A Psychodynamic Perspective on Treatment-Refractory Mood Disorders

Preliminary findings from an ongoing naturalistic, longitudinal study of treatment outcome support the notion that work with patients who have treatment-refractory mood disorders is enhanced by the careful integration of a psychodynamic therapeutic approach into the customary biological approaches. Review of a series of cases led to the identification of nine core psychodynamic principles associated with good outcome in these cases.

Advances in the treatment of patients with mood disorders have led to the recognition that about half these patients fail to respond adequately to biological treatment approaches (Fava and Davidson, 1996). Only a minority of patients recover fully with medications (Rush and Trivedi, 1995). Augmentation strategies and the use of algorithms to maximize therapeutic response in patients with mood disorders have emerged, but a subset of patients remain refractory to treatment. Treatment for these patients often becomes chronic crisis management, organized around fending off the next crisis or recovering from the last, with suicide attempts, repeated hospitalizations and significant strain on treaters.

Thase and colleagues (2001) suggested including a psychosocial component in treating these patients. However, their psychosocial approach does not include such central psychodynamic concepts as the role of unconscious mental processes or transference and countertransference and how they may relate to treatment-refractoriness.

This article offers a set of psychodynamic principles for working with patients with treatment-refractory mood disorders. The principles were derived from a study of patients with treatment-refractory mood disorders treated at the Austen Riggs Center, a national referral center for patients with treatment-refractory disorders. The Riggs continuum of care includes individual psychodynamic therapy four times a week and a sophisticated milieu program within a completely open setting. Treatment averages seven months in duration. Riggs is conducting an ongoing, naturalistic, longitudinal follow-along study of patients during and after treatment. The study uses reliable measures of symptom change, but also reliable measures of such psychodynamic constructs as defenses and conflicts.

Christopher Perry, M.D., M.P.H., and I now have data on the first 57 patients with three to five years of follow-up. Eighty percent of patients in the sample had reliably diagnosed mood disorders. The patients were a disturbed group, with 40% having six or more self-destructive episodes, 50% having made at least one serious suicide attempt, and 60% with three or more previous hospitalizations. Close to half the patients had histories of childhood trauma, while an increased frequency of trauma was associated with a greater number of Axis I disorders. Eighty-six percent of the patients had personality disorders, usually borderline personality disorder.

Preliminary findings suggested significant improvement over three to five years, with most of the change occurring after the first two years. Conservatively estimated effect sizes ranged from 0.4 (for self-report measures), to 0.5 to 1.0 (for employment, defenses and the Hamilton Rating Scales for Anxiety and Depression [HAM-A, HAM-D]) to 1.5 (for Global Assessment of Functioning [GAF]).

Based on the evidence that these patients were improving, evidence from the literature (Bremner, 1999; Kaufman et al., 2000) and clinical experience suggesting that character issues play a role in treatment-refractoriness, I undertook a review of such cases to discern psychodynamic principles associated with good outcome.

The principles were extracted from cases studied by direct review of the medical records, supplemented by my participation in extended case conferences and/or serving as psychotherapy supervisor. In two cases, the review was based on my own reassessment of the medical records of patients I had treated in the past. This methodology is limited by my own unwitting clinical biases, but it is a reasonable starting point for such a complex endeavor. A summary of the psychodynamic principles follows.

Our experience confirms clinical experience that patients with treatment-refractory mood disorders often present with complex comorbidity, making them unlike the patients generally included in medication trials (Perry et al., 2001). Our first 57 patients met criteria for an average of six Axis I and II disorders. Eighty-six percent met criteria for one or more personality disorders, 70% had underlying dysthymia, 41% had substance abuse problems, and 36% met criteria for posttraumatic stress disorder (PTSD).

Patients with complex comorbid clinical presentations require an interdisciplinary treatment plan. Medication, psychotherapy, family work, substance abuse treatment, group work, case management services and medical management may all be required, regardless of the treatment setting. Often a different clinician provides each service, creating a de facto treatment team. When there is such a team, communication is essential. This is particularly important in cases with prominent Axis II pathology because of the frequency of splitting and projective defenses. Team members may find themselves split in ways that reflect and, potentially, illuminate the patient's life history (Shapiro and Carr, 1991). Patients with good outcomes generally had treatment teams that used a psychodynamic formulation to notice and integrate splits, while deepening understanding of the patient.

In psychodynamic treatment, the alliance includes negotiation of an agreement to explore the patient's mind and the meaning of their actions, verbalizations and symptoms. It also includes careful delineation of the roles and responsibilities of therapist and patient. Therapists of patients with treatment-refractory mood disorders who had good outcomes made the alliance the foundation on which the rest of the treatment was built. Particularly when Axis II pathology complicates treatment of a mood disorder, acting-out and assaults on the boundaries of the treatment often unfold. A clearly negotiated alliance, in the face of a strong transference attachment, helps contain acting-out within the therapy instead of spilling over into the patient's life (Plakun, 1994). This can avoid chronic crisis management during therapy sessions.