Patient-Important Outcomes in the Long-Term Treatment of Bipolar Disorder: A Mixed-Methods Approach Investigating Relative Preferences and a Proposed Taxonomy.

Abstract

In patient-centered healthcare, the assessment and selection of treatment should be based on outcomes important to patients and the relative importance patients place on these outcomes. The evidence base on long-term treatment outcomes important to patients with bipolar disorder is inconclusive. The aim of this study was to investigate the relative importance of patient-important outcomes in bipolar disorder, and to construct a holistic and logically sound shortlist of treatment outcomes relevant in the evaluation and selection of pharmacological treatment in bipolar disorder. Overall, 22 outpatients from southern and eastern Norway participated in four focus groups, and suggested outcomes important in treatment decisions. Quantitative, relative importance weights for treatment outcomes identified in literature reviews were elicited from each participant, employing a self-explicated approach (SEA). The method combined a ranking- and rating-stated preference exercise and resulted in a 0-100 SEA-score for each outcome. Outcomes from the literature accommodated the outcomes suggested in the focus groups. Mean age in the sample was 42 years and 64% were women. All patients completed the exercises with consistent results. The most important outcomes were severe depression (median SEA 95 [interquartile range 26]), severe mania (76 [40]), quality of life (65 [53]), work/school functioning (58 [48]), and social functioning (54 [50]). Avoiding severe mania was significantly more important to patients with bipolar disorder type I compared with patients with type II. Outcome scores correlated strongly (p < 0.01) across the ranking and rating exercises. Based on the results, a simplified and consistent set of outcomes was constructed. Patients' preferences for outcomes in the long-term treatment of bipolar disorder vary considerably. To advance patient-centered healthcare, we propose that researchers, clinical guideline producers, and patient-clinician dyads integrate a taxonomy of patient-important outcomes, such as constructed in this study, when assessing treatment options.