How Therapists Fail: Why Too Many Clients Drop Out of Therapy Prematurely

Earn CE Credit

• The "Infallibility Error"
• The “Pathology Orientation” Error
• Emphasizing Therapeutic Techniques Over Relationship Building
• The Homework Assignment Trap
• A Final Note
• Public Domain Assessment Tools

Depending on which study you read, between 20 and 57 percent of therapy clients do not return after their initial session. Another 37 to 45 percent only attend therapy a total of two times. Although many factors contribute to premature client termination, the number one cited reason by clients is dissatisfaction with the therapist. The problem of the “disappearing client” is what Arnold Lazarus has called “the slippery underbelly to the successful practice of psychotherapy that is almost never discussed in graduate programs or medical schools.”

As clinical supervisors of interns at a university community clinic, we are painfully aware of the high rate of client dropout, and thus the idea for our book How to Fail as a Therapist was born. What we found in doing the research for the book is that high dropout rates are not just common amongst interns, but are equally prevalent among experienced therapists regardless of training and clinical orientation.

When clients drop out early, everyone loses. We clinicians lose a chance to help someone in need and our wallets and reputation suffer as well. The consequences for clients are even more dire. Those clients who drop out early display poor treatment outcomes, over-utilize mental health services, and demoralize clinicians.

Now the good news (after all, therapists should be optimistic): there are a number of well-researched strategies which have been proven to reduce dropout rates and increase positive treatment outcomes. For example, in one study a simple phone call to confirm a client’s first appointment resulted in a two-thirds reduction in dropouts. Unfortunately, it is often labor intensive to seek out and review much of the relevant research because it is scattered throughout the literature--a journal article here, a chapter in a book there. And, unfortunately, most mental health clinicians, with and without a PhD, rate reading research as a very low clinical priority.

Thus, a major task in writing the book How to Fail as a Therapist was to assemble, organize and condense the vast body of research addressing therapeutic effectiveness. Of the 50 therapeutic errors described in the book, here we present five of the most common ones made by clinicians--both beginners and “master” therapists.

The "Infallibility Error"

One of the most distinguishing characteristics of therapists who have low dropout rates is that they actively seek feedback--both positive and negative--regarding the effectiveness of their clinical work. On the other hand are those therapists who believe that after years and years of study, comprehensive exams, postgraduate supervision, and licensing exams, they do or should have all of the answers to clinical matters. So when their clients voice concerns about their progress, or worse yet, when they drop out or deteriorate under the therapists’ care, there is a tendency to avoid accepting responsibility for committing a possible therapeutic error. It is easier to point the finger elsewhere: “maybe the problems were too severe”; “the patient was not ready or willing to change”; there was too much transference operating.” The possibility for rationalization and denial is endless. These explanations, even when partially valid, may soothe the ego, but they protect clinicians from engaging in an honest and comprehensive exploration of what might have gone wrong in a particular case.

A group of interns were asked to describe a case in which a client of theirs terminated early in therapy. One intern described the case of a 10-year-old male client, who had been referred by his teacher because he seemed disconsolate over his parents’ divorce. When, in the first session, the intern probed about the effect of the parents’ separation, the client became emotional and wanted to change the subject. The intern persisted, however. The client stood up, tears falling, and refused thereafter to return to therapy. The supervisor responded to the case presentation by emphasizing the need for therapists to be very cautious during early sessions, particularly when eliciting difficult material from clients. Before the supervisor could get very far, the intern interrupted by stating: “I am already discussing this case with my other supervisor, so I probably shouldn’t get input from both of you.”

Clearly, this intern was desperate to avoid facing the possibility that he did not handle the case as delicately as perhaps he should have. None of us really relishes the idea that we may have blundered, but if we deny this possibility, we deny ourselves the chance to grow as clinicians.

