Preventing Psychotherapy Dropouts with Client Feedback

“I need to you to slow down.”

“I was sad and you cut me off.”

These words of dissatisfaction are from my clients. They weren’t easy to hear, but they have changed how I practice psychotherapy and have significantly reduced my dropout rate.

Back to Top

Anne: A Case Study

I had been treating Anne, a Latin-American woman in her early 20s, in psychotherapy for six months. She presented with weekly panic attacks, daily cutting, severe sleep disturbances, a range of somatic symptoms that she attributed to her anxiety, and persistent interpersonal difficulties. She presented as attentive and likeable, though beneath her mask of smiling and compliance she clearly hid a tremendous amount of pain. Anne has a history of sexual abuse by multiple family members over a six-year period starting before age four. Her mother had been a prostitute for most of Anne’s life, and both her biological father and stepfather are in prison for sexual assault. Despite these and many other challenges, Anne demonstrated tremendous resiliency and had just graduated from college with a very strong GPA.

Anne had been in individual and group therapy for much of her childhood and teens, but by her own report she had never really tried to make it work. After graduating from college, Anne decided she wanted to find a solution to her anxiety, sought out individual therapy, and found me.

Anne’s treatment progressed well at first. In the first few months her panic attacks stopped, her general anxiety decreased, she stopped cutting, her somatic symptoms decreased, and her sleep gradually improved. Anne’s interpersonal difficulties, however, persisted. We had been digging into that material for a few months but had made little progress. In fact, her social and romantic life was getting worse. Anne was becoming restless and frustrated. I pulled out my two favorite “getting therapy unstuck” tools: consultation groups and additional training. Neither helped. As a dynamic therapist, I knew what I was supposed to do: work in the transference, bring insight to the dynamics in the room, monitor my counter-transference, and above all hold the frame. But the frame of a therapy case cannot be stronger than the frame of a therapy practice, and mine was starting to splinter.

Back to Top

Existential Threat

In the same month that my treatment of Anne was getting stuck, I had two new clients drop out after one session in the same week. I knew about the research that we are all told in graduate school about how the modal number of psychotherapy sessions nationwide is one, and how not every client and therapist is a good match, and yada yada. But for a new therapist trying to build a practice during a recession, having two new clients drop out in one week is an existential threat. I decided something had to change.

On my commute home one evening that week, I listened to a recording of Scott Miller’s presentation at the 2009 Evolution of Psychotherapy Conference regarding his pioneering work on feedback-informed psychotherapy. Scott got my attention when he referred to dropouts as the “largest threat to outcome facing behavioral health” in the United States and Canada. He was talking about my practice! I realized that I was not the only therapist with a dropout problem, and there was no reason to hide it out of embarrassment. I resolved to seek counsel from my colleagues and mentors.

Back to Top

The Ubiquitous Scourge

In the first, difficult year of building my private practice, I ate a lot of lunch. Networking lunches are like lottery tickets: one in ten results in a few referrals, and every referral was worth its weight in gold in that difficult first year. I enjoy networking lunches, because it’s fun to meet senior clinicians and hear their war stories. They tell me that they enjoy the lunches because they get to pass on the gift of mentoring that was once given to them. Senior clinicians are a generally calm, relaxed and self-assured bunch; they have established referral sources and can easily afford to lose a client here and there. Want to make some highly regarded pillars of the therapeutic community stop eating their free lunch and sweat a bit? Ask about their dropout rate. It’s as if you’re asking what sexually transmitted diseases they may have. It’s not polite. Never mind that dropouts are one of the ubiquitous scourges of our profession, affecting all diagnoses and treatment modalities. Therapy dropouts are the dirty secret of our profession: everyone has them yet few want to talk about them. Unfortunately, avoidance has not proven to be an effective solution to the problem. With few exceptions, the overall psychotherapy dropout rate is as bad now as it was fifty years ago, despite decades of treatment research and empirical certification.

Back to Top

What Counts as a Dropout?

For 2010, the overall dropout rate for my private practice was 37%. Unfortunately, it is hard to know whether this number is good, average or poor, because there is no general consensus in the literature on what exactly constitutes a “dropout.” The average psychotherapy dropout rate has been reported to be from 15% to 60%, or higher, depending upon whether you define dropout as quitting therapy before all treatment goals were achieved, terminating without the therapist’s agreement, or a variety of other definitions. For my own practice, I define dropout as any time a client terminates therapy without telling me that they are stopping because they have achieved enough positive results. I chose this definition because I think it points most directly to the problem I want to resolve: clients who could benefit from more therapy but choose to not be in treatment with me anymore. Of course, this definition is not precise and won’t work for all therapists. If a client terminates due to factors that make continued treatment impossible, such as moving out of town, then I do not count it as a dropout; but if the given reason is that he or she cannot afford therapy anymore, but isn’t interested in talking about a sliding scale, then I do count this.

