Psychiatry residency programs historically include training in psychotherapy in addition to training in various forms of "biological" psychiatry. The Accreditation Council for Graduate Medical Education (ACGME) currently requires that psychiatry residency programs demonstrate "that residents have achieved competency" in several forms of psychotherapy (1). This has provided the focus of the recent literature on psychotherapy training, including discussions about how such competency should be assessed, whether training programs are ready to assess competency, and whether it is even possible to assess competency in psychotherapy (2, 3). Publications describing competency assessment include descriptions of written tests (4), rating scales, checklists, and videotapes (5—7). Other studies evaluated residents’ perceptions of their training (8, 9, 10). None, however, has studied the ultimate goal of psychotherapy training—its inclusion in psychiatrists’ actual practice. We address this question by describing how training in one particular evidence-based psychotherapy, dialectical behavior therapy (DBT) (11, 12, 13), has influenced the practice of psychiatry residents after graduation from training.
This study tested the assumption implied in the mandatory psychotherapy competencies that psychiatry residents, once taught psychotherapeutic skills, actually use these skills after graduation. Our hypotheses were as follows: 1) residents who participated in the University of Washington (UW) DBT resident training program would report that they had performed formal or informal DBT since graduation; 2) there would be a positive relationship between the amount of time spent learning DBT and the respondents’ postgraduation practice of DBT; and 3) respondents with more training would report greater confidence and less burnout treating chronically suicidal patients or patients with borderline personality disorder.
DBT Training at University of Washington (UW)
At the UW Department of Psychiatry and Behavioral Sciences, residents learn psychotherapy throughout their training through required and elective didactic lectures, seminars, and individual and group supervision. Except for a single 2-hour R-1 didactics lecture and a few outpatient clinical rotations that make use of DBT to varying degrees, the bulk of the DBT training opportunities at the UW occur as part of the UW DBT Residency Training Program. These training opportunities and the years in which they occur are outlined in F1. This program builds upon itself, beginning with an intensive workshop, a 4-month evening seminar, and the option for continued clinical work in DBT with expert supervision. Those who complete the workshop and seminar are eligible to participate in the clinical consultation team meeting, see individual therapy patients, and/or lead a DBT skills group.
The UW Psychiatry Residency Program provided graduation records of psychiatry residents who finished training between 1998 and 2003, and thus had been working in fellowship or psychiatric practice for 1—5 years. The residency program sent the survey to the 36 graduates who had completed the DBT workshop with a cover letter making the appeal that participation would "… help future residents who receive DBT instruction as we strive to ensure that the curriculum remains relevant." Responses were kept confidential but not anonymous. The UW Human Subjects Review Committee approved the study.
The authors (one who organized [K.A.C.] and one [J.T.F.] who participated as a resident in the DBT training) developed the survey. The survey was five pages in length and consisted of 16 questions. Data collected included: settings in which DBT training occurred, number of patients and months of individual therapy conducted in DBT training program, current practice setting, details of postgraduation DBT practice, impediments to postgraduation practice of DBT, and current confidence and "burnout" when treating borderline personality disorder and/or chronically suicidal patients. Question format consisted of forced-choice (yes/no) responses, Likert scales, checklists, and narrative comments. One of the two questions involving Likert scales was reverse scored to limit response set. The survey was not evaluated psychometrically, but one of the authors (K.A.C.) confirmed that responses fit her records of "dose" of training received.
The survey responses were calculated to assess whether the amount of time spent in DBT training during the general residency program (i.e., "dose" of training) predicted the amount of DBT performed postgraduation. To do this, the training was first operationalized in terms of ascending levels of exposure: 1) workshop alone, 2) workshop plus opportunity to practice DBT during a general residency rotation, 3) workshop plus seminar but no practice, or 4) workshop plus seminar plus opportunity to practice individual or group psychotherapy as part of the UW DBT Residency Training Program or during a clinical rotation. Second, the amount of DBT training was assessed by the number of DBT patients seen and the number of months of DBT therapy conducted during residency.
The question of whether DBT training affected graduates’ behavior in their postgraduation practice was examined first from the perspective of whether respondents practiced "formal" versus "informal" DBT. Formal DBT practitioners practiced DBT as part of a structured DBT program that included a therapist consultation group. Respondents who were not part of a structured DBT team were considered informal practitioners.
Second, the amount of DBT practiced postgraduation was operationalized as the number of DBT interventions ever used since graduation. To avoid the possibility that any association could be due to interventions practiced in other therapy models taught during residency, the analysis only included the interventions that are commonly used in DBT but uncommonly used in the psychotherapies taught as part of the UW psychiatry residency during 1998—2003. That is, these respondents did not have an opportunity to learn these interventions in the residency outside of the DBT program. These 11 "uniquely DBT" interventions are as follows: dialectical strategies, contingency management, behavioral analysis, solution analysis, orienting, commitment, negative reinforcement, shaping, extinction, training an incompatible behavior, and mindfulness. (All respondents who did formal DBT were presumed to have done all of these interventions, as they are common to DBT practice.)
To evaluate our second hypothesis that the dose of DBT training predicted use of uniquely DBT interventions after graduation, an analysis of variance (ANOVA) was conducted to evaluate the effect of the 4-level independent variable and a linear regression was conducted for the number of DBT patients seen for individual therapy and number of months treating patients. Respondents’ degree of confidence and their degree of burnout when working with individuals who have borderline personality disorder or who are chronically suicidal were rated on a 0—5 rating scale with 5 indicating "very much" and 0 indicating "not at all." An ANOVA was conducted predicting confidence and burnout from the 4-level dose of DBT training. Responses were coded using Microsoft Excel 2002 and analyzed with SPSS 12.0 software.