One way to avoid the infallibility error is to seek feedback from clients who have dropped out prematurely. Arnold Lazarus describes in his book, Multimodal Behavior Therapy, how he has gained great insights by writing “early terminators” and suggesting that they come in for a “feedback session” for which he doesn’t charge. In one such case, a client reported that she felt the therapist had not been sympathetic when she was recounting the loss of a beloved pet. The therapist apologized for the insensitivity and the client decided to continue in therapy.

One crucial statistic to keep is mind is that the majority of clients who drop out do so after the first or second session. Thus, we must elicit client feedback, positive and negative, early on to head off any misunderstandings or negative feelings about the therapist, the therapeutic process or the therapists. Clients can be asked directly at the end of the first session if they feel therapy is on track and if they feel liked, understood and respected. Asking for direct feedback may feel a little awkward; however, a little awkwardness is better than losing a client before he or she can be helped.

The “Pathology Orientation” Error

In the field of psychotherapy, the term “The Bible” has become synonymous with the publication known as Diagnostic and Statistical Manual. This definitive compendium of emotional disorders was first published in 1952. Since that time, the Manual has gone through a number of revisions (four major and several minor ones) and has continued to add new diagnostic categories. In addition, it has really bulked up over the decades, growing from a mere 138 pages at the outset to over 800 pages in its most recent incarnation.

Currently every student entering the field of psychiatry, psychology, social work or counseling is required to virtually memorize the DSM-IV-TR, and thus professionals in our field have greatly increased their knowledge base of diagnostic criteria, demographics and prognoses of emotional disorders. Alas, these advances have a downside as well: it has created an overemphasis on pathology to the near exclusion of what is healthy, resilient, and capable in the clients that we treat.

At the same time that the fields of diagnosis and assessment were becoming more sophisticated, an alternative view of human potential was also advancing. Theorists such as Carl Rogers, Abraham Maslow and Victor Frankl were among the forerunners of those who tended to take a broader view of the client, looking beyond pathology toward human capability. Milton Erickson’s work, which emphasized client resources, was in the vanguard of this new perspective.

Following Erickson’s lead, a number of other clinicians and researchers have explored the idea of utilizing client strengths as a resource in the treatment of emotional problems. Narrative Therapy avoids the exclusive focus on problems and pathology by instead exploring clients’ alternative stories--occasions in which healthy, productive behaviors were enacted instead of the usual counter-productive responses.

Ryan was described as “incorrigible” by his teachers. He spent as much time in the principal’s office as he did in the classroom. His main transgressions revolved around aggressive and bullying behavior. Ryan’s counselor applied a narrative approach by first asking Ryan about his “problem story”--the things that get him in trouble. They then gave a name to his problem story--“Mr. Trouble.” In addition to gathering the nasty details of his misbehavior, the counselor also inquired about occasions when a different Ryan, a kinder Ryan, surfaced. The question itself seemed to shock the 10-year-old. However, after reflection he confessed that on occasion he had shown care to his younger brother when he was ill, or was lonely and needed a playmate. The counselor then asked follow-up questions to explore the way “Kind Ryan” felt after demonstrating care to his brother.

“What did you think of yourself for being helpful to your brother?”
“How did your brother respond to your help?”
“What did your parents think of you?”
“What does it say about you that you show care to your brother?”

Unfortunately, despite the advent of “positive psychological” approaches to therapy, we have been programmed to look more at what clients are lacking and less at client strengths. Most intake forms have a space in which the client’s clinical diagnosis is supposed to be entered. To avoid the pathology orientation, we need to expand the initial interview to include a thorough assessment of clients’ skills, talents and resources. We need to know what challenges they have surmounted, what kinds of accomplishments they have attained, what special abilities they have developed. When therapists and clients shift their focus from the pathologized victim to the heroic victor, therapy becomes a much more creative and productive process.

