Get access

PATIENT TREATMENT PREFERENCE AS A PREDICTOR OF RESPONSE AND ATTRITION IN TREATMENT FOR CHRONIC DEPRESSION

Authors


  • Contract grant sponsor: NIMH; Contract grant numbers: UO1 MH62475 (James H. Kocsis), UO1 MH61587 (Michael E. Thase), UO1 MH62546 (Daniel N. Klein), UO1 MH61562 (Madhukar Trivedi), UO1 MH63481 (Philip Ninan and Barbara O. Rothbaum), U01 MH62465 (Alan J. Gelenberg), UO1 MH61590 (Martin B. Keller), UO1 MH61504 and T32 MH019938 (Alan F. Schatzberg), U01 MH62491 (James P. McCullough, Jr.)

Correspondence to: Dana Steidtmann, Department of Psychiatry and Behavioral Sciences, Stanford University, 401 Quarry Road, Stanford, CA 94305–5722. E-mail: dsteidt@stanford.edu

Abstract

Background

Findings regarding the relationship between patient treatment preference and treatment outcome are mixed. This is a secondary data analysis investigating the relationship between treatment preference, and symptom outcome and attrition in a large two-phase depression treatment trial.

Methods

Patients met DSM-IV criteria for chronic forms of depression. Phase I was a 12-week, nonrandomized, open-label trial in which all participants (n = 785) received antidepressant medication(s) (ADM). Phase I nonremitters were randomized to Phase II, in which they received 12 weeks of either cognitive-behavioral system of psychotherapy (CBASP) + ADM (n = 193), brief supportive psychotherapy (BSP) + ADM (n = 187), or ADM only (n = 93). Participants indicated their treatment preference (medication only, combined treatment or no preference) at study entry. Symptoms were measured at 2-week intervals with the 24-item Hamilton Rating Scale for Depression (HAM-D).

Results

A large majority of patients reported a preference for combined treatment. Patients who preferred medication only were more likely to endorse a chemical imbalance explanation for depression, whereas those desiring combined treatment were more likely to attribute their depression to stressful experiences. In Phase I, patients who expressed no treatment preference showed greater rates of HAM-D symptom reduction than those with any preference, and patients with a preference for medication showed higher attrition than those preferring combined treatment. In Phase II, baseline treatment preference was not associated with symptom reduction or attrition.

Conclusions

Treatment preferences may moderate treatment response and attrition in unexpected ways. Research identifying factors associated with differing preferences may enable improved treatment retention and response.

Ancillary