A total of 30 graduates returned their surveys (response rate=83%). Responders (N=30) were comparable in gender, ethnicity, amount of DBT didactic training, and amount of participation on DBT consult teams across the residency compared to the full sample (N=36). However, responders were younger and 5% more likely to participate in the DBT elective clinical training.
The survey showed that 23% (N=7) of the total (N=30) respondents were performing formal DBT at the time they completed the survey. Twenty seven percent (N=8) of the respondents indicated that they had completed formal DBT at any point since graduation.
Those who did not do formal DBT were asked to indicate which of the "DBT interventions"—DBT strategies, DBT skills, or behavioral principles—they have used since graduating from general residency. All twenty two respondents who reported that they have never completed formal DBT indicated that they have used at least one of the DBT interventions, and more than three-quarters reported using six or more—supporting our hypothesis (t1).
As can be seen in F2, the amount of workshop, seminar, and practical training significantly predicted the number of uniquely DBT interventions used after graduation [F =6.03, df=3, p<0.01]. Tukey’s post-hoc test found significance only when comparisons were made between the two groups with lower training dose (i.e., those groups that did not participate in the seminar) and the group with the highest training dose (the workshop plus seminar plus practice group). Thus, our hypothesis was supported: those residents who received more training in DBT were more likely use uniquely DBT interventions than those who just participated in the workshop and/or were exposed to DBT during a clinical rotation.
As shown in F3, the number of patients treated with DBT during residency constituted another measure of training dose. Results indicated both the number of patients seen [F =4.09, df=1, p=0.001] and the total months of DBT individual therapy conducted across patients [F = 3.70, df=1, p=0.001] strongly predicted the number of uniquely DBT interventions used after graduation. This also supported our hypothesis that higher exposure to DBT training during residency is associated with greater implementation of DBT skills after graduation.
Amount of training showed a trend in predicting the psychiatrists’ confidence in treating this population [F = 2.65, df=3, p<0.10], and an examination of the plotted means (F4) indicates a notably lower level of confidence reported by those who only participated in the 2-day workshop and practiced DBT during a clinical rotation during general residency. The respondents belonging to the other 3 groups all reported a higher level of confidence. Burnout was not significant [F =0.74, df=3, n.s.] but followed the same pattern. Thus, the hypothesis that respondents who had completed the seminar portion of the training program would be more motivated to treat this population was not supported.
Finally, the survey included a question regarding barriers to performing DBT. Responses were separated between those from respondents who had performed formal DBT and those who had not, as we expected the barriers might be different for these psychiatrists. As can be seen in Table 2, the barriers that discriminated formal from informal practitioners were related to practical access issues as opposed to disinterest.
This survey was designed to gather information regarding the practice of DBT by former psychiatry residents who participated in DBT training during residency. With a response rate of 83%, the survey surpassed the target response rate of 70% or more recommended by Sierles (14). To our knowledge, this represents the first attempt to gather data on the use psychotherapeutic skills by graduates of a residency. The responses suggest that at least some DBT interventions taught to residents have been put into practice. Since all respondents endorsed using at least one DBT intervention (validation), it can be said that all graduates have used at least some of what they were taught, and 26%—83% used the interventions unique to DBT. The data suggest that residents who were exposed to more DBT training reported using more uniquely DBT interventions since graduation. This held true both for formal and informal practitioners.
The examination of barriers to conducting DBT since graduation showed that a lack of access to DBT skills and consultation groups was the main impediments to performing DBT. This is particularly important because the efficacy of including informal DBT interventions in standard practice is not known and should not be assumed to be beneficial. These results also show that residents were well oriented to the DBT model that requires a means of teaching the DBT skills as well as a consultation team to be considered adherent to the model. Respondents who have completed formal DBT were more likely to list fear of legal liability as a barrier to the practice of this therapy.
Limitations of the survey include its small sample size and lack of a control group. Therefore, no information exists as to whether other graduates from the UW residency are also using DBT interventions. The information received was based upon self-report that may be more a measure of psychiatrists’ perceptions of their work than what they are actually doing. Respondents knew they were participating in a survey performed by former peers and supervisors who were not blinded. This may have caused graduates to overestimate their use of DBT interventions, thereby introducing response bias, a potential threat to the study’s validity. However, representativeness analysis suggested that those who participated were not notably more committed to the program. The survey also requested narrative comments about details of the training process (not reported here) that included many critical comments, suggesting willingness to give negative as well as positive responses.
While other reports about the training of psychiatrists in specific treatments exist, it seems that little has been published about the implementation of such treatments after graduation. Further investigations about the impact of various curricula taught during residency are warranted to see if these results are a function of psychotherapy training in general or DBT in particular. Future research should also extend beyond DBT to investigate the range of psychotherapy models used by graduate psychiatrists in diverse settings. Evaluations of postgraduation practice would provide useful information for residency directors, especially in the age of ACGME-required competencies in various psychotherapy modalities.
In conclusion, the data collected by this survey suggest that the DBT interventions taught to UW residents continue to be used by them after graduation. The majority of respondents to this survey have used 16 or more DBT interventions during the DBT Residency Training Program. They have also indicated that they remained moderately motivated to continue to treat patients with borderline personality disorder. This survey was based on self-report, and it was not without limitations. Future research should investigate the use of other psychotherapeutic skills used by residents after graduation.