Of course, there are many reasons a client may drop out. Most of the research on dropouts has focused on what we call client factors, such as the client’s diagnosis, demographics, rate of progress in therapy, etc. But this research doesn’t help my dropout problem because I’m trying to keep my practice full, and I don’t have the luxury of excluding clients who are at high risk of dropout. So instead I have to focus on therapist factors: what can I change about how I work to reduce my dropout rate.

Back to Top

Insisting on Feedback

“Of course I ask for feedback from my clients. I do it every session!” Every therapist believes they ask for client feedback. True for you too? Then tell me why your last three dropouts happened. Sure, we ask for feedback, in the same way that my previous dentists asked—as an offhand, pro-forma fly-by at the end of the root canal. “Was that ok?” And the information we get is usually as meaningful as the effort we expend asking. “Yeah, that was great,” or “You’re a great therapist,” or “I’m really feeling better.” Vague and general; even worse, polite. Just enough for the client to think that they have satisfied the therapist and just enough for the therapist to keep the specter of dropout in the closet. It’s a mutual con-job—a wink and a nod to accountability. But if we don’t embrace accountability in the therapy room, then it will make itself known in dropouts.

Sure, some clients are tripping all over themselves to give you feedback. Sometimes you can’t stop the feedback. But those aren’t the clients I’m worried about losing to dropout. Maybe some therapists are able to get meaningful information through informal soliciting of feedback, but I’ve found the hard way that if I don’t make a Big Formal Procedure out of it, I end up with empty, vague generalities.

Another fruitless session had just ended with Anne, and I was pretty sure that she was about to drop out. I handed her a feedback form and asked her to complete it. She looked at the piece of paper, snorted and said, “Are you kidding me?” As a beginning therapist, I have a lot of practice hiding my nervousness. I replied, “I need your feedback in order to learn how to help you better, but also to become a better therapist overall, so I appreciate your time and candor in filling this out.” Anne snorted again, rolled her eyes, and completed the Session Rating Scale, an ultra-brief tool that measures the working alliance along four dimensions. She handed the form back to me and I saw that our working alliance, as I would have guessed, was a sinking ship. I asked what specifically I could do to help her better. Anne replied, “You could listen.”

I said, “More specifically, tell me how I don’t listen and how I can help you better.”

She gave me the look clients give you when they’re not sure if you really mean what you say or if you’re just doing a canned intervention. “You understand me thirty percent of the time,” she said, visibly angry. I asked for an example. “When I mentioned my cousin you cut me off,” Anne said. “That was important.”

I couldn’t remember Anne mentioning her cousin. “What else?” I said.

“You tuned out two or three times this session. I can always tell you’re tired when we meet this time of day.” I thought I had managed to hide my mid-afternoon fatigue.

“What else?”

“There are times when I am sad that you really don’t understand how I’m feeling—even though I can tell that you think you do.”

None of Anne’s feedback struck me as accurate. Above all, I pride myself on accurate empathy. What kind of therapist am I if I don’t feel a client’s sadness?

Back to Top

Four Rules for Receiving Feedback

We all have areas of known weakness. Take cultural diversity, for example. I am a straight, white, middle-aged male. Anne is a young bisexual Latina. I would expect for her to tell me about culturally based misunderstandings. This would be ego-syntonic for me and not cause anxiety. But tuning out or missing sadness—that’s not me!

The feedback I get from clients that is confusing or seems inaccurate is the most important feedback I get. Why is it that we trust our supervisors to point out our blind spots, but not the people who are actually in the room with us? It’s odd how we spend so much effort and money getting feedback from peers and experts, yet so little effort on getting formal feedback from our customers.

I’ve come to see that there were two major problems with how I had been using feedback. First, my collection of feedback was pro-forma. I wasn’t invested in getting it, and my clients could tell. Second, I interpreted the feedback. I conceptualized it as part of the therapeutic process, which meant that it was ultimately about the client, not about me. Of course, getting and using feedback affects and informs the therapeutic process. I needed to learn, however, to set aside the process for a moment to accurately hear the feedback as it pertained to me.

Since then I have developed a four-step feedback rule. First, I make a Big Deal out of it. I use a paper form (the Session Rating Scale) because the act of pulling out the paper and pen serves as a symbolic shift in focus away from the client’s process towards my performance. If a client always gives me high marks on the form, or responds with platitudes like, “Tony, everything is great,” I’ll say, “Well, there’s always something I can improve. Can you give me one or two specific ideas on what I could be doing better?” In therapy, it’s all about the client. In feedback, it’s all about me—I’m downright selfish!

The second rule of feedback is that I don’t interpret. If I make the feedback about the therapeutic process then I am missing the actual feedback. As a dynamic therapist, all my training was telling me to interpret Anne’s response as transference or a projection: she was reliving her past pathological attachments in our relationship. But I’m convinced this approach would have caused Anne to drop out, because she would have seen (correctly) that I was ignoring her.

Scott Miller calls this kind of attribution “burden shifting”—when we misattribute our mistakes to client factors. He warns therapists that blaming dropouts on client demographics or diagnostic categories can block our insight into our own mistakes.

