Determinants of clinical outcome and length of stay in acute care forensic psychiatry units.

Criminological and sociodemographic variables, such as previous criminal convictions, increased risk of violence, early onset of mental disorder, antisocial personality, psychosis and low social support, have all been related to longer length of stay (LoS) and poorer outcome in long stay forensic services. The factors impacting on LoS and clinical response in acute care specialized units are poorly documented. To address this issue, we examined the psychiatric records of all cases admitted between January 1st and December 31th 2020 in the sole acute ward for detained persons located in the central prison of the Geneva County, Switzerland. Information on judicial status included pre-trial versus sentence execution, previous incarcerations, and age of the first incarceration. Sociodemographic data included age, gender, marital status, and education attainment. Previous inpatient stays prior to incarceration were recorded. All of the ICD-10 clinical diagnoses were made by two independent, board-certified psychiatrists blind to the scope of the study. The standardized assessment was based on the HoNOS (Health of Nation Outcome Scales) at admission and discharge, HONOS-secure at admission, HCR-20 (Historical Clinical Risk 20) version 2, PCL-R (Psychopathy Checklist Revised), and SAPROF (Structured Assessment of Protective Factors). Stepwise forward multiple linear regression models predicting the LoS and delta HONOS respectively were built with the above mentioned parameters. The selected variables were then used in univariate and multivariable regression models. Higher HCR-scores (mainly on clinical items), and longer LoS were related to higher delta HONOS scores. In contrast, cases in pre-trial detention showed a worst clinical outcome. In multivariable models, all three variables remained independent predictors of the clinical outcome and explained 30.7% of its variance. Only education and diagnosis of borderline personality were related to the LoS and explained 12.6% of its variance in multivariable models. Our results suggest that the use of acute wards specialized in forensic psychiatry are mainly useful for patients with prior inpatient care experience, and higher violence risk during sentence execution. In contrast, they seem to be less performant for persons in pre-trial detention that could benefit from less restrictive clinical settings.