing to put in effort and pursue something that I don't believe is going to be particularly useful. So I think they go together, and generally the case that if my outcome expectations are lower, then that's probably going to come along with just being less motivated.
So what would be, if we made it extremely practical, if I'm a therapist and I have a client who has low outcome expectations, would be something that you would suggest that I would do?
Well, I would want to explore what's contributing to that. It could be related to previous attempts at therapy that didn't go well, so they had a negative experience with previous therapists, and part of them expects that it's going to be the same. So what might be contributing to those expectations?
It could be just in talking about every course of therapy is different. Let's talk a little bit more about what you did in that previous therapy. How is what we will do together potentially different, or you tell me how you want it to be different, or we can make a difference.
Provide more evidence about the rationale for the therapy that you're expected to provide and what the evidence would show for it. So I have a little bit of OCD, and I also have social anxiety, and that makes me more depressed when my OCD gets periods where it's worse, and I get kind of demoralized, and I think I'm just too complicated for this. I just have too many problems.
But to be able to say that, well, actually, getting back to transdiagnostic cognitive therapy or other approaches, clients often come in with multiple problems, and there are probably ways for us to figure out how these problems interact and influence one another, and there actually is research evidence to show that people who come in with that kind of pattern of difficulties can benefit from therapy. We can talk a little bit about what that would look like. I can't guarantee a positive outcome, but generally speaking, I have a sense for how to be helpful to you, and I do think that the odds are probably pretty good that you'll find something about this helpful.
And so you explore, provide a rationale, examine the evidence together, share that with them. Maybe they don't know what the outcome data would show.
You've probably never worked with somebody as sick as me. Obviously, you wouldn't want to minimize that perspective on the part of the client, but talking through maybe that's not exactly the case. Let's talk about that.
It seems like in terms of tailoring, in terms of the if-then approach, I think there's the spotting the ifs. Spotting the alliance rupture, the low motivation, the diminished outcome expectation, etc. And then the then oftentimes seems to be to bring it into the room, to process it.
No, I think so. I mean, I think the getting back to the training conversation we were having earlier, that it's especially more novice trainees, right? They want to, things that they can really latch on to, you know, they like the if-then framework because they're like, oh, I can learn to attend to certain things, and I know that if I see it, then I should do X, you know, and especially more cognitive behaviorally oriented trainees, like having the roadmap, you know, of the manual.
I like using manuals too, but it takes experience, I think, to learn how to flexibly apply these principles or strategies, kind of the difference in some ways between adherence and competence. And so I think the super anxious want to be helpful, want to know exactly what to do. It's not just trainees, it's all of us, frankly.
But that telling a trainee, in particular in supervision, that I know that you know that on page 10, that explains what to do on session three, that this is kind of where you want to go, what would be helpful for this client. But part of being a responsive therapist is exercising flexibility, meeting the client where they're at, maintaining a sense of collaboration, that you're not wasting time if you actually slow things down and try to explore your client's experience. That is still therapy.
And I think some people think that it's supportive chit-chatting.
But that's not what we're talking about here. There's in fact a conceptual framework and some evidence to show that you need to hit the brakes and shift for a little bit and gather more information and explore before you know what you need to know in order to make a decision about how to move forward, or not.
I think that's a really interesting point that you bring up about the idea of talking about the process of therapy, this perception of this isn't therapy. This is something that in my training, I feel like one of my criticisms of the more cognitive behavioral approaches is that on average, there are certainly exceptions to this, don't use the relationship, the interaction between client and therapist explicitly as much as they could in the therapy process itself. And I've learned so much from the more psychodynamic people or the interpersonal folks, and then just of how to use this me you in this space right here, right now.
And even if I'm using a cognitive approach, that I can still use this really effectively to affect change using a cognitive approach.
Absolutely. I was fortunate because in my graduate training, the faculty came from identifying with multiple orientations or different orientations. It was a program that was very welcoming toward psychotherapy integration.