Emphasizing Therapeutic Techniques Over Relationship Building

One of the best things about attending continuing education seminars is learning about the latest therapeutic interventions. And every year or so, such new “breakthroughs” arrive—EMDR, DBT, ACT—you name it. We rush home from the seminars, and can hardly wait for the first patient that we can try out our newfound knowledge on. Many of these innovations do have credibility, but there is one glitch in all of the focus on techniques. Decades of research have consistently demonstrated that the most powerful predictor of positive therapeutic outcome depends less on what type of therapeutic interventions you employ, and more on what kind of therapist-client bond you develop.

An intern related to her ever-patient supervisor that she had been learning about the use of “paradoxical intentions” in her advanced counseling class. She was hoping to try out this new dramatic technique with one of her clients, and did so with a patient during their very first session. The patient had returned to school after a recent divorce, and complained of being totally overwhelmed. She couldn’t get herself to do any homework and was no longer the organized housewife she used to be--failing to do even the simplest of chores like laundry or dishes. The intervention the intern tried was to “ join the symptom” and prescribe the homework assignment to do “absolutely no work at all this week,” then report back at the next session about how this went.

Unfortunately, there was no next session--the client was never heard from again. The lesson here is one that is all too commonly missed: the therapeutic relationship trumps technique. To be more precise, no other single factor affects therapy outcomes more than the quality of the client-therapist relationship. Although exact percentages of therapeutic effect are difficult to ascertain, one study did attempt to do just that. After reviewing over a hundred outcome studies, Lambert and Barley¹ derived an estimate of the relative contribution of the myriad factors which have been studied in outcome research. Surprisingly, the specific techniques employed by therapists (cognitive, psychodynamic, etc.), accounted for only 30 percent of therapeutic outcome. However, the quality of the client-therapist relationship predicted results 40 percent of the time.

In the case discussed above, the paradoxical intervention might have proven effective in the long run, if the therapist and client had developed enough rapport and a trusting relationship before implementing the approach. The tendency to rush into the therapist tool kit and resolve the problem quickly is of course exacerbated by the current emphasis on brief or time-limited therapy. Suffice it to say, this bottom-line, time-is-money orientation is not always in the patient’s best interests. Relationship building begins with the first hello and handshake. In fact, in one study of medical doctors, the handshake was cited by patients on an exit questionnaire as the most positive factor in the office visit.

One of the best (and least utilized) methods to ensure that the therapist and client are on the same page is to employ a relationship assessment tool such as the Working Alliance Inventory developed by Horvath and Greenberg. This user-friendly tool predicts with a high degree of accuracy whether or not a client is at risk of dropping out of therapy. It also points to the areas of disconnect which can be addressed sympathetically with the client.

The Homework Assignment Trap

Providing clients with opportunities to apply what they have learned in therapy is one of the keys to therapeutic effectiveness. This makes good sense, given that clients spend only an hour or two per week in therapy and 165+ hours in the real world. So it would stand to reason that the majority of therapists would regularly utilize out-of-session activities as part of their therapeutic arsenal. However, the sad truth is that the majority of therapists report never using such assignments. Why would there be this disconnection between what the research shows and what most therapists do?

What the research doesn’t show is that creating homework assignments that clients actually comply with is a tricky business--and there are a multitude of therapeutic errors that can interfere with the process.

A case history will help illustrate:

Dr. Doom was working with Sabrina, whom he diagnosed as socially phobic. Sabrina had particular difficulty in her college classes, worrying excessively about bringing attention to herself. To avoid the possibility of embarrassment, she always arrived early to class, sat in the last row, and never raised her hand. After several weeks of therapy in which he gave her no assignments, Dr. Doom decided it was time for action and suggested that Sabrina arrive five minutes late to her next class meeting. At her next session, Sabrina at first told her therapist that she forgot to do the assignment but later admitted that she was able to comply with the first part of the assignment--being late--but could not muster the courage to actually enter the classroom, so she ended up cutting class.

