Journal: Frontiers in Health Services, 2026, doi: 10.3389/frhs.2026.1781703
Authors: Kammarauche Aneni, Ching-Hua Chen, Jenny Meyer, Christina Mavromichali, Emmanuel Scaria, Gaoqianxue Liu, Youngsun T. Cho, & Lynn Fiellin
Abstract:
Introduction: Early identification of adolescents at risk for substance use is critical for timely intervention. However, standard screening tools like CRAFFT (Car, Relax, Alone, Forget, Family and Friends, Trouble) and S2BI (Screening to Brief Intervention) face significant barriers, including adolescent disclosure reluctance, limited clinic privacy, and administrative challenges with paper-based forms. Digital games offer a promising alternative by generating behavioral data that may serve as digital biomarkers of substance use risk through engaging, low-burden gameplay. The objective of this proof-of-concept study was to explore the utility of game log data collected during gameplay to predict substance use.
Methods: We analyzed game log data from 160 adolescents aged 11-14 years who played an HIV prevention game targeting high-risk behaviors including substance use, drawn from a larger randomized controlled trial (N = 333) conducted in schools. We extracted 240 gameplay-derived behavioral metrics reflecting executive function, decision-making, and inhibitory control-cognitive domains affected by substance use. Machine learning models (Support Vector Machine, Logistic Regression, Gradient Boosting, Neural Networks, Decision Tree, Random Forest) were trained to predict lifetime substance use and drug-refusal self-efficacy. A validation study with 36 newly recruited high school students tested associations with standardized measures of emotion regulation, impulsivity, and cognitive performance. Model performance was assessed using discrimination (e.g., AUC) and clinical utility (e.g., sensitivity) metrics.
Results: Predictive accuracy across all models was insufficient for clinical translation. For drug use prediction, AUC (SD) ranged from 0.458 (0.11) to 0.593 (0.07) (mean AUC <0.6 across all models), sensitivity ranged from 0 to 0.265 (0.18), and F-1 scores ranged from 0.398 (0.0) to 0.482 (0.08). For drug-refusal self-efficacy prediction, AUC ranged from 0.425 (0.10) to 0.592 (0.09), sensitivity ranged from 0.386 (0.14) to 0.785 (0.19), and F-1 scores ranged from 0.431 (0.06) to 0.598 (0.07). No model achieved the minimum threshold (AUC ≥0.7) suggested for clinical utility.
Discussion: Our study elucidates the challenges associated with extracting behavioral markers from naturalistic digital environments. We discuss the potential of game-based digital biomarkers as scalable, low-burden tools for screening and monitoring, and the limitations of our study that can inform future studies seeking to understand the feasibility of using in-game data as digital biomarkers of substance use risk in adolescents.
To read the full text of the article, please visit the publisher’s website.
Journal: Drug and Alcohol Dependence Reports, 2026, doi: 10.1016/j.dadr .2026.100436
Authors: Paul T. Harrell, Nicholas T. Williams, Kara E. Rudolph, Rachel Ayala Guzman, Ateeqa Ijaz, Nicole Rychagov, Aaron Null, & Silvia S. Martins
Abstract:
Background: Globally, over 10 million youth use e-cigarettes; United States use, in particular, dramatically rose in the late 2010’s, exceeding year-over-year increases from any other substance in over four decades. This rise has been partially attributed to youth online e-cigarette marketing exposure, but has not been appropriately studied.
Methods: We examined youth data from the Population Assessment of Tobacco and Health (PATH) spanning the dramatic increase period (Waves 4, 4.5, and 5; 2016-2019). We estimated average risk differences (RD) for Wave 5 e-cigarette harmfulness perception and use comparing all youth versus no youth reporting past-month online e-cigarette marketing exposure in Waves 4 and 4.5. We used a doubly robust, nonparametric targeted minimum loss-based estimator (TMLE) to estimate RD, incorporating PATH survey weights. Initial analyses adjusted for demographics, mental health issues, and other forms of e-cigarette marketing. Subsequent analyses adjusted for frequency of social media use, other substance use, and tobacco (non-e-cigarette) use.
Results: Initial analyses estimated that online marketing was associated with a 9% decrease in e-cigarette harmfulness perception (RD=-0.09, 95% C=-0.12, -0.05), or a Risk Ratio (RR) of 0.85 (95% CI=0.80, 0.91), as well as a 4% increase in current e-cigarette use (RD=0.04, 95% CI=0.02, 0.06; RR=1.36, 95% CI=1.15, 1.62). However, after adjusting for additional potential time-varying confounding variables, point estimates were close to null with 95% confidence intervals spanning the null.
Discussion: Frequency of social media use, other substance use, and/or tobacco use may be important confounding variables related to marketing and e-cigarette use that require further investigation.
To read the full text of the article, please visit the publisher’s website.
Journal: The American Journal of Emergency Medicine, 2026, doi: 10.1016/j.ajem.2026.05.025
Authors: Jennifer Mondle, Ashlee De Leon, Benoit Stryckman, & R. Gentry Wilkerson
Abstract:
Background: Take-home naloxone distribution is a key harm-reduction strategy in emergency departments and community programs to prevent opioid overdose deaths. Intranasal naloxone is commonly provided as either single-step devices or multi-step kits requiring assembly. Device complexity may influence the effectiveness of take-home naloxone programs, particularly among lay responders without prior training, yet usability among individuals most likely to witness or experience overdose is not well characterized.