And so, as mentioned before, even the integrative therapy that was being tested in this trial was all about, how do you develop a cognitive behavioral therapy that codifies some of those more inner personally oriented dynamic therapy principles to navigate the relationship and metacommunicate with clients about how things are going. We also had practica and supervision that was more by the book, but then there was a general openness. And I think that's somewhat unique.
I think there are a lot of programs that don't quite... That's something that us who are more cognitive behaviorally oriented, we don't tend to get exposure to that perspective until later on, like internship, post-doc, just additional client experience.
Yeah, I've always found it interesting that, I don't know, well, this may be a deeper conversation or longer, maybe a different conversation. But I think that there's so much value there in those sort of, I don't know, ways of being, I don't know, is that too philosophical? Ways of being in the therapy room from those other approaches that can really be effective and useful for influencing, or sort of even when working from a cognitive behavioral approach.
And so I really, I often encourage folks to get that diversity of training or experience. And I always say, sometimes we'll get students who are very sort of CBT-oriented. And I always say, if you have a supervisor who's not, take advantage of it.
Like, just pretend that you are psychodynamic for that semester, for that year. Just dive into it and learn so much from it, because then you can totally be a cognitive therapist, you know, and ascribe to that and be a zealot and all that. That's great.
And you will, but you will extract so much that you'll be able to apply as a cognitive therapist by just sort of diving into this rather than being resistant to it.
Yeah, absolutely. I agree 100%. And it's, yeah, openness, sort of that general, having openness and flexibility is so important.
And thinking about the people who are known as cognitive therapists or cognitive behavioral therapists who are widely regarded as experts and do a lot of training and consultation, their best sellers, so to speak, are books that are all about interpersonal process in cognitive therapy and responding to resistance in cognitive therapy. The real classics out there that are really useful are just saying, okay, you're a cognitive therapist doing your cognitive therapy, but we've all been there where clients are ambivalent about change or they're resistant or they lack motivation or there's a rupture, it's like, how do we talk about these things? Even the tongue and cheek in some of the writing is really so interesting to me because even saying like, well, our dynamic colleagues would call this kind of transference, and I see no reason why we can't use that term, too, because that's exactly what we're seeing in the room with our clients when X happens.
And I do think that cognitive behavioral therapy often, unfortunately and unfairly, gets sort of painted as a reductionistic and kind of, you know, paint by numbers. I think it's actually much more complicated than it's sort of portrayed. But it is true that we could sit down with the trainee and walk through the steps of actually setting up an exposure hierarchy or something like that and practice with various cases and vignettes.
And you can get that down. Like with some practice and some feedback, you can get that down. But it's when you're actually sitting with a client and the client tells you that this isn't working, that doesn't fit, they see it differently, they don't want to do it.
You know, that's all relational, process-oriented. You know, it's like that's where you really need to bring in a high level of this interpersonal skill in order to, again, explore and figure out how to be helpful for the client in that moment. And if you're a more cognitive behavioral theory therapist who just thinks that that's a waste of time or something like that, then I think you're really going to struggle to be helpful with your clients.
Those are the things that people really want to latch onto and learn more about. How do I navigate those impasses?
Yeah, I think that that's a really good point. I see that too, where folks who don't have much training in the cognitive behavioral approach do perceive it in a very reductive kind of a way. It's just, you know, you identify the thoughts, you challenge the thoughts, and then session's over, and you do that for 10 sessions with some homework in between, and then we're good to go.
And it's sort of like, no, I mean, that's bad cognitive therapy. And you brought up the exposure hierarchy thing, which is great, you know, where you sort of identify what are the specific, you know, what are the, from the least feared to the most feared stimuli, if you will, that even that, like that, for some clients, that can be a very quick process. But for other clients, that could take forever, because it's so difficult for them to do that.
And so you have to be so interpersonally skilled with many clients to be able to do it and do it in a way that it's not just identifying the hierarchy, then that is going to be so much the focus of your work with that client for X many weeks. And so it has to be done in a way that they're going to have buy-in to it. And obviously you can adjust it, but they're going to have buy-in to it for the next, you know, what could be months of work.