The American Psychological Association is moving towards requiring trainees to learn how to collect clinical outcome data. Likewise, Michael Lambert¹ and others have developed tools to predict and reduce dropout by tracking clients’ session-by-session clinical progress throughout treatment. This data is valuable, but still focuses on client factors, and thus can miss important information that only the client has on what the therapist is doing wrong. I need to know my part in the story so I can stay ahead of potential dropouts. Without session-by-session feedback, when a client drops out, it is already too late to find out why.

As therapists we claim clinical legitimacy by using empirically certified treatments. We advertise our professional trainings and certifications proudly. But just as important are our personal treatment data, including our dropout rate, which we generally hide in the closet. Krause, Lutz and Saunders² have argued that instead of having empirically certified therapies, we should have empirically certified psychotherapists. As public health providers, assessing outcome is an ethical responsibility. If we continue to hide to our mess then we run the risk of others exposing it for us. (For example, teachers’ unions across the country are getting clobbered for their resistance to incorporating meaningful outcome evaluations into their work.)

Back to Top

Incorporating Feedback

How do I actually use feedback? Sometimes it is easy. For example, in response to Anne’s feedback, I moved her appointment to a time of day when I wouldn’t be tired. (Now I use her previous time for a midday nap, so other afternoon clients are benefiting from Anne’s feedback as well.) Other feedback can be harder to use, especially when it is about my own unconscious behaviors. Anne insisted that I cut her off when she had brought up her cousin, but I couldn’t remember doing so. Likewise, I had no awareness of avoiding her sadness. While I did want to take her comments seriously, I also didn’t want to automatically assume her perceptions were correct.

However, feedback that points to my unconscious behaviors is also the most valuable. This is the third rule of feedback, which is the hardest rule to follow: to focus most on the feedback that seems inaccurate, confusing, or anxiety-provoking. This is where the treasure is buried.

When I’m unsure about the accuracy of the feedback I am getting, I use a strategy I call perspective triangulation. First, I videotape my sessions with that client and review the video myself. I then review it with colleagues in consultation groups. Comparing the perspectives of the client, myself and my colleagues usually results in a definitive answer.

In my experience, the client’s perceptions are correct at least two-thirds of the time, and I make consequent course corrections in their treatment. It is important to note, however, that even when I think the client’s perceptions are incorrect, I still have to substantively address their feedback, or else there is a growing risk of dropout.

My review of the video showed that, yes, I had cut her off. Colleagues in a consultation group watched the video and pointed out multiple instances where Anne was about to have a rise of sadness, but I had blocked her sadness by refocusing on her anger. (Later sessions revealed that the two were in fact connected, as her sadness was about being unable to protect her cousin from abuse.) This was the hardest feedback for me to receive; I never would have believed it, had it not been clear as day on the video. Investigation of videos revealed that I had an unconscious pattern of re-directing from sadness with a range of other clients in addition to Anne. I never would have found out had I not insisted on feedback.

The fourth step in my feedback process brings it back to the client. If I agree with their comments, then I make appropriate course corrections in our work. If I disagree, then we discuss our different points of view. Either way, I make sure to be clear and transparent in my process, and to let clients know that I take their feedback seriously. So in this case Anne and I had a discussion about her feedback. I agreed to be more attentive to not cutting off her sadness. She agreed to let me know, in the moment, if she saw me doing it.

I was trained to get a review of my clinical weaknesses from my trainers and supervisors. Now I also get it from my clients. They have given me an amazing gift: an empirically validated list of my clinical weaknesses. I can’t think of a better resource to prevent dropouts.

Now, six months later, Anne has made significant progress on her interpersonal challenges. She has improved her relationships with friends, roommates and employers. She started setting firm boundaries with previously abusive family members. Her sleep, anxiety and somatic symptoms all continue to improve. Every session Anne teaches me how to better help her.

Before using feedback, I had one to three dropouts per month. Since getting serious about feedback, I’ve had only one dropout in over three months. While this is too soon to draw definitive conclusions, the results so far are very encouraging.

The client sitting across from me knows something about my dropout problem that I don’t. All I have to do is ask, and listen.

Back to Top

2011 Update

 I am pleased to report that my dropout rate for 2011 was 18%, one-half what it was in 2010. I'm confident that getting serious about client feedback contributed to this improvement. This raises the question: how low can a dropout rate realistically go? Besides improving as a therapist, what else can help lower the rate further? (One of my clients recently suggested offering coffee in the waiting room for night sessions!) Hopefully we will find answers to these questions from future research.


Back to Top

Footnotes

1. Lambert, M. J., Harmon, C., Slade, K., Whipple, J. L., & Hawkins, E. J. (2005). Providing feedback to psychotherapists on their patients' progress: Clinical results and practice suggestions. Journal of Clinical Psychology, 61, 165–174.

2. Krause, M.S.; Lutz, W. & Saunders, S.M. Empirically certified treatments or therapists: The issue of separability. (2007). Psychotherapy: Theory, Research, Practice, Training. 44, 347-353.

Back to Top

Further Reading

“When I’m good I’m very good , but when I’m bad I’m better”: A New Mantra for Psychotherapists. by Barry Duncan, PhD and Scott Miller, PhD.