Methods: We conducted a randomized usability study in an urban emergency department. Participants without prior naloxone training were randomized to administer naloxone using either a commercially manufactured single-step intranasal device or an improvised multi-step kit during a standardized simulated overdose scenario designed to assess first use usability. The primary outcome was successful completion of predefined critical steps; secondary outcomes included time to administration and participant-reported usability.
Results: Forty participants were enrolled (20 per group). During the pre-education simulation, successful completion of all critical steps occurred in 17/20 (85%) participants assigned to the single-step device compared to 4/20 (20%) assigned to the multi-step device (risk difference 65 percentage points; 95% CI 42-83; p < .001). Median time to successful administration was shorter with the single-step device (30 s [IQR 24-38] vs 58 s [45-75]; p < .001). Following a brief structured educational intervention, success rates improved and no longer differed significantly (100% vs 90%; p = .29), although administration time remained shorter with the single-step device (22 s [18-29] vs 35 s [28-47]; p = .002).
Conclusion: In this simulated overdose scenario among at-risk individuals, a single-step intranasal naloxone device produced higher first attempt success rates and faster administration than an improvised multi-step kit. Although brief structured education substantially improved performance with the multi-step device, administration times remained shorter with the single-step device. These findings suggest that device complexity may influence the real-world effectiveness of emergency department and community naloxone distribution programs, particularly when structured training is not consistently available.
To read the full text of the article, please visit the publisher’s website.
Journal: Human Brain Mapping, 2026, doi: 10.1002/hbm.70545
Authors: Jessica S. Flannery, Elizabeth S. Escalante, Ruofan Ma, Kristen A. Lindquist, & Eva H. Telzer
Abstract:
Adolescence is a period of social development marked by increased sensitivity to social feedback and substance use experimentation. Although reinforcement learning (RL) models provide a powerful framework for examining how individuals learn from experience, they have rarely been applied to understand how adolescents learn from social experiences or how these processes relate to real-world behavioral outcomes. In a sample of 261 youth (11.0 ± 1.6 years old), we applied computational modeling to a novel social RL fMRI paradigm. Q-learning models estimated individual differences in learning and trial-by-trial prediction errors, which were used as a parametric modulator to assess brain activity that tracked the degree to which social feedback was better or worse than expected. Among older participants (n = 73; 12.9 ± 0.9 years old), we assessed associations between RL metrics and measures of substance use propensity. Greater substance use curiosity and household exposure to substance use were both linked to weaker striatal prediction error tracking of better-than-expected outcomes. However, among youth with substance-using peers, curiosity was associated with elevated striatal prediction error signals and better positive RL performance. Findings suggest that both hypo- and hyper-sensitivity to positive reinforcement learning signals may confer an increased propensity for substance use, possibly through distinct pathways.
To read the full text of the article, please visit the publisher’s website.
Journal: Substance Abuse Treatment, Prevention, & Policy, 2026, doi: 10.1186/s13011-026-00731-8
Authors: Jennifer D. Hall, Maria N. Danna, Viviane Cahen, Andrea Baron, & Camille C. Cioffi, & Deborah J. Cohen
Abstract:
Background: Substance use has increased among pregnant and postpartum people in the last decade, yet few pregnant individuals receive prenatal care and treatment for substance use disorder (SUD). The Nurture Oregon model aims to integrate medical care with SUD treatment and provide destigmatized care through regular visits in an integrated care setting. We studied one family medicine, one behavioral health, and two SUD treatment organizations in rural counties with high SUD rates and limited resources that were funded to implement the model.
Methods: To examine the startup and early implementation phases of Nurture Oregon, we used a prospective, observational design to appreciate the effort from multiple perspectives. We observed program development meetings, team operations, and conducted semi-structured interviews with organization leaders, team members, and community partners. We used an inductive and comparative approach to identify sites’ startup and early implementation activities and challenges.
Results: Each site started with different program elements at baseline. This influenced the model elements each organization worked on during startup and the implementation challenges they experienced. The SUD and behavioral health organizations did not fully integrate care due to difficulty developing partnerships with medical organizations; they leveraged peers and doulas to provide cohesion for patients. The family medicine site was the only site that fully implemented the model, but they experienced barriers to financially supporting their peer workforce due to licensing and reimbursement policy constraints. All sites experienced challenges collaborating with hospital labor and delivery departments, and they all took steps to address patient housing needs by connecting patients to housing resources or acquiring housing units.
Conclusions: Implementing a care model to integrate medical and SUD treatment for pregnant individuals is difficult to accomplish but has the potential to make a significant difference in maternal and child health outcomes, recovery success, and prevention of foster care placement. Experiences during the startup and early implementation phases can shape the entire trajectory of a program and determine its long-term success. Our work shows the early challenges that need to be addressed to build an integrated program.
To read the full text of the article, please visit the publisher’s website.