To do so effectively for, hell, even for the average client, I'm going to say for the median client is, you know, it does take a lot of skill, a lot of art, a lot of flexibility to be able to do that effectively. Okay, so matching.
So in one of the courses I teach, which is sort of this basic teaching, basic helping skills, where I have usually either advanced undergrads or sort of like early grad students, this topic comes up often, and I'm often not too adept to answer it, but my general sense of it is with matching clients and therapists, so that a certain type of client might do well with a therapist who's similar with them in certain characteristics, race, ethnicity, age, gender, whatever, etc., religion. My understanding is that the research on this sort of suggests like there's some studies that have maybe found a little bit there, but most of them have found that it doesn't really matter so much.
Do you have a sense of this?
I mean, my understanding is very consistent with yours. I mean, I think if we go back to preferences, in general, there's this small but significant effect of preference accommodation on outcome across the board. The preferences could be for, I really would prefer working in this particular way of therapy, or it could be, I really prefer to work with a therapist with these characteristics.
It could be interpersonal personality, but it could be identity related. And so, in general, if there's an opportunity to accommodate a preference, if a client has a strong preference, then if we can accommodate it, that's great. And generally speaking, it's better to do so, it's not always possible.
But that is different from looking at actually, when you have a match on those variables, does that actually translate to better outcomes? And for the most part, it's very mixed, it's not clear that it does. The one group where it seems like there might be a benefit, even though it's not huge, is that black clients with a strong preference for working with a black therapist, if those individuals are matched, then there's some research to show there's kind of a bump in effectiveness.
But for the most part, the matching on that doesn't really yield much. But again, that doesn't discount trying to accommodate a preference if you can do so. I mean, in our work, again, not that any of this is irrelevant, but in the work that we've been doing, we've been more focused on problem-building and more focused on outcomes.
And so thinking about the best place to start here. Like I mentioned before that you had also brought up that one of these if-then markers has to do with responsiveness to reaching out to monitoring feedback. And so we've done some work in this area with a particular multidimensional tool.
So it's an outcome measure. It's a self-report measure that patients complete on a weekly basis. But rather than just ask about general symptoms of functioning, it has targeted questions that cut across different commonly presenting domains.
Sorry, I have a feed being happening in the living duck. But things like depression and anxiety and suicidality and sleep, as well as interpersonal functioning, work functioning, quality of life. And so when you're using those tools on an individual client basis, that accumulates a lot of data.
Like on the one hand, the routine outcome monitoring feedback work is very focused on individual patients and tracking their outcomes and making decisions based on that. But then you end up with a very large data set that includes not just patient level outcomes, but then also you can look at therapist level outcomes. Because in most trials or data sets, you have therapists who are working with multiple patients.
And so we can apply some statistical modeling to try to isolate how much of the outcome differences accounted for by differences between therapists. And so that's what's called the therapist effect. It's a statistical observation or phenomenon.
And just to see how much of this difference is accounted for by differences between therapists. And then we can start to get a sense of, in very large samples or within particular systems of care, certain therapists seem to be very effective with their clients. Certain therapists seem to be much less effective.
Most therapists, again, if you're assessing outcomes across multiple domains, tend to have a pattern of relative strengths and weaknesses. With a multidimensional measure that assesses outcomes across different dimensions, we can actually look to see that certain therapists are really good at getting clients less depressed. Those same therapists might be actually really bad at helping clients with clinically severe substance use.
Whereas another group of therapists seem to be really excellent with clients who come in with clinically elevated substance use problems. But clients who come in with work functioning impairment, let's say, that tends to not move very much for them. And so if we can identify relative patterns of outcome strengths for therapists, then kind of the next question for us was, okay, we can sort of identify these different patterns of more versus less effective therapists in these different domains.
But then the question is, if this is actually going to be harnessed in some way, is it actually a stable observation, so to speak? So is there stability and effectiveness in one's effectiveness profile over the course of time? And so we've done some research with secondary data and some prospectively collected trial data, where we get a group of therapists with a bunch of routine outcome data with their patients, and then we apply a classification algorithm to sort of label them as effective, neutral or ineffective in these different domains, sort of look at a cohort of patients and their outcomes.
