[Longitudinal analysis of factors influencing direct costs of schizophrenia treatment].

In a prospective longitudinal study with 5 points of measurement the direct service costs as well as the clinical and social characteristics of 307 outpatients with the diagnosis of schizophrenia (ICD 10 F 20.0) were assessed over a period of 2.5 years. A random-effect regression model was computed for the assessment of the impact of clinical and social variables on cost variance. Within-effect and between-effect of independent variables on the cost variance were identified by comparing of within and between effect models. Results of the study show that 10 % of the within variance and 32 % of the between variance were explained by the random effect model. Significant within effects were found for a deterioration of emotional health status and for general level of functioning. Significant between effects were found for living situation, partnership, number of inpatient episodes, psychopathological symptoms, depression, alcohol abuse and number of life events. The study results suggest that interventions aiming at the reduction of direct costs should not be restricted to medical treatment but must also include psychosocial interventions for the enhancement of general functioning. However, the explained proportions of within and between variance reveal that the possibilities of manipulating direct treatment costs by means of patient-related interventions may be limited.