Faculty Bibliography

BACKGROUND:Neighborhood-level overdose risk may vary over time. In Rhode Island, we developed and validated a machine learning model to identify the 20 percent of census block groups (CBGs) at the highest predicted risk of future overdose death. We updated this model periodically between November 2021 and August 2024 to generate six sets of predictions. This study aims to characterize the trajectory of each CBG's predicted overdose risk over time across these six periods. METHODS:In each prediction period, CBGs were designated as "high risk" or not designated as "high risk" based on our model's 20 percent predicted overdose risk threshold. We implemented sequence analysis to describe unique trajectories in each CBG's risk designation over each prediction period. We then calculated optimal matching distances to estimate dissimilarity between each pair of trajectories and applied agglomerative hierarchical clustering to group similar trajectories. RESULTS:The 809 CBGs included in this study followed 60 unique trajectories in predicted overdose risk designation over the six prediction periods. Clustering of trajectories favored a solution with five trajectory groups. Most CBGs (73.4 %) were rarely or never designated as "high risk", 7.9 % of CBGs were always designated as "high risk", and the remaining 18.7 % were designated as "high risk" in multiple prediction periods, represented by trajectory groups with different patterning over time. CONCLUSIONS:Given the substantial variability in which CBGs were at highest overdose risk over time, dynamic machine learning predictions may inform harm reduction resource allocation by identifying neighborhoods with emerging needs.

IMPORTANCE/UNASSIGNED:There has been a national increase in fatal and nonfatal overdoses involving stimulants, and 4.5 million US individuals meet criteria for stimulant use disorder (UD), with the highest prevalence in young adults. However, limited information exists on trends in diagnosed stimulant UD. OBJECTIVE/UNASSIGNED:To estimate trends in the proportion of adolescent and young adult Medicaid enrollees diagnosed with a stimulant UD from 2001 to 2020. DESIGN, SETTING, AND PARTICIPANTS/UNASSIGNED:A repeated cross-sectional study (2001-2020) was conducted using administrative health care claims data from Medicaid (public insurance program in US). Publicly insured adolescents (aged 13-17 years) and young adults (aged 18-24 or 25-29 years) from 42 US states were included. Data were analyzed from January 2025 to July 2025. MAIN OUTCOMES AND MEASURES/UNASSIGNED:Stimulant UD was defined as an inpatient or outpatient International Classification of Diseases diagnosis code in the year, with cocaine UD and noncocaine psychostimulant UD evaluated separately. The annual proportion with stimulant UD diagnoses was stratified by age group, sex, race and ethnicity, and presence of attention-deficit/hyperactivity disorder (ADHD) diagnosis or stimulant prescription in the year. Differences comparing 2001 with 2020 were summarized. Characteristics of those diagnosed with stimulant UD in 2020 were described. RESULTS/UNASSIGNED:The sample included 5.7 million (2001) to 16.1 million Medicaid enrollees (2020) per year; in 2020, 54.2% were female, and 7.1 million were adolescents. From 2001 to 2020, the proportion diagnosed with noncocaine psychostimulant UD increased from 0.09% to 0.49% (prevalence ratio [PR], 5.47 [95% CI, 5.20-5.75]) in those aged 18 to 24 years, from 0.13% to 1.63% (PR, 12.55 [95% CI, 11.83-13.31]) for ages 25 to 29 years, and from 0.10% to 0.91% among young adults aged 18 to 29 years. Among adolescents, the proportion diagnosed with noncocaine psychostimulant UD varied between 0.03% and 0.07%. The proportion diagnosed with cocaine UD was stable in young adults (range, 0.17%-0.34% [18-24 years] and 0.53%-0.79% [25-29 years]) and declined in adolescents (from 0.04% to 0.01%). Cocaine and noncocaine psychostimulant UD diagnoses were 2 to 4 times more common in patients with an ADHD diagnosis or stimulant prescription. Most patients diagnosed with a stimulant UD in 2020 were also diagnosed with a mental health disorder (68%-82%) or other substance UD (72%-78%). CONCLUSIONS AND RELEVANCE/UNASSIGNED:The prevalence of noncocaine psychostimulant UD diagnoses in young adult Medicaid patients increased over the last 2 decades, potentially associated with an increasing use of prescription and illicit stimulants along with increased clinical detection. These trends raise concerns given recent rises in stimulant-involved overdose fatalities and stress the need for evidence-based stimulant UD treatments for young people.

