Length of stay in a German psychiatric hospital as a function of patient and organizational characteristics – a multilevel analysis

Hierarchical linear modelling (HLM) was used to compare the influence of individual and organizational variables on patients' length of stay (LOS) in a large psychiatric hospital in Germany. Organizational variables used were characteristics of the different wards, such as capacity, staff workload, and diagnosis mix. About 28% of the variance in the final hierarchical model was explained by the wards (between‐wards variance) and about 13% by individual or ‘level‐1’ variables. Individual predictors with significant influence included age, diagnosis, level of social functioning and its improvement during therapy, medication, and discharge to complementary services or to subsequent outpatient treatment. Significant organizational variables influencing LOS were number of nursing personnel per bed day, average level of social functioning among patients of each ward, and proportion of patients with schizophrenic disorders treated in each ward. The results are compared with other studies from the recent literature on LOS. General advantages of HLM in the comparisons of hospitals or other health care organizations are illustrated. It is concluded that HLM analysis should be employed regularly in quality management and cost‐containment studies of different healthcare providers.