Identifying subpopulations in forensic addiction care: A latent class analysis

Purpose

Forensic patients with substance use disorder (SUD) vary in clinical.

characteristics like psychiatric comorbidity, including mild intellectual disability (MID). In this study, we examined whether different patient classes could be identified based on type of psychiatric diagnosis (including MID) and historic risk factors at treatment start, using Latent Class Analysis (LCA); whether these classes differed on risk behavior during treatment and treatment outcomes; and whether MID was associated with risk behavior and treatment outcomes.

Method

Data were retrieved from health records in a forensic addiction treatment centre in the Netherlands (n = 252). Information included DSM-5 diagnoses, historical risk factors for recidivism and treatment outcomes (urine toxicology, number of aggression incidents and drop-out).

Results

We identified four patient-classes, including one with a high prevalence of psychopathology and high historic risks, one with severe past substance use and long treatment history and two classes with low historic risks. These classed did not differ in risk behavior or treatment outcomes. MID was associated with risk behavior during treatment, but not with treatment outcomes.

Conclusions

These data suggest that though subgroups of forensic addiction patients are identifiable, historic risks do not predict variations in treatment outcomes, and that co-occurring MID might be clinically more relevant.