Predicting discharge in forensic psychiatry: the legal and psychosocial factors associated with long and short stays in forensic psychiatric hospitals

In Germany, both the number of patients treated in forensic psychiatric hospitals and the average inpatient treatment period have been increasing for over thirty years. Biographical and clinical factors, e.g., the number of prior offences, type of offence, and psychiatric diagnosis, count among the factors that influence the treatment duration and the likelihood of discharge. The aims of the current study were threefold: (1) to provide an estimate of the German forensic psychiatric patient population with a low likelihood of discharge, (2) to replicate a set of personal variables that predict a relatively high, as opposed to a low, likelihood of discharge from forensic psychiatric hospitals, and (3) to describe a group of other factors that are likely to add to the existing body of knowledge. Based on a sample of 899 patients, we applied a battery of primarily biographical and other personal variables to two subgroups of patients. The first subgroup of patients had been treated in a forensic psychiatric hospital according to section 63 of the German legal code for at least ten years (long-stay patients, n=137), whereas the second subgroup had been released after a maximum treatment period of four years (short-stay patients, n=67). The resulting logistic regression model had a high goodness of fit, with more than 85% of the patients correctly classified into the groups. In accordance with earlier studies, we found a series of personal variables, including age at first admission and type of offence, to be predictive of a short or long-stay. Other findings, such as the high number of immigrants among the short-stay patients and the significance of a patient's work time before admission to a forensic psychiatric hospital, are more clearly represented than has been observed in previous research.