A community driven modeling approach for identifying community & policy-level interventions to address the impact of structural racism and discrimination on adolescent substance use and mental health

Project Summary
Our overarching goal is to identify and assess policy and community-level interventions that counteract the
effects of SRD within neighborhoods, schools, and peer networks, on adolescent substance use and mental
health outcomes. We will leverage data from The Context Study, a large (n= 7,174), seven wave panel study of
public-school children in three racially diverse, predominantly rural counties in North Carolina. At wave one, all
6th, 7th, and 8th graders attending a public school in the three counties were eligible; the final wave occurred
when participants were in 10th-12th grade. Context includes extensive panel data on peer networks and
comprehensive residential geocodes for participants. We have linked these data to students’ school and criminal
records, Census, county tax-value property assessments, Google Maps, and the American Community Survey
to construct measures of SRD. Guided by the NC State Health Improvement Plan to increase health equity and
empirical evidence, we will use Context data, group model building, and an agent-based model (ABM) to project
the long-term impact of policy and community-level interventions (e.g., resource distribution, social integration,
and school discipline policies) that target SRD and its consequences on adolescent substance use and mental
health. Our specific aims are: Aim 1: Develop a detailed theoretical model of the systems rooted in SRD that
drive adolescent substance use and mental health outcomes and determine how these may be addressed
through policy and community-level interventions. We will use a community-based systems dynamics approach
to convene a group model building workshop with diverse community stakeholders to develop a causal loop
diagram of the system and identify a set of community co-created candidate policy and community-level
interventions that could modify the system and reduce the impact of SRD. Aim 2: Use ABM to compare the
expected long-term outcomes of various types and combinations of candidate policy and community-level
interventions to address SRD and its effects on adolescent substance use and mental health. We will use
information from the group model building workshops and leverage national models and a representative
synthetic population to develop an ABM to examine spatial and temporal effects of SRD on adolescent health.
We will validate the model on Context data, conduct sensitivity analyses to evaluate the robustness of model
results to uncertainty in model structures and parameters, and project the impact of candidate interventions. Aim
3: Determine key implementation factors of candidate policy and community-level interventions. For selected
interventions, we will conduct focus groups and in-depth interviews with community stakeholders to gain insight
on implementation factors including feasibility, acceptability, and scalability. We will analyze these data and
develop intervention maps of the most promising candidate interventions. Identifying and projecting systems-
level interventions that disrupt adolescent exposure to SRD has strong potential to move the field beyond
individually focused solutions and ultimately support strategies to reduce adolescent health inequities.