A precision prevention approach to identifying and addressing behavioral health treatment gaps to reduce the population burden of suicide

About This Grant

The United States is facing a suicide crisis. Despite investments by national agencies, suicide death rates rose 37% between 2001 and 2023. This pattern signals a critical need to alter existing prevention strategies. Rising rates of alcohol use disorder (AUD), along with co-occurring mental illnesses and drug use disorders, are an important contributor to rising rates of suicide. However, which subgroups are at highest risk of suicidal thoughts and behaviors (“suicidality”) and the types and combinations of mental and substance use disorders (MSUDs) that place distinct subgroups at greater risk are not well understood. Increasing uptake of behavioral health services to treat MSUDs is a critical component of the 2024 National Strategy for Suicide Prevention and the NIAAA Strategic Plan. Yet most adults with MSUDs do not receive evidence-based treatments, such as naltrexone for AUD or therapy for depression, and we do not know which types of treatment gaps mediate suicide risk for distinct subgroups. For example, co-occurring AUD and depression may drive suicide risk for urban women, whereas untreated opioid use disorder may put disabled male veterans at higher suicide risk. Health departments and health systems have evidence-based interventions to address barriers to accessing MSUD treatment, such as health insurance enrollment outreach, telehealth, stigma reduction campaigns, and mental health service navigators. However, we do not know which barriers or interventions are the most important for distinct subgroups. To maximize impact with limited resources, a precision prevention approach would develop tailored interventions that address the unique barriers to care faced by each subgroup. This study aims to address the suicide crisis by identifying the optimal tailoring and targeting of interventions to expand MSUD treatment uptake among the subgroups at highest risk of suicide. We will leverage data from the National Surveys on Drug Use and Health and machine learning tools developed by our team to: (1a) Characterize the subgroups of the US population at highest risk of suicidal ideation, plans, and attempts across a range of intersecting characteristics (e.g., age, sex, education, geography, employment, disability, military service); (1b) Identify the specific MSUDs (e.g., AUD, depression) that affect suicidal individuals within each subgroup; (2) Assess the mediating effects of unmet need for treatment for each MSUD, and co-occurring combinations of MSUDs, as mechanisms underlying disparities in suicidality; (3) Determine the optimal combinations and targeting of evidence-based interventions that increase MSUD treatment uptake to maximally reduce population suicidality. Aim 3 will generate a tool that models the hypothetical impacts of alternative interventions and, given a target population, indicates which combinations of interventions should be used, to whom they should be targeted, and for what MSUDs to effectively mitigate barriers to MSUD treatment and maximize population-level reductions in suicidality. The tool will be piloted with several already- recruited suicide prevention agencies and disseminated via a web-based dashboard.