Guideline concordance and outcome in long-term naturalistic treatment of bipolar disorder - a one-year longitudinal study using latent change models

Background: Only few studies investigated the relation between concordance with treatment guidelines and treatment outcome in everyday treatment of bipolar disorder (BD). Prospective studies are scarce. Methods: A nationwide, naturalistic, prospective study on the relation between guideline concordance and treatment outcome in the long-term outpatient treatment of patients with BD. Participants completed a survey on treatments received and various outcome measures at baseline and after one year. Results: Of 839 patients who completed the baseline survey, 615 (73.3%) also completed the follow-up survey. Consistent with our a priori hypothesis, cross-sectional analyses at baseline showed correlations between guideline concordance with quality of life (r =. 17, p < .001), treatment satisfaction (r =. 17, p <.001), and impaired functioning (r = -.10, p = .04). At follow-up, guideline concordance was correlated with severity of illness (r = -.10, p = .05), quality of life (r =. 18, p < .001), and treatment satisfaction (r =. 15, p < .001). Concerning three additional hypotheses on longitudinal relations between concordance and outcome measures, only a positive relation was found between change in guideline concordance and change in quality of life. Limitations: Selection bias may have occurred by inclusion of patients with neither a very severe nor a very mild course of illness. Conclusions: Although guideline concordance was high throughout the study, change in guideline concordance was positively associated with change in quality of life, suggesting that especially in long-term treatment, continuous efforts to optimize ongoing treatment is essential.