Artificial intelligence to identify harmful alcohol use after early liver transplant for alcohol-associated hepatitis.

Early liver transplantation (LT) for alcohol-associated hepatitis (AH) is the fastest growing indication for LT, but prediction of harmful alcohol use post-LT remains limited. Among 10 ACCELERATE-AH centers, we examined psychosocial evaluations from consecutive LT recipients for AH from 2006 to 2017. A multidisciplinary panel used content analysis to develop a maximal list of psychosocial variables. We developed an artificial intelligence model to predict post-LT harmful alcohol use. The cohort included training (N = 91 among 8 centers) and external validation (N = 25 among 2 centers) sets, with median follow-up of 4.4 (IQR 3.0-6.0) years post-LT. In the training set, AUC was 0.930 (95%CI 0.862-0.998) with positive predictive value of 0.891 (95%CI 0.620-1.000), internally validated through fivefold cross-validation. In the external validation set, AUC was 0.692 (95%CI 0.666-0.718) with positive predictive value of 0.82 (95%CI 0.625-1.000). The model identified specific variables related to social support and substance use as highly important to predict post-LT harmful alcohol use. We retrospectively developed and validated a model that identified psychosocial profiles at LT predicting harmful alcohol use post-LT for AH. This preliminary model may inform selection and post-LT management for AH and warrants prospective evaluation in larger studies among all alcohol-associated liver disease being considered for early LT.

Identifying individual and environmental predictors of opioid and psychostimulant use among adolescents and young adults following outpatient treatment.

BACKGROUND: The United States (US) continues to grapple with a drug overdose crisis. While opioids remain the main driver of overdose deaths, deaths involving psychostimulants such as methamphetamine are increasing with and without opioid involvement. Recent treatment admission data reflect overdose fatality trends suggesting greater psychostimulant use, both alone and in combination with opioids. Adolescents and young adults are particularly vulnerable with generational trends showing that these populations have particularly high relapse rates following treatment. METHODS: We assessed demographic, psychosocial, psychological comorbidity, and environmental factors (percent below the poverty line, percent unemployed, neighborhood homicide rate, population density) that confer risk for opioid and/or psychostimulant use following substance use disorder treatment using two complementary machine learning approaches-random forest and least absolute shrinkage and selection operator (LASSO) modelling-with latency to opioid and/or psychostimulant as the outcome variable. RESULTS: Individual level predictors varied by substance use disorder severity, with age, tobacco use, criminal justice involvement, race/ethnicity, and mental health diagnoses emerging at top predictors. Environmental variabels including US region, neighborhood poverty, population, and homicide rate around patients' treatment facility emerged as either protective or risk factors for latency to opioid and/or psychostimulant use. CONCLUSIONS: Environmental variables emerged as one of the top predictors of latency to use across all levels of substance use disorder severity. Results highlight the need for tailored treatments based on severity, and implicate environmental variables as important factors influencing treatment outcomes.