Psychosomatic inpatient treatment: Real-world effectiveness, response rates and the helping alliance

While inpatient mental health treatments in real-world settings have shown to be generally effective with moderate to large pre-post effects, little is known about rates of response in inpatients.

Inpatients routinely treated at a university hospital for psychosomatics and psychotherapy in Germany were assessed before and after receiving a psychodynamically oriented multimodal treatment. As primary outcome response rates based on the Symptom Checklist-90-Revised (SCL-90-R) were used. As secondary outcomes, pre-post effect sizes on measures of symptoms and stress (SCL-90-R, Hospital anxiety and depression scale, Perceived stress questionnaire) as well as interpersonal problems (IIP) were determined. Additionally, the relationship of the helping alliance to outcome was examined.

A total of 709 patients with various primary diagnoses and high mental (72.4%) and physical (61%) comorbidity, receiving on average 7.9 weeks of treatment, were included. The response rate based on the SCL-90-R global severity index was 62.9% at posttreatment and 60.4% at short-term follow up. Pre-post effect sizes on symptom and stress measures were large (Cohen's d ≥ 1.0) while change regarding interpersonal problems was small (d = 0.34). For patients responding to treatment a significantly better helping alliance was found, corresponding to a large effect (d = 0.84).

A majority of patients benefitted considerably from a psychosomatic inpatient treatment in terms of response and pre-post effects. Response was significantly related to the quality of the therapeutic alliance. Monitoring and improving alliance may enhance treatment outcome.