With the overdose crisis continuing to pose significant challenges in North America, harm reduction strategies are critical for public health systems to reduce mortality and morbidity. Despite the considerable strides in harm reduction research, high-quality evidence for decision-making is limited. This is compounded by a variation in reported outcomes, drug supply, administration changes, and policy and social impacts, which further challenge researchers and practitioners in their efforts to implement effective, nimble harm reduction interventions. Adoption of common data elements (CDEs) and common outcome measures (COMs) helps researchers standardize and enhance data collection and outcome reporting, ultimately improving the comparability and generalizability of research findings. To accelerate the pace and use of CDEs, members of the NIDA HEAL Research on Interventions for Stability and Engagement (RISE) engaged in prospective semantic harmonization and consensus on CDEs and COMs using a rigorous pragmatic Delphi community informed approach. This process resulted in a set of CDEs and COMs that standardized data collection and reporting across 10 harm reduction research projects. This paper describes this process and presents the derived CDEs and COMs, along with key considerations, challenges encountered, and lessons learned.

Between May 2020 and December 2021, there were 159,872 drug overdose deaths in the US. Higher eviction rates have been associated with higher overdose mortality. Amid the economic turmoil caused by the COVID-19 pandemic, 43 states and Washington, DC, implemented eviction moratoria of varying durations. These moratoria reduced eviction filing rates, but their impact on fatal drug overdoses remains unexplored. We evaluated the effect of these policies on county-level overdose death rates by focusing on the dates the state eviction moratoria were lifted. We obtained mortality data from NCHS and eviction moratoria dates from the COVID-19 US State Policy Database. We employed a longitudinal targeted minimum-loss-based estimation with Super Learner to flexibly estimate the average treatment effect (ATE) of never lifting the moratoria. Lifting state eviction moratoria was associated with a 0.14 per 100,000 higher rate of monthly overdose mortality (95%CI: -0.03, 0.32), although confidence intervals were wide and included zero. Eviction moratoria may not be sufficient to prevent overdose mortality during crises such as the COVID-19 pandemic.

BACKGROUND/UNASSIGNED:Understanding sex differences in the effects of cannabis legalization and increasing risk for cannabis use disorder (CUD) is important. We hypothesized that from 2005 to 2019, increases in CUD prevalence due to state medical or recreational cannabis laws (MCL; RCL) would differ among male and female veterans treated at the U.S. Veterans Health Administration (VHA), with greater increases among females. METHODS/UNASSIGNED:Data obtained through the VHA Corporate Data Warehouse included veterans 18-75 years with ≥1 VHA primary care, emergency department, or mental health visit in a given year, 2005-2019. Staggered-adoption difference-in-difference analyses were used to estimate the role of MCL and RCL on trends in CUD diagnostic prevalence, fitting a linear binomial regression model with fixed effects for state and categorical year, time-varying cannabis law status, state-level sociodemographic covariates, patient-level age group (18-35, 36-64, 65-75 years), race and ethnicity. RESULTS/UNASSIGNED:CUD prevalences increased in both sexes. CUD increased more in states enacting MCL and RCL than in states that did not enact CL. However, no CUD prevalence increases attributable to the change from no-CL to MCL-only or MCL to RCL differed significantly by sex, with one exception (greater in males aged 35-64). CONCLUSIONS/UNASSIGNED:Increases in CUD prevalence following MCL or RCL enactment were greater than in states with no-CL, but generally did not show differences by sex. The increases in CUD prevalence occurring for males and females throughout the study years indicate the need for cannabis use screening by medical providers and the importance of offering evidence-based treatments for CUD.