Was Sabrina’s case just another example of client resistance, lack of commitment, or lack of readiness to change? In fact, a careful analysis of the approach the therapist used reveals several therapeutic errors that greatly decrease the likelihood of compliance.

Unilateral Assignments (“Here’s what you need to do…”)
For starters, Dr. Doom “decided” on his own, without input from his client, that it was time for action, and then he chose what that action should be. This one-sided approach helped guarantee noncompliance. Just as the entire therapeutic process should be collaborative, each assignment needs to be arrived at by a joint meeting of the minds. Thus, the term “assignment” is not really appropriate at all because it connotes one person doing the assigning and the other person complying. Far better are concepts such as “experiments,” “activities,” or “tasks.” Therapists certainly can take the lead in developing possible strategies, but clients must be encouraged to provide their input and feedback as the tasks are developed. Clients who feel they have participated in the process of generating the activity are more likely to attempt it, complete it, and maintain whatever they have learned from it. Leaving the client out of the decision-making process increases the likelihood that the task may be beyond the reach of the client’s capabilities. In this case, suggesting the client arrive late to class was an attempt to hit a home run with one pitch instead of moving gradually toward the ultimate goal.

Failing to Prepare Clients for the Assignment
All too often, clinicians employ a “take two aspirin and stay out of drafts” approach to therapy. That is, they act as if mental health work is identical to the medical model in which clients ask the all-knowing physician for a diagnosis, prognosis, and treatment recommendations. In reality, most therapy clients need information about the efficacy of specific interventions. In the course of Dr. Doom’s assignment-giving, he neither sought Sabrina’s input nor gave her even a clue what this fear-inducing activity was supposed to accomplish. What might have seemed obvious to the therapist was probably not at all clear to the client. For those with phobias such as Sabrina’s, education about the efficacy of gradual exposure should have preceded any specific homework recommendations.

Failing to Provide Backup Support to Increase Compliance
As any therapist quickly learns, just because clients say they will perform an activity outside of session, this does not mean they will actually follow through with the commitment. Getting clients to comply with homework (even those assignments they have helped design) is about as difficult as getting students to complete school assignments on time. Understanding this, successful therapists utilize a wide array of approaches designed to overcome the numerous obstacles to completing out-of-session activities.

1. Use Post-it notes. At the conclusion of a session, suggest that the client write down the assignment and then post it at home in a convenient location. The therapist should also make a note of the assignment so it can be reviewed at the next session.

2. Encourage the client to tell a trusted individual about the task, asking the friend to check back and see how the assignment is going. This person should not be a guilt inducer or have any vested interest in the activity other than the welfare of the client. Typically spouses, children, and parents are not useful choices.

3. Determine whether the client has a buddy who is also willing to engage in the desired activity. This can be especially helpful with assignments such as increased exercise or attending classes or support groups.

4. Frame the assignments as a way to learn about oneself while trying new things. Emphasize the possibility of enjoying the opportunity to develop new skills that could be beneficial for a lifetime.

5. Leave little or nothing to chance by carefully clarifying the how, when, and where components of the assignment.

6. Do a thorough assessment of any an all obstacles which might prevent the client from following through with the assignment. Make no assumptions. For example, one client committed to doing an online search for employment during the week. However, an inspection of barriers revealed that the client had never used the internet and in fact did not even have an internet connection for his computer!

Underutilizing Clinical Assessment Instruments

Assessment tools, used early in therapy to measure the type and intensity of the initial problem and occasionally during the course of treatment, can aid in treatment effectiveness, client morale and reduction of termination by resistant clients.

Despite this, clinicians by and large are often skeptical about the value of utilizing assessment tools. For example, one clinical supervisor described a case where a postdoctoral intern was not following agency policy to administer a well-known and highly validated instrument. The trainee stated that she did not “believe in” the assessment because it was not particularly useful and took a lot of time to score--despite the fact that the specific instrument had proven its validity and utility in dozens of studies.

