Dropping Out of Cognitive Behavioral Therapy

Cognitive Behavioral Therapy (CBT) has earned a position of high regard in the scheme of evidence-based treatment of psychological disorders. Yet, it shares some of the same difficulties faced by psychotherapy and medical practice in general. One such major problem is dropout.

Dropout is the client’s discontinuation of treatment against the recommendations of the clinician. When this happens, many questions arise. Firstly, what is the dropout rate in general, are there particular times when it is especially likely to occur, what are the reasons for dropping out, and what diagnostic and treatment factors could explain dropout?

We undertook a meta-analysis of over a hundred studies incorporating a total of almost 21,000 clients receiving CBT for a range of mental health disorders (Fernandez, Salem, Swift & Ramtahal, in press). This revealed that almost 16% dropped out even before treatment started; in other words, they were screened and offered treatment, but did not proceed to session 1. So, this may be the time to introduce motivational interviewing. Of treatment starters, a further 26% dropped out, implying that taking the first step is no indication that the person will stay on course.

Of the different categories of disorders, depression turned out to be associated with the highest dropout rate (regardless of treatment phase). This may be tied to some of the very symptoms of depression (e.g., general loss of interest, hopelessness, helplessness, and slowing down of motor behavior) which collectively decrease the client’s likelihood of initiating or continuing treatment.

Format of treatment delivery also made a significant difference. Dropout was about 10 to 15 percentage points higher in computer-based therapy as compared to in-person therapy. On average, individual therapy had the lowest dropout rate. Might this actually mirror similar patterns in the education system where the initial enthusiasm over mass online learning has waned in favor a return to the classroom, and individual instruction still remains the most effective medium for learning and student retention?

The setting of treatment also made a significant difference. As expected, inpatient settings were associated with lower dropout rates than outpatient settings. The former is inherently more effective for continuity whereas the latter does allow greater latitude for discontinuity.

No significant differences in dropout occurred as a function of client development stage (adolescent versus adult). Neither did therapist type (licensed versus trainee) make any significant difference.

One noteworthy nonsignificant result relates to methodological rigor. The strict Randomized Controlled Trial (RCT) did not produce any appreciable reduction in dropout rates. Most of the research in this field was published over the last 3 decades and there is a notion that earlier studies were less well-designed. Whether this is tenable or not, it turned out that recency of publication did not make a difference to dropout rates.

Retention is the flip side of attrition. If the latter is correlated with diagnosis, treatment delivery, and treatment setting, then CBT retention could be boosted by giving special attention to these same factors. To take an idiographic example, a client who is depressed may require special efforts to remedy motivational deficits as soon as treatment is offered and before it actually commences. This is because a significant minority of patients don’t even show up for the first session. Individualizing therapy in residential settings may further improve retention. Methodological rigor as implicit in RCT’s and recency of publication is hardly a promise of better participant retention.

We started out by defining dropout as willful discontinuation of therapy against professional advice. There must be a complexity of reasons that go beyond the scope of our study. These encompass process variables (e.g., therapeutic alliance, psychological mindedness, and treatment aversiveness) to mere circumstances (e.g., relocation, scheduling difficulties, and financial limitations). All these make up a further dimension that awaits future investigation.

Author Bio

Ephrem Fernandez is Professor of Clinical Psychology at the University of Texas at San Antonio. He has pioneered CBT for chronic pain and cognitive behavioral affective therapy (CBAT) for emotional disorders such as dysfunctional anger. Within the field of psychotherapy, his efforts are directed at innovation and integration of therapeutic techniques, and the empirical evaluation of their effectiveness.

Discussion Questions

  • Is dropout a reflection on the client or the therapist, or both?
  • Some degree of dropout is almost inevitable in psychotherapy. What are the reasons for such attrition of clients/participants in CBT?
  • What steps can be taken to minimize dropout at different phases, from pre-treatment to follow-up?
  • Why does dropout tend to be higher in E-therapy as compared to in-person therapy?
  • How does dropout in the mental health field compare with that in mainstream medical care?
  • What research standards may help bring about consistency in operational definitions of dropout, routine reporting of dropout data, and the way in which such data are handled in statistical analyses of treatment outcome?



Fernandez, E., Salem, D., Swift, J. K., & Ramtahal, N. (2015). Meta-analysis of dropout from cognitive behavioral therapy: Magnitude, timing, and moderators. Journal of Consulting and Clinical Psychology, 83(6), 1108-1122. doi:http://dx.doi.org/10.1037/ccp0000044

Self, R., Oates, P., Pinnock-Hamilton, T., Leach, C. (2005). The relationship between social deprivation and unilateral termination (attrition) from psychotherapy at various stages of the health care pathway. Psychology and Psychotherapy: Theory, Research and Practice, 78, 95-111.

Swift, J. K., & Greenberg, R. P. (2012). Premature discontinuation in adult psychotherapy: A meta-analysis. Journal of Consulting and Clinical Psychology, 80(4), 547-559.