Identifying patients at risk of complete nonresponse in the outpatient treatment of depression.

Background: Although complete nonresponse in depression treatment is considered to be a major problem in clinical practice, research in this area is very limited. The objective of this preliminary study was to determine the frequency and predictors of complete nonresponse in different treatments for depression.

Methods: Post hoc analysis of the pooled data of 3 consecutive randomized controlled trials of outpatient depression treatment was conducted. The subjects were 313 patients with major depressive disorder and 17-item Hamilton Rating Scale for Depression (HAM-D-17) scores between 14 and 25 who were treated for 6 months with either pharmacotherapy, short-term psychodynamic supportive psychotherapy or combined therapy. Complete nonresponse was defined as a <25% response according to the HAM-D-17. Sociodemographic factors, depression features and adherence were investigated as predictors in a multivariate stepwise logistic regression analysis.

Results: Overall, nonresponse occurred in 34% of the patients. In pharmacotherapy this was 46%, in psychotherapy 39% and in combined therapy 28%. The severity of somatic symptoms was associated with nonresponse in both combined therapy and psychotherapy. No predictive factors were found in the case of pharmacotherapy. In psychotherapy, nonresponse was related to age above 40 years, chronic depression and nonadherence by the patient. In the case of combined therapy, younger age, previous use of an antidepressant and having a previous depressive episode were associated with nonresponse.

Conclusion: Easily measurable patient characteristics may help to identify patients at risk of complete nonresponse to treatment. It is suggested that predictors may differ across treatment modalities. However, head-to-head comparisons are required before it can be recommended to take this into account when selecting the most appropriate treatment for individual depressed patients.