There are a number of factors that contribute to the effectiveness of utilizing assessment instruments:

1. The therapist gains information from a source that allows comparisons to other clients regarding the severity of the problem.

2. Repeating the test at periodic intervals can help demonstrate to the therapist and client whether treatment is being effective.

3. If the results indicate improvement, positive expectations are reinforced. If there is no improvement, the client and therapist can adjust the treatment approach appropriately.

4. Clients tend to see assessment utilization by the therapists as an act of caring, and it enhances client regard for a clinician’s expertise.

All of this and more--and yet clinicians often ignore assessment tools like the plague. Two common reasons for the underutilization of these instruments involve the perception that they require a lot of time to take and score, and that they cost an arm and a leg. To counter this problem we have compiled a list of short, easy-to-score tests which are in the public domain--meaning they are free for the taking. (These are listed at the end of this article.)

While utilizing assessment tools is a good starting point for improving therapeutic outcome, there are two other factors which can enhance their use. First it is crucial to explain to clients that just like medical doctors, therapists utilize assessments in order to pinpoint possible problem areas. Lastly, results of assessments should not be kept secret from the client. It would seem quite odd if your medical doctor did not provide any feedback after a patient had a series of tests such as blood work or X-rays. Similarly, several studies have shown that an open discussion of the results of psychological tests enhances therapeutic outcome by increasing client engagement in the therapeutic process.

A Final Note

All clinicians have no doubt experienced something like the following scenario: You provide your client with some helpful information--“for all the reasons we have discussed, maybe now is not the time to start a new romantic relationship”; your client nods his head in agreement; and at the following session the client announces that he has fallen head over heels in love. The helpful information somehow went in one ear and out the other. Our hope in writing this article and the book upon which it is based is that it will actually impact clinician behavior, that readers will not just nod their heads in agreement, but also put one or two concepts into practice.

To help clinicians move beyond the conceptual to the behavioral involves some self-assessment. This assessment involves taking a few minutes to answer the following questions: What is your clinical batting average?—or conversely, what percentage of your clients are dropping out prematurely? What type of clients are the dropouts? What is it about those clients that makes them more difficult to work with? What type of clients do you tend to do well with?

Addressing questions such as these enables us to take stock of our clinical strengths and weakness and can help us locate the therapeutic errors we may be making with clients – errors such as the ones discussed in this article. This in turn can lead to the implementation of new therapeutic practices and better outcomes for clients and ourselves.

Public Domain Assessment Tools

Following is a list of just a few of the many public domain assessment tools available:
Assertiveness: The Assertiveness Inventory
Depression: Center for Epidemiologic Studies. Depression Scale (CES_D)
Eating Disorders (Anorexia Nervosa): Eating Attitudes Test (EAT)
Social Anxiety: Fear of Negative Evaluation (FNE)
Post-Traumatic Stress Disorder: Impact of Event Scale - Revised (IES - R)
Substance Abuse (Alcohol): Michigan Alcoholism Screening Test (MAST)

While utilizing assessment tools is a good starting point for improving therapeutic outcome, there are two other factors which can enhance their use. First it is crucial to explain to clients that just like medical doctors, therapists utilize assessments in order to pinpoint possible problem areas. Lastly, results of assessments should not be kept secret from the client. It would seem quite odd if your medical doctor did not provide any feedback after a patient had a series of tests such as blood work or X-rays. Similarly, several studies have shown that an open discussion of the results of psychological tests enhances therapeutic outcome by increasing client engagement in the therapeutic process.

¹Lambert, M., J. & Barley, D., E. (2001). Research Summary on the therapeutic relationship and psychotherapy outcome. Psychotherapy, 38, 4, 357-361.

This article was adapted from the book How to Fail as a Therapist: 50+ Ways to Lose or Damage Your Patients, © 2010, Bernard Schwartz and John. V. Flowers. Reprinted with permission.