We say, okay, you're classified as effective in this domain, this domain, this domain, average or neutral in these domains. And then we take a completely separate, subsequent cohort of patients, and then we apply the same classification, and we see are they actually correlated? Is there predictive validity to the initial classification?
And across all of the domains, we have statistically significant medium to quite large associations in the effectiveness profile between the first cohort and the second cohort. And so very consistent with the other research showing soberingly that clinician outcomes don't tend to improve over time. We also find that these profiles don't tend to change much over time.
And so if that's the case, then maybe we can actually then start matching clients based on presenting problem domain in therapist outcomes. And so in that instance, even though we might ask a new patient, do you have a preference for working with a therapist with these sorts of characteristics? Okay, great.
Let's take that into consideration. Can you complete this multidimensional tool at the start of the therapy? They complete the tool.
It gives us a sense of what their most clinically elevated domains are. You say, okay, your most elevated domains are depression and sleep. Which therapists in our system have we identified as being really, really excellent at helping clients improve their mood and sleep?
Assign that new, that client to one of those therapists. We also in the background have a small set of therapists who have reliably shown that their, their clients get more depressed with them. At the very least steer that new client away from those therapists.
And so we've found that in a prospective trial that actually matching based on outcomes, not necessarily characteristics, but actual outcomes leads to more effective therapy. When we compare this sort of match approach to referrals and case assignments as usual with therapists. And so a long way of saying that it's like the other sorts of matching were made very relevant.
And there are ways to make these decisions even more personalized. But most of the work that we've been doing has been actually focused more on the outcome itself.
So then, and I know that we're going to sort of speak beyond what the data can tell us for sure. But then I would assume then that you're using a particular assessment, which has particular categories of challenges, problems, difficulties that clients are going through. And so that's what makes sense.
And so that's how you're looking at it. Presumably, this phenomena would generalize to other frameworks for thinking about problems. So that, you know, I mean, obviously, one would be using, you know, the psychiatric diagnosis was sort of an obvious one that one might think about.
But it could be any other sort of ways of framing. I'm sure some would be better than others or more useful than others. But would you assume that that's true?
I mean, I think the idea can be generalized to different constructs and measures. Absolutely. I think that focusing on outcome is, you know, because we've been doing work with this particular tool and implementation, outcome monitoring, that's kind of where we started.
But certainly other measures and other constructs of interest could be relevant. Because again, even though I guess one potential benefit of this measure is that it's not just looking at symptoms of area, it's also looking at functioning across several domains. But you could be working in a population where those sorts of outcomes are less important or value or relevant.
And so some alternative measure could be used, definitely. And not to say that this isn't worth doing, but I think one distinction that we make is that if you're using a measure that's more global, that's intended to give you a snapshot of overall severity or impairment, you can still look, see that there are differences among therapists, and you can still identify therapists who are really good at getting clients in general or global distress down.
But the level of precision tailoring sort of stops there, right? And then you're kind of immediately at this issue of like, okay, this group of therapists seems to be pretty effective on this general indicator. Any new client that shows up or makes a phone call, refer them or assign them to those therapists, this group of therapists not doing so well, don't assign or refer to those individuals.
And then it's such a general indicator. It's to say to those therapists, the outcome data show that actually you're generally ineffective with most of your clients. You need more training.
Have you considered?
But if you have more fine-grained information, then you might have a little bit more to go on. Not to say that there's a ton of evidence to support it, but let's just say as an example, if somebody is really struggling with dealing with sleep, then you might say, okay, here are some evidence-based approaches for dealing with insomnia or sleep-related problems. Let's make sure you get more training and consultation on working with sleep.
I have a little bit more to go on in terms of how to help you if you want to work with clients who have sleep problems.
Well, so that was a question I was going to follow up with, which is this makes sense in terms of directing clients towards or away from, but is there, I guess, a follow up study would be, so if we identify the therapist who stink at this particular thing, is there a way to help them stink less? Is there a way to help them get better or not?
