Hallucinogen use predicts reduced recidivism among substance-involved offenders under community corrections supervision

The investigators analysed intake and supervision data collected 2002–2007 from the Treatment Accountability for Safer Communities (TASC) programme in the Southeastern United States. The sample comprised 25,622 individuals charged with a felony and under community corrections supervision (for example, probation or parole). TASC is a case-management intervention for people with a history of substance involvement; participants lived in the community and continued to have access to substances. The programme emphasised sustained abstinence verified by random urinalysis, but services were idiographic and could include outpatient counselling, residential treatment, vocational rehabilitation and other supports as determined by caseworkers. At intake participants completed a structured interview recording sociodemographic and psychosocial variables and current substance use diagnoses. The primary independent variable was any hallucinogen use disorder (DSM-IV abuse or dependence) versus no hallucinogen use disorder; the authors used the disorder classification because the dataset did not include mere-use indicators and caseworkers likely recorded illicit use as a disorder. Outcome was supervision failure (versus non-failure), a composite measure that could reflect noncompliance with TASC requirements or legal conditions, failure to appear, incarceration, remand to jail/prison, or death. Median time under TASC supervision and the overall failure rate were reported. Initial group comparisons used analyses of variance for continuous measures and chi-square tests for categorical measures to identify variables associated with hallucinogen-use-disorder status. Variables associated with hallucinogen status were entered simultaneously as covariates in a logistic regression predicting supervision failure; these covariates included age, race, employment status, marital status, living situation, primary means of financial support, violent-offence history, substance-related offence history, and diagnostic status for other substance-use disorders (cannabis, cocaine, alcohol, opiates, sedative/hypnotics, amphetamines). The extracted text does not report specific modelling diagnostics or whether analyses accounted for clustering or missing data.

In the full sample (n=25,622) participants were 71.0% male with a mean age of 31.24 years (SD=10.55, range 15–85). Racial composition was 59.9% African American, 38.6% Caucasian/white, and 1.5% other ethnoracial groups. Any hallucinogen use disorder was uncommon (1.0%) compared with other substance-use disorders (for example cannabis 44.0%, cocaine 28.2%, alcohol 22.6%). Median time under TASC supervision was 124.23 days (range 1.0–5139.57) and 38.4% of participants experienced supervision failure. Bivariate comparisons showed that having any hallucinogen use disorder (versus none) was associated with younger age, white/Caucasian race, current unemployment, never having been married, living with relatives/friends or in a shelter, relying on others as the primary financial support, a history of violent offence, a history of substance-related offence, and co-occurring diagnoses for cannabis, cocaine, alcohol, opiates, sedative/hypnotics, and amphetamines. Only one individual with a hallucinogen use disorder had no other substance-use disorder. Other variables (gender, educational attainment, matriculation status, income, number of prior arrests, prior felonies, pending offences) were not related to hallucinogen diagnostic status. In the multivariable logistic regression that included the listed covariates entered simultaneously, any hallucinogen use disorder was associated with a decreased probability of supervision failure. By contrast, diagnoses of cannabis, cocaine, alcohol, opiate and amphetamine use disorders were each associated with an increased probability of supervision failure; sedative/hypnotic disorder was not related to outcome. Additional independent predictors of supervision failure were younger age, non‑Caucasian race, never married status, current unemployment, living with relatives/friends, and reliance on others for financial support. The authors note that any hallucinogen use disorder was the third strongest predictor of supervision outcome, after cocaine and cannabis use disorders (both of which predicted failure). The extracted text does not provide odds ratios, confidence intervals or p-values for these associations.

Hendricks and colleagues interpret their findings as evidence that hallucinogen use is prospectively associated with a reduced likelihood of recidivism among substance-involved individuals under community supervision. They situate the results within both the historical literature on hallucinogen interventions in correctional settings and recent clinical work showing enduring psychological and behavioural changes after carefully administered hallucinogen sessions. The authors suggest that, if causal, the mechanisms discussed in the Introduction (for example mystical-type experiences, changes in openness, enhanced autobiographical recall and related neuropsychological effects) could account for reduced criminal behaviour. At the same time the investigators acknowledge important caveats. As an observational study, causal inference is limited: a common predisposition could have led some individuals both to use hallucinogens and to have lower recidivism, and although the regression controlled for many sociodemographic and psychosocial covariates this possibility cannot be eliminated. The extracted text also notes that type, dose, frequency and duration of hallucinogen use were not standardised or recorded; the diagnosis did not specify which hallucinogens were used, and generalisability to other populations is uncertain. The authors explicitly disclaim any endorsement of recreational hallucinogen use. Despite these limitations, the researchers argue that the naturalistic setting increases real-world relevance and that the findings justify randomised controlled trials of hallucinogen-based therapies in criminal justice populations. They propose that supervised, clinician-administered treatments could potentiate beneficial effects and identify the need for mechanistic studies (for example examining personality change) to improve treatment design. Overall, the authors present the results as a noteworthy exception to the typical positive association between substance use and criminal behaviour and call for continued scientific investigation of hallucinogens as potential therapeutic agents in this context.