Abstract
Dynamic deconstructive psychotherapy (DDP) is a time-limited manual-based individual treatment for borderline personality disorder (BPD) that incorporates major psychodynamic principles including affect focus, free association, transference, and countertransference (Gregory and Remen 2008). A twelve-month randomized controlled trial (Gregory et al. 2008) and an eighteen-month naturalistic follow-up study (Gregory, DeLucia-Deranja, and Mogle 2010) demonstrated the efficacy of DDP for co-occurring BPD and alcohol use disorders. Five of the six therapists participating in the study were third-year psychiatry residents. Given the success of DDP using psychiatry resident trainees as study therapists, the present analysis attempts to determine whether DDP could be a useful model for meeting residency training program requirements in psychodynamic therapy. To accomplish this task, this study addresses (1) trainee outcomes and treatment adherence as compared with the performance of expert DDP practitioners, and (2) the experience of those residents regarding not only their conduct of this treatment, but also how it influenced their professional development in general.
Method
Treatment model
DDP attempts to remediate three neurocognitive deficits putatively responsible for adaptive processing of emotional experiences by using the following four interventions: (1) Association—fostering the verbalization of recent interpersonal experiences into simple narratives, and labeling and elaborating any associated emotions; (2) Attribution—exploring and integrating alternative or opposing attributions toward self and other; (3) Alterity-ideal—building therapeutic alliance in the early stages of therapy through empathy, mirroring, framing, and summarizing; and (4) Alterity-real—repairing ruptures in the therapeutic alliance by providing novel experiences that deconstruct enactments and promote self-other differentiation. Learning about DDP theory, the treatment contract, stages of recovery (Gregory 2004), attribution states (Gregory 2007), and the nondirective, nonjudgmental therapeutic stance (Gregory 2005) aids the therapist, especially the trainee, in avoiding enactments, providing therapeutic deconstructive experiences, and maintaining adherence to the treatment model (for further description of these terms, see the online treatment manual [Gregory 2011]).
Study
Ten subjects (ages 18–45) dually diagnosed with BPD and alcohol abuse or dependence completed a twelve-month RCT of DDP versus optimized community care (OCC) (Gregory et al. 2008) with an additional eighteen months of naturalistic follow-up (Gregory, DeLucia-Deranja, and Mogle 2010), for a total of thirty months in the study. Subjects attended weekly individual sessions with either the expert therapist (RJG) or one of five resident trainees. Research assistants blinded to group allocation used structured interviews and questionnaires to assess outcomes at baseline, at three, six, nine, and twelve months, and after eighteen months of naturalistic follow-up.
Investigators chose the Borderline Evaluation of Severity Over Time (BEST; van Wel et al. 2006; Blum et al. 2002) a priori as the primary outcome measure. The BEST is a 15-item questionnaire assessing core symptoms of BPD, and patients were assessed as having a clinically meaningful change in core BPD symptoms if the BEST score decreased at least 25% from baseline. Additionally, a subset of the Addiction Severity Index (ASI; McLellan et al. 1992) was used to quantify the number of heavy drinking days over the preceding month. Also, two independent observers used video recordings of sessions to rate working alliance using the Working Alliance Inventory (WAI; Tichenor and Hill 1989; Tracey and Kokotovic 1989) and treatment adherence using a 25-item DDP adherence scale at each time interval. The DDP adherence scale assessed a session for both the number of positive interventions fostering [A]ssociation, [A]ttribution, [A]lterity-ideal, and [A]lterity-real, and the number of negative [E]nactments. The overall percentage of adherence was then calculated as (A + A + A + A)/(A + A+ A + A + E). Adherence to DDP technique has been demonstrated to positively correlate with symptom improvement (Goldman and Gregory 2009). At the end of the twelve-month trial, trainees anonymously evaluated the treatment protocol using a standardized protocol implementation questionnaire (PIQ; Najavits 1996).