We don't know. I mean, I think people have ideas about how you might do that. We actually, years ago, submitted a grant proposal to a particular federal funding agency to do just this, to test it, to collect data, look at benchmark therapist outcomes, and then based on areas where they're not doing so well, randomized people to get this training versus this training, and then see if their outcomes improve.
We did not get funded. The main piece of feedback was that therapists would never agree to participate in the study. Because they don't want to hear that they're not doing so well in certain things.
But so we've done some, right now we're in the process of a large scale implementation project where we're implementing this routine outcome monitoring tool and matching in a large system. And we've collected a bunch of survey data. We've also done some interviews with clinicians and staff.
And they are concerned with how the data will be used. They're concerned with, like, if you find out that my supervisor or my administrator finds out that I'm not doing well in certain areas, what are the consequences of that going to be? But not saying that they don't want to know.
In fact, they're generally pretty clear that it's better to know than not know. I'm intrigued by the idea. It's just if you're going to use this information, then you need to provide resources so that I can get better at doing these things.
And so, in interviews, therapists might say, if you find out that in our clinic, in general, we're not doing so well on X, like, could you make sure that our administrator brings somebody in to do some workshops on X so the whole group gets the benefit? But then also, you could have more, again, more tailored training opportunities where you, as a specific clinician, have this relative pattern of strengths and weaknesses, so let's try to focus the training on those more personalized domains. But then we start bumping up against just the limits of what we know about.
Training. And if we stopped at just, okay, this particular intervention for insomnia has some evidence base behind it. You're struggling with insomnia, so we're going to train you on how to do that.
With a lot of support, consultation and feedback, there's research to show that clinicians can learn to do new things, but then we have to actually see over time if that translates to better outcomes. And even if it does, unless you have a really carefully designed study, it might be hard to know exactly if the improvement is due to that specific training that you just conducted.
Well, and I was thinking, too, getting back to what we were talking about earlier with sort of this if-then approach, that if you found that I was really bad with clients with, I didn't help very much with mood, so that their depressive symptoms didn't increase or decrease, that, yeah, this is time intensive, but one could theoretically could take, if there were, in a perfect world, if there were videos of sessions, one could take, okay, let's find a recent client you've had over the last six months who fit this profile, who didn't get better, and someone could sit down and watch through some sessions. And it might not, alternatively to, let's learn this whole treatment for insomnia, it could be that this helper just sucks at dealing with alliance ruptures, right? And so that it could be that alliance ruptures tend to show up more often in, or maybe a certain type of alliance rupture, because it hits me as a clinician in a certain spot, so I'm just not as effective.
And I'm sure that's true for all of us, that there's going to be certain things that we're just not as good at. And so identifying those, it could just be like, oh, we just need to work at these types of alliance ruptures, which tend to show up with clients who have these types of needs, and then we could see a change. Now, of course, that's a very sort of generous in terms of the time and energy and motivation to do those things.
But it's such a great point. It's such a great point because it could also be the interaction. It's kind of what you're saying.
It's like maybe it isn't, I mean, it's the problem domain to some extent, but it's not a sort of intervention basics issue. It's more about navigating the process of clients who have this sort of problem, or maybe there are, say, non-diagnostic factors that tend to co-vary with that particular process or something that are more likely to come up. And so you'd be really focusing on the wrong thing, if the issue is really something else, like alliance negotiation with this particular type of client.
Yeah, that's really... I mean, it's been a major struggle with trying to... Because so much of this work has been large datasets, very high-level systems focused, and knowing that there's something going on there, like something's differentiating therapists, at least on outcomes, but what's accounting for it?
I mean, there's other research to show that maybe there are some candidate factors that help explain differences between therapists, but it's been really a struggle to do a deeper dive and actually gather session-level data to try to really explore and unpack what the therapists are doing or not doing in the session. That's been a big limitation of this work, because the obvious question is, what is making these therapists really, really effective at addressing insomnia versus these therapists who not only are not just neutral or average, their clients' sleep gets much worse when they see them. What's the difference between these two?