Following federal regulatory changes during the COVID-19 pandemic, Rhode Island expanded methadone access for opioid treatment programs (OTPs) in March 2020. The policy, which permitted take-home dosing for patients, contrasted with longstanding restrictions on methadone. This study used patient-level OTP admission and discharge records to compare six-month retention before and after the policy change. We conducted a retrospective cohort study of 1,248 patients newly admitted to OTPs between March 18 and June 30 of 2019 (pre-policy) and 2020 (post-policy). We used logistic regression to estimate associations with retention before and after the policy and used a machine learning approach, the Heterogeneous Treatment Effect (HTE)-Scan, to explore heterogeneity in retention across subgroups. Overall, we found no change in retention following the policy, with an adjusted OR of 1.08 (95% CI: 0.80-1.45) and adjusted RR of 1.03 (0.90-1.18). Using HTE-Scan, we identified two subgroups with significantly increased retention above the overall cohort: (1) patients with below high school education and past-month arrest and (2) male, non-Hispanic white or Hispanic/Latino patients reporting heroin or fentanyl use with past-month arrest. We identified no subgroups with significantly decreased retention. Collectively, findings suggest that expanded methadone access may benefit vulnerable populations without harming overall retention.

PURPOSE/OBJECTIVE:In this article, we studied whether pathways in substance use disorder (SUD) treatment differ among people admitted to residential versus ambulatory settings. METHODS:We analyzed a retrospective cohort of 84,755 adults (ages ≥ 18) in Chilean SUD treatment during 2010-2019, creating a comparable sample of 11,226 pairs in ambulatory and residential treatment through cardinality matching. We used a nine-state multistate model, stratifying readmissions by baseline treatment outcome (i.e., completion vs. noncompletion) from admission to the third readmission. We estimated transition probabilities and lengths of stay in states at three-month, one-year, three-year, and five-year follow-ups. Sensitivity analyses tested different model specifications and estimated E-values. RESULTS:Patients in residential settings (vs. ambulatory) had greater treatment completion probabilities (difference at three months; 3.4% [95% CI: 2.9%, 3.9%]), and longer treatment retention (e.g., 1.6 days longer at three months, 95% CI: 0.8, 2.3). Patients in residential vs. ambulatory settings had higher first readmission probabilities regardless of baseline treatment outcome (e.g., three-month difference: 5.7% if completed baseline [95% CI: 4.4%, 7.0%] and 8.0% if did not complete baseline [95% CI: 6.7, 9.3%]). Third readmission probabilities were higher only among patients in residential settings with an incomplete baseline treatment (at least 3.7%; 95% CI: 0.2%, 7.3% at 1-year). CONCLUSION/CONCLUSIONS:Patients in residential settings at baseline were more likely to experience a second treatment and a third readmission among patients with incomplete treatments. Findings underscore the importance of completing initial SUD treatments to reduce readmissions. Residential treatments might require additional strategies to prevent readmissions.

BACKGROUND/UNASSIGNED:We investigated whether the associations of state medical and recreational cannabis legalization (MCL, RCL enactment) with increasing prevalence of Cannabis Use Disorder (CUD) differed among patients in the United States (US) Veterans Health Administration (VHA) who did or did not have common psychiatric disorders. METHODS/UNASSIGNED:Electronic medical record data (2005-2022) were analyzed on patients aged 18-75 with ≥1 VHA primary care, emergency department, or mental health visit and no hospice/palliative care within a given year (sample sizes ranging from 3,234,382 in 2005 to 4,436,883 in 2022). Patients were predominantly male (>80%) and non-Hispanic White (>60%). Utilizing all 18 years of data, CUD prevalence increases attributable to MCL or RCL enactment were estimated among patients with affective, anxiety, psychotic-spectrum disorders, and Any Psychiatric Disorder (APD) using staggered difference-in-difference (DiD) models and 99% Confidence Intervals (CIs), testing differences between patient groups with and without psychiatric disorders via non-overlap in the 99% CIs of their DiD estimates. FINDINGS/UNASSIGNED:Among APD-negative patients, CUD prevalence was <1.0% in all years, while among APD-positive patients, CUD prevalence increased from 3.26% in 2005 to 5.68% in 2022 in no-CL states, from 3.51% to 6.35% in MCL-only states, and from 3.41% to 6.35% in MCL/RCL states. Among the APD group, DiD estimates of MCL-only and MCL/RCL effects were modest-sized, but the lower bound of the 99% CI for the DiD estimate for MCL-only and MCL/RCL effects was larger than the upper bound of the 99% CI among the no-APD group, indicating significantly stronger MCL-only and MCL/RCL effects among patients with APD. Results were similar for MCL-only and MCL/RCL effects among disorder-specific groups (depression, post-traumatic stress disorder [PTSD], anxiety or bipolar disorders) and for MCL/RCL effects among patients with psychotic-spectrum disorders. INTERPRETATION/UNASSIGNED:Cannabis legalization contributed to greater CUD prevalence increases among patients with psychiatric disorders. However, modest-sized DiD estimates suggested operation of other factors, e.g., commercialization, changing attitudes, expectancies. As cannabis legalization widens, recognizing and treating CUD in patients with psychiatric disorders becomes increasingly important. FUNDING/UNASSIGNED:This study was supported by National Institute on Drug Abuse grant R01DA048860, the New York State Psychiatric Institute, and the VA Centers of Excellence in Substance Addiction Treatment and Education.

