A latent class analysis of mental disorders, substance use, and aggressive antisocial behavior among Swedish forensic psychiatric patients

Background: Patients in the forensic mental health services (FMHS) with a mental disorder, a co-occurring substance use disorder (SUD), and high risk of aggressive antisocial behavior (AAB) are sometimes referred to as the ‘triply troubled’. They suffer poor treatment outcomes, high rates of criminal recidivism, and increased risk of drug related mortality. To improve treatment for this heterogeneous patient group, more insight is needed concerning their co-occurring mental disorders, types of substances used, and the consequent risk of AAB. Methods: A three-step latent class analysis (LCA) was used to identify clinically relevant subgroups in a sample of patients (n = 98) from a high-security FMHS clinic in Sweden based on patterns in their history of mental disorders, SUD, types of substances used, and AAB. Results: A four-class model best fit our data: class 1 (42%) had a high probability of SUD, psychosis, and having used all substances; class 2 (26%) had a high probability of psychosis and cannabis use; class 3 (22%) had a high probability of autism and no substance use; and class 4 (10%) had a high probability of personality disorders and having used all substances. Both polysubstance classes (1 and 4) had a significantly more extensive history of AAB compared to classes 2 and 3. Class 3 and class 4 had extensive histories of self-directed aggression. Conclusions: The present study helps disentangle the heterogeneity of the ‘triply troubled’ patient group in FMHS. The results provide an illustration of a more person-oriented perspective on patient comorbidity and types of substances used which could benefit clinical assessment, treatment planning, and risk-management among patients in forensic psychiatric care.