Long-Term Patterns of Depressive Symptomload and their Predictors in Hospitalised Patients Undergoing Psychodynamic Psychotherapy: Findings of the STOP-D Study

Abstract

Depression is one of the most common mental disorders. While the general effectiveness of in- and outpatient psychotherapy is proven, different long-term patterns in treatment of symptoms of depression have been described. The aim of the present study was to show different patterns of benefit in the context of inpatient psychodynamic psychotherapy of depressive disorders and to detect predictors of different types of response that help to identify possible non-responders and adjust treatments accordingly. Data of the naturalistic multicentre intervention study were collected in 15 German psychosomatic hospital units employing a predominantly psychodynamic approach to treatment. The sample includes n=432 patients (women: age 25-45 years) with typical symptoms of depression. The patterns of outcome were identified using a latent state model with a method factor and a latent class analysis; potential course predictors were analysed using regression analysis. Three long-term patterns of outcome were identified: patients with significant treatment benefit, whose symptom decline was stable even in a 6-month catamnesis (Responders: 76.9%), patients without a significant symptom decrease during treatment and in the follow-up survey (Non-responders: 18.8%), as well as patients with a significant symptom decrease but showing an increase of symptoms in the catamnesis (Backsliders: 4.4%). The severity of baseline depressive symptom load was determined as a predictor for the pattern of Backsliders. Non-responders differed from responders in having had psychosomatic pre-treatments more frequently. In the case of backsliders, further studies should, for instance, verify whether relapses can be explained by the patient's symptoms, treatment, or social environment. In the case of non-response due to numerous unsuccessful pre-treatments, the question arises whether psychosomatic treatment offers the right setting for these patients or how therapy settings should be modified. Long-term patterns reported in the literature were partially confirmed. There are indications of an influence of the initial symptom-load severity on the outcome of treatment. It is important to consider how treatment settings can be modified accordingly.