The Impact of Semaglutide on Health, Employment, and Involvement in the Criminal Justice System in the Alcohol and Opioid Use Disorder Population

2027-03-10

2030-02-28

Substance use: Number of heavy drinking days (among participants with alcohol use disorder) and days of opioid use (among participants with opioid use disorder), each assessed over the last 30 days via Timeline Followback, measured six months from enrollment.

Criminal justice involvement index using days in jail, any charges, number of charges, and any jail booking (all based on administrative data), over the 12 months following enrollment.

Labor market outcomes index constructed using administrative data on quarterly earnings, quarterly employment, number of quarters employed, and cumulative earnings over the 12 months following enrollment, and survey data on usual weekly earnings, any employment (usual weekly hours >= 5), and usual weekly hours.

For all outcome families (administrative and survey), we will aggregate outcomes into an inverse covariance-weighted index, in addition to reporting each outcome separately. We will adjust for multiple hypothesis testing within each family of outcomes.

Substance use: any and number of days of heavy drinking; any and number of days of opioid use; any days and number of days of stimulant use. Additional validation outcomes dependent on funding (not in index): any drinking and number of heavy drinking days using wearable alcohol monitoring (subset of individuals with alcohol use disorder at baseline); percentage of failed urine drug screenings (subset of individuals with a documented opioid use disorder at baseline).

Substance cravings: Penn Alcohol Craving Scale components; Opioid Craving Scale components.

Substance-related adverse events and high-acuity health care utilization: Self-reported nonfatal overdose or alcohol poisoning; Medicaid or county-funded drug or alcohol-related hospitalization or emergency department visit, residential treatment utilization, or inpatient detox or crisis substance use disorder treatment services utilization; drug- or alcohol-related death.

Criminal justice involvement: Individual components of criminal justice index; police encounters.

Labor market outcomes: Individual components of the labor market index; self-reported employment; self-reported earnings; self-reported hours worked.

Mental health and subjective well-being: mental health (PHQ-4); stress (PSS-4); life satisfaction (Cantril Ladder).

Behavioral measures of impulsivity and self-control: survey measures of impacts of substance use on relationships; survey and revealed-preference measures of impulsivity (BIS-Brief) and present focus; gambling, social media usage, high-interest debt (payday loans, credit card debt); sophistication about present bias; demand for commitment (not in index).

Housing and family stability: housing stability; net household savings; child protective services involvement; married or partnered; number of kids living in the same household.

All-cause mortality.

Measures of social benefits and costs (in dollars): In addition to these outcomes, we plan to measure the impact of our intervention by measuring the marginal value of public funds (MVPF) of the intervention. Inputs into this calculation will include (in dollars): Total public program expenditures [TANF, SNAP, Medicaid, county-funded health care, WIC, SSI, SSDI, UI, and housing (Housing Choice Vouchers, Low-Income Public Housing, Permanent Supportive Housing, Rapid Re-Housing, and rental assistance)]; predicted taxes; monetary cost of crime; all-cause mortality (Value of Statistical Life); cost of semaglutide.

This trial will enroll 1,200 individuals with alcohol use disorder and/or opioid use disorder at elevated criminal justice risk across multiple sites in Allegheny County, Pennsylvania. Half of the sample will be randomized to the treatment group to receive a one-year regimen of semaglutide therapy and will also have access to normal treatment services offered in the county. Individuals randomized to the control group will not receive any intervention as part of the trial and will have access only to standard treatment services offered in the county. Both groups will receive equal monetary compensation for participating in the trial and completing surveys. Outcomes will be measured using administrative data and baseline, midline, and endline surveys as well as monthly short surveys. The trial will also be registered at ClinicalTrials.gov.

Not available

Randomization done in office by a computer

Individual

No

1200 participants

1200 participants

600 participants control, 600 participants treatment