[Predicting long-term outcome of in-patient psychosomatic treatment].

Abstract. The outcome of in-patient psychosomatic treatment was evaluated by longitudinal assessment at hospital admission, discharge, and 1-year-follow-up. Out of n = 1829 patients, n = 712 responded (38.9 %). Response rates co-varied with duration of in-patient treatment. Responder-analyses characterise the sample as fairly representative. Outcome criteria included standardised measures of complaints, mood quality, global quality of life, life satisfaction, everyday functioning, and social integration. Disease-related parameters (duration of illness, medical certification, number of doctors consulted, number of somatic resp. psychosocial diagnoses) and generalised outcome expectancies (self-efficacy, optimism, pessimism) were included as predictors. Patients' retrospective estimations at follow-up as well as longitudinal assessments show a successful and mostly stable therapy outcome. Negative moods and complaints show a new increase at follow-up, though, still remaining below the level at hospital admission. Although there is much accordance among the diagnostic groups, patients with anxiety disorders (ICD-10: F40/41) and patients with adaptation disorders (F43) show distinctly varying courses. Regression analyses yield specific correlations between disease-related parameters and long-term outcome: Duration of treatment does not predict treatment outcome. The number of somatic diagnoses is a predictor of complaints, the number of psychosocial diagnoses a predictor of negative moods at follow-up. Pessimism and - to a lesser degree - self-efficacy prove to significantly predict numerous outcome criteria. Generalised expectancies should be considered in the process of prognosis and differential indication.