And you actually have to look at what they do to know.
Yeah, maybe as we move forward in these sort of advancements in these, I don't know, AI is sort of the phrase. I hate to use that phrase as it could probably be better conceptualized as something else. But that when we have the ability to be able to sort of automate a lot of this, because as you know, the amount of time and energy it takes to code individual sessions just on one construct is a lot.
And so to be able to code it on all the possible constructs, to be able to get the, you know, but I think that I feel very optimistic about sort of the future of process research and being able to do it with these, you know, advances in technology.
Yeah, I mean, it's, yeah, you're with the how it's so intensive that if you, even if it's theoretically driven, hopefully it would be like, there's almost a needle in a haystack quality of just like, oh, we're going to focus on this construct. And then you don't really find much and it could be a million other things. But yeah, I mean, one benefit of just, again, I mean, there are obviously complicating factors and ethical questions associated with it.
But if we were to just collect a lot of information, you know, based on audio, let's say from a session, you know, there are methods, you know, now that can help us sort through patterns and identify potentially relevant information, you know, relatively quickly that, you know, spend an entire career just focusing on a couple of constructs, you know, to do the same amount, essentially. But so, yeah, it's a little bit of a mess right now, but I think it's inevitable, actually, if the field is going to progress and process research is going to move forward, then we have to start figuring out how to make use of these tools.
Just to clarify one point you made a few minutes ago about black clients and black therapists, about there being a small but significant increase in terms of outcome. Did I hear you right that that's for... So it's not simply black clients and black therapists, but it's black clients with a preference for black therapists, is that right?
And then actually working with that black therapist.
Right, right, of course, of course. So maybe a way to reframe some thinking about matching rather than thinking about a client therapist who match on a particular variable. It's rather a slight reframing is to think about, as you sort of said, think about client preferences.
What actually do they think, you know, what actually matters to them? And being able to accommodate that could result, so this for any number of characteristics, could account for a small improvement that does seem to be significant. Or I'm sorry, that while being small, it is still significant.
So when you can, you should. But it doesn't seem to be huge. Okay, great.
Yeah, preference, identity, salience, even though it's not really measured often, you know, just think about how, like, oh yeah, if you happen to have a therapist who, you know, checked out these boxes, that'd be great. But if not, then no big deal, versus I absolutely have to work with a therapist with this identity, because it's really important to me. You know, so it's what gets baked into the match and how it really can vary.
It's funny, I have one of my graduate students, she's just finishing up a paper where she measured identity salience, and we were looking at it with cultural humility, and it didn't matter at all. At all, like we were shocked. But anyway, that's a story for another day.
So what pushback, if any, have you gotten for your work?
Yeah, continuing to talk more about the matching work. I mean, what's been interesting, again, we've done at this point over the last decade or so, we've done multiple studies involving patient interviews and surveys, therapist focus groups and interviews, and I would say from the patient or the client perspective, when we talk about these things, they tend to be on board. They tend to actually like the idea.
They are intrigued by using outcome data to help inform therapist selection. And they also use other research to show that clients actually don't mind completing measures generally either. They tend to have positive attitudes toward measure of base care generally.
But therapists are a little bit more mixed, and maybe it's not surprising to hear that. Because I think the pushback comes out of a variety of levels, including the things I was mentioning a moment ago. What are the implications at the level of employment, of having this information and making this determination of classifying me as ineffective in particular areas?
The particular areas in the profile piece is really important here, and it's something that we try to stress with implementation too. It's that every therapist has a relative pattern of strengths and weaknesses. Nobody's telling you that you're a bad therapist.
It's just that relative to what we're seeing at the average level in terms of outcome, there are some areas where you're underperforming, and you can make use of that what you like. You can either refrain from taking on clients who have those problems as primary clinical characteristics, or you can get more training and come back. And so, within a system, who is going to see this information, and how is it going to be used to make decisions?
That is a concern. And then also, even if therapists are, this is an interesting idea. So far, when I get people assigned to me, my caseload in this clinic, or when I take referrals, I'm not really basing it on anything other than what I think I'm good at or what I'm interested in.
