Journal Article

DZNE-2026-00459

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Non-European migrants with schizophrenia spectrum disorders in Swiss forensic and general psychiatric care facilities - A comparative study using machine learning.

Hofmann, A. B. ; Dörner, M.DZNE* ; Stottmeister, F. ; Machetanz, L. ; Kirchebner, J.

2026
Elsevier Science Amsterdam [u.a.]

This record in other databases:    

Please use a persistent id in citations: doi:10.1016/j.forsciint.2026.112976

Abstract: Personal history of migration poses an important risk factor for schizophrenia spectrum disorders (SSD), which are also associated with a higher rate of criminal behavior. To enhance care for migrants, a vulnerable and often stigmatized group in both general and forensic psychiatry, this study investigates clinical, therapeutic, and psychopathological differences between non-European migrants diagnosed with SSD treated in forensic and general psychiatric settings. The aim is to identify factors that may influence pathways into the criminal justice system and pose challenges for the therapeutic process, rather than directly predicting criminal behavior. We compared retrospectively obtained data of 52 general (GPP) and 104 forensic psychiatric (FPP) inpatients - all with a history of migration from non-European countries and treated in Zurich, Switzerland. To detect complex variable patterns, supervised machine learning models were applied to a training dataset. The best algorithm was then used to assess the predictive power of nine out of 174 possible predictor variables in a validation dataset. Two positive and negative syndrome scale (PANSS) items (uncooperativeness, poor impulse control) and the modified sum score upon discharge (mean 14.5 for GPP vs. 10.8 for FPP), previous treatments (more frequent treatment for GPP: inpatient 96% vs. 69%; outpatient 81% vs. 37%), medication strategies regarding antidepressants and antipsychotics, attendance to occupational therapy, and failed attempts of expanding patients' freedom influenced the model. The final model yielded satisfactory statistical properties (area under the curve (AUC) of 0.86 and balanced accuracy of 71.9%, including a specificity of 88.9%), demonstrating its strong predictive performance. General and forensic psychiatric inpatients with a non-European background diagnosed with SSD differ in various clinical variables. These findings may inform future psychiatric practices by identifying clinical and therapeutic variables that could support targeted interventions and potentially reduce the risk of criminal justice involvement among migrant patients with SSD.

Keyword(s): Criminal behavior ; Forensic psychiatry ; Machine learning ; Migration ; Offending ; Schizophrenia spectrum disorders

---

Contributing Institute(s):

1. Mixed Cerebral Pathologies and Cognitive Aging (AG Schreiber)

Research Program(s):

1. 353 - Clinical and Health Care Research (POF4-353) (POF4-353)

Appears in the scientific report 2026

---

Database coverage:
; ; ; BIOSIS Previews ; Biological Abstracts ; Clarivate Analytics Master Journal List ; Current Contents - Clinical Medicine ; Ebsco Academic Search ; Essential Science Indicators ; IF < 5 ; JCR ; Nationallizenz ; SCOPUS ; Science Citation Index Expanded ; Web of Science Core Collection ; Zoological Record

---