Predicting treatment outcome in difficult-to-treat depressed patients

Introduction: The treatment of depression has a very high rate of non-responding patients. The recurrence of the disorder and a poor response to the treatment configure difficult-to-treat depression. This is a widespread subgroup, which should be detected earlier for a better formulation of the therapeutic project. The efficacy of a day hospital treatment in containing moderate to severe disorders seems to be comparable to full hospitalisation. However the need for a strict definition of the programme itself and of the psychopathological features of participants emerged as crucial issues. If the integration of different therapeutic approaches seems to improve outcome, there are several limitations due to the severity of the disorder. The aims of the study were to measure the efficacy of a day treatment programme on mood disorders and analyse several clinical features as predictors of treatment outcome. Methods: The present study was conducted in the outpatient service for depressive disorders in Gemelli Hospital on 102 patients (mean age=45.4±15.5; M/F = 43/59). Most of them were affected by difficult to treat depression, that includes patients with history of at least two MDE, risk of chronicization and who did not achieve remission in previous treatments. They underwent a Day Treatment Programme focused on: (1) a prompt reduction of symptoms, based on pharmacological therapy and emotional support; (2) promoting the development of new coping skills; (3) improving relational capacities; (4) psycho-educational rehabilitation. The programme was carried out over 12 weeks. Multidimensional assessments were made at baseline (t0), after 1 month (t1) and after 3 months (t3), using clinical interviews and psychometric tests as: 21-Hamilton Depression and Anxiety Rating Scales; BPRS for the global psychopathological load; Social Adaptation Self-evaluation Scale for social and professional functioning; Connor-Davidson Resilience Scale for resilience. Treatment outcome was measured considering response to the therapy and remission. Involvement of relatives or partners in the treatment programme, social functioning, comorbidities and resilience were taken as predicting clinical features. Results: Patients recruited were moderately depressed (HDRS = 20.2±5.5). Response and remission rate were comparable to those obtained through inpatient care (60% had HAMD<10 in t3). Baseline resilience and social functioning were measured in patients achieving remission and compared to non-responders (CD-RISC 53.3±15.2 vs. 41.2±19.6, p < 0.05; SASS 36.8±7.6 vs. 30.9±9.5, p = n.s.). Taking into account moderate-to-severe depressions only (HAM-D>18), higher baseline SASS above the average (34) is correlated to a higher remission rate (56.7% vs. 36.7%). Conclusions: Patients who had a better response to the treatment, achieving remission by three months from initial assessment, had the following favourable prognostic elements: (1) single diagnosis of Major Depressive Disorder, (2) better relational support, (3) higher scores at SASS and CD-RISC rating scales. Patients who were still affected by a significant psychopathological load by the end of the treatment had at least two of the following unfavourable prognostic elements: (1) diagnosis of Bipolar Depression, (2) comorbidity with Personality Disorders, (3) physical diseases that limited the use of pharmacological therapy, (4) chronic disorder (at least 2 years since the mood disorder started). (5) social isolation, (6) lower scores at SASS and CD-RISC.