So if you're going to bring some data on board to help guide these decisions, I think that that's probably worth doing. But then there are concerns about what, it's pretty interesting how consistently this very specific term comes up in interviews, is the concern about pigeonholing. So one might classify in this classification process as much more dynamic than static because we're always collecting data and adjusting it.
But if we sort of freeze it, and we say, okay, these therapists are really good at working with depressed clients, getting depressed clients less depressed, every new depressed client, and you get a lot of them in outpatient therapy, gets assigned to those therapists, don't assign them to these other therapists, you're only going to send me depressed clients. I'm going to have a practice that's all depressed clients. Like that will get boring very, very quickly.
And then you're not going to send me these other types of clients. If the goal is for me to improve, how am I supposed to improve if I'm never going to see a client with this other problem again? That seems inconsistent.
I mean, again, the match approach that we take is more complex than I've laid out so far, because clients usually come in with multiple problem areas. And so you might be assigned a client based on being really good at all of those things, or you're the only available therapist that's really good at at least one of those clinically severe things. And so that's why you're being...
So there's still going to be variety in your case. The one other thing to note, before I ramble too much, is that clinicians, again, when you explain the intention of this and the general principle of, how do we bring data on board from your own patients to guide decisions, because what we're doing up to this point is based on convenience and self-identified interest on the part of therapists. Do you agree that this is something, that having some information to guide these decisions is better than none?
Most therapists will say yes. But then you still have some therapists, and again, I'm not necessarily... I can understand the perspective because there are a lot of researchers who have this perspective too, that I'm not willing to even entertain the matching that you're suggesting because I don't think that the outcome data that you're using to inform the match is valid at all.
And so for clinicians who just don't believe in that you can measure outcomes, they're not on board with any of it. And so that's not very many, but there are some who have that perspective.
If there's somebody listening who, or for the person listening, who would like to learn more about thinking about tailoring therapy to the individual client, will it be one or two resources, books, articles, YouTube videos, trainings, anything that you might think to send them to? And then I can, I apologize, I didn't prepare you for this ahead of time, but then I can link them in the show notes as well. So that way they could just jump right there.
Well, I mean, I don't know how helpful this is in the moment, but I mean, as you mentioned, we've written some articles on this topic, colleague Michael Constantino, who is the person who coined and started to develop this context responsive integration framework. And so I don't have the citation memorized, but if you are looking for articles with Constantino and Boswell and cognitive responsive integration, you sort of put that in your favorite search engine, you will have at least a few articles pop up. We talk about this.
There's a chapter that I co-authored on responsiveness in cognitive behavioral therapy, and then a chapter that I was the lead author on that's responsiveness in integrative therapy in an edited volume by Watson and Wiseman, so Gene Watson and Hottest Wiseman, on responsiveness in psychotherapy. That's a really nice book because it does kind of walk through, we haven't talked all that much about responsiveness, even though it can be kind of a slippery construct. But it is ultimately about how do you, what would guide decision making in a way that's personalized and relevant to the client in this moment or in this treatment.
And so each chapter kind of takes a slightly different perspective or focus on this, what is ultimately a tailoring question. So that book, look up some of those articles, try to think if there's anything else that's, I mean, as in just thinking out loud, I'm not necessarily going to suggest these very large handbook volume chapters that are narrative reviews or even meta-analyses about certain literatures, because we've written some of these on like patient variables, you know, that could be moderators or tailoring variables, and you review the literature, and there just isn't a lot to hang your hat on, you know. And so I hesitate, if somebody really wants something to stick their teeth into, I hesitate to suggest this thing.
But I think if you find that responsiveness book from Watson and Weisman, you can find some of these context-responsive integration articles from Michael and myself, and it's a good start.
Great. Well, Dr. James Boswell, thank you so much for your time.
That's a wrap on part two of my conversation with Dr. James Boswell, whom I am most appreciative. And as I said at the outset, I'd love to hear from you, so please reach out by sending a text or leaving a voice message. Until next time!