IMPORTANCE/UNASSIGNED:In the context of the US opioid crisis, factors associated with the prevalence of opioid use disorder (OUD) must be identified to aid prevention and treatment. State medical cannabis laws (MCL) and recreational cannabis laws (RCL) are potential factors associated with OUD prevalence. OBJECTIVE/UNASSIGNED:To examine changes in OUD prevalence associated with MCL and RCL enactment among veterans treated at the Veterans Health Administration (VHA) and whether associations differed by age or chronic pain. DESIGN, SETTING, AND PARTICIPANTS/UNASSIGNED:Using VHA electronic health records from January 2005 to December 2022, adjusted yearly prevalences of OUD were calculated, controlling for sociodemographic characteristics, receipt of prescription opioids, other substance use disorders, and time-varying state covariates. Staggered-adoption difference-in-difference analyses were used for estimates and 95% CIs for the relationship between MCL and RCL enactment and OUD prevalence. The study included VHA patients aged 18 to 75 years. The data were analyzed in December 2023. MAIN OUTCOME AND MEASURES/UNASSIGNED:International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) or International Statistical Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) OUD diagnoses. RESULTS/UNASSIGNED:From 2005 to 2022, most patients were male (86.7.%-95.0%) and non-Hispanic White (70.3%-78.7%); the yearly mean age was 61.9 to 63.6 years (approximately 3.2 to 4.5 million patients per year). During the study period, OUD decreased from 1.12% to 1.06% in states without cannabis laws, increased from 1.13% to 1.19% in states that enacted MCL, and remained stable in states that also enacted RCL. OUD prevalence increased significantly by 0.06% (95% CI, 0.05%-0.06%) following MCL enactment and 0.07% (95% CI, 0.06%-0.08%) after RCL enactment. In patients aged 35 to 64 years and 65 to 75 years, MCL and RCL enactment was associated with increased OUD, with the greatest increase after RCL enactment among older adults (0.12%; 95% CI, 0.11%-0.13%). Patients with chronic pain had even larger increases in OUD following MCL (0.08%; 95% CI, 0.07%-0.09%) and RCL enactment (0.13%; 95% CI, 0.12%-0.15%). Consistent with overall findings, the largest increases in OUD occurred among patients with chronic pain aged 35 to 64 years following the enactment of MCL and RCL (0.09%; 95% CI, 0.07%-0.11%) and adults aged 65 to 75 years following RCL enactment (0.23%; 95% CI, 0.21%-0.25%). CONCLUSIONS AND RELEVANCE/UNASSIGNED:The results of this cohort study suggest that MCL and RCL enactment was associated with greater OUD prevalence in VHA patients over time, with the greatest increases among middle-aged and older patients and those with chronic pain. The findings did not support state cannabis legalization as a means of reducing the burden of OUD during the ongoing opioid epidemic.