[Duration of inpatient depression treatment--fair benchmarking between hospitals]

Process and outcome quality of inpatient treatment of depression in Germany was described in a multicenter study of 10 hospitals in North-Rhine Westphalia, Baden-Württemberg and Bavaria. The treatment of more than 2000 depressive patients was assessed by quality indicators and outcome was compared between the hospitals (benchmarking). Results show great variance in length of stay between the hospitals. While in one hospital patients with depressive episodes were discharged after 36.8 days (average), the average length of stay in another hospital was 64.3 days. Furthermore the study revealed that hospitals differ strongly regarding their case-mix. Using stepwise multiple regression analyses potential confounding variables (sociodemographics, history of previous treatment, severity of depression) were identified and their influence on length of stay was calculated. After cross validation the regression analyses model explained 7% of the variance and included 5 predictors. Length of stay is prolonged by patients with a recurrent depressive disorder, by patients with impairment of social functioning and by severity of the depression. Length of stay is reduced if the indication of inpatient treatment was crisis intervention and if there was a previous suicide attempt. It was shown that differences in patient case-mix only account for a small percentage of hospital differences in length of stay. Method, effort and benefit of the regression analyses approach are discussed.