Resident training
Five third-year psychiatry residents were trained to competency in DDP through a combination of didactics, case supervision, and a training manual. Didactics included five hour-long presentations incorporating DDP theory, treatment contract, stages of recovery (Gregory 2004), attribution states (Gregory 2007), and therapeutic stance (Gregory 2005). Every week each trainee received an hour of individual case supervision with the principal investigator (RJG), in which video recordings of therapy sessions were employed. The amount of case supervision needed to achieve competency was 16.2
Results
Results from the one expert therapist and the five resident trainee therapists in the study show that DDP has an efficacy superior to that of OCC (Gregory, DeLucia-Deranja, and Mogle 2010; Gregory et al. 2008). When results from the trainees were isolated out for the purposes of this study, the degree of improvement in BPD symptoms and heavy drinking was intermediate between the degrees achieved by the DDP expert and the community practitioners (Figure 1).
(Top) The actual percentages of participants achieving clinically meaningful change in core BPD symptoms (25% reduction in BEST scores). (Bottom) Percent heavy drinking days during previous 30 days.
Regarding adherence, the mean number of DDP-related interventions per session was compared between the trainees and the expert with the following results: Association: 4.7 ± 4.1 and 2.0 ± 0.9 interventions for the trainees and the expert respectively; Attribution: 5.7 ± 1.6 and 5.2 ± 0.6; Alterity-ideal: 1.4 ± 1.5 and 1.3 ± 0.8; and Alterity-real: 2.3 ± 2.1 and 1.9 ± 2.0. The mean number of negative enactments was 6.6 ± 3.9 for the trainees and 3.6 ± 1.2 for the expert. Despite a trend for trainees to engage more often in negative enactments than the expert, the overall percentage of adherence did not differ statistically between the two: 69% ± 11% and 76% ± 5.3%, respectively. Working alliance (WAI) was good for both trainees (5.1 ± 0.7) and expert (5.8 ± 0.4), and these values were not statistically different from each other. Thus, resident trainees were able to demonstrate good adherence to the treatment model and a good working alliance with the patient.
An anonymous standardized interview with the resident trainees (Table 1) demonstrated that residents were able to become comfortable with the treatment protocol within about seven weeks. By the end of twelve months of treating patients, they reported being strongly satisfied with the treatment model, felt comfortable and effective implementing it, felt that it took only moderate effort to implement, developed a strong allegiance to it, and felt they would use it in the future. They also left with the impression that the treatment model was a good fit for nearly all the patients and clinical situations they faced (97% ± 7%). Regarding how the treatment study has influenced them professionally (Table 1), the resident therapists felt the treatment model had a strongly positive influence on their development as therapists, as psychiatrists, and on their attitudes toward psychodynamic psychotherapy. Moreover, the study had a strongly positive influence on how they evaluated and treated not only BPD patients outside the study, but non-BPD patients as well.
Anonymous trainee evaluation of treatment protocol using standardized Protocol Implementation Questionnaire (PIQ)
Discussion and Conclusions
The outcome and adherence findings support the hypothesis that the DDP treatment/training model can be used effectively not only to train psychiatry residents to competency in psychodynamically oriented psychotherapy, but also to treat patients with a complete treatment protocol within the time constraints of psychiatry residency training. Further, the study demonstrates that trainees can deliver this care to a severely impaired patient population commonly seen in psychiatry residency training clinics. Thus, DDP training has the potential for being immediately relevant to trainees, and may contribute to their rapid achievement of competency in psychodynamic psychotherapy more generally.
Trainee assessment of the treatment/training model was very favorable. Not only did residents feel they were able to execute it comfortably and effectively; they also felt that they were able to generalize the principles they learned to other treatments and patient populations and that the treatment/training model had a strongly positive influence on their overall professional development. DDP is at its core psychodynamic psychotherapy much like that used with many other patient populations. Thus, DDP learned to treat BPD patients is easily extended to other clinical situations that respond well to dynamically oriented treatments. In our opinion, the training difference lies mainly in the model’s ability to help residents readily highlight, frame, and intervene against many common therapy-interfering factors that therapists, especially trainees, get themselves into when treating patients. We conclude that DDP may be an ideal psychodynamic psychotherapy training model for psychiatry residents.
Footnotes
This poster received the 2012 APsaA Poster Session Award.