Intersectional Inequities of Unmet Substance Use Disorder Treatment Need Based on Race, Sex, Age Group, and Rurality

Background and Purpose Substance use disorder (SUD) treatment uptake remains below 12% among those who could benefit from treatment in the US. Although SUDs present a significant public health concern across populations, individuals from marginalized populations bear a disproportionate burden. However, little is known about the potential impacts of individuals’ intersecting identities on SUD treatment disparities. We examined disparities in unmet SUD treatment need at the intersections of race, sex, age, and rurality.

Methods: This cross-sectional study used data from the National Survey on Drug Use and Health (NSDUH), conducted annually among noninstitutionalized US civilians aged 12 years or older. For the current analyses, we pooled data from the 2015-2019 NSDUH surveys given methodological changes in 2020 due to COVID-19. To define the population with unmet treatment needs for SUD, we focused on the individuals who needed SUD treatment in the past year but did not receive it. Respondents were classified as needing SUD treatment if they met criteria for a past-year SUD and/or expressed past-year need for SUD treatment. Specifically, the 2015-2019 NSDUH included questions assessing SUD symptoms including for illicit substances (e.g., cannabis, hallucinogens, methamphetamine, cocaine, heroin, etc.), prescription drugs (e.g., pain relievers, sedatives, tranquilizers, etc.), and alcohol, based on the DSM-4 standard diagnostic criteria. We used intersectional multilevel analysis of individual heterogeneity and discriminatory accuracy (MAIHDA) models which employed two-level Bayesian multilevel logistic regression models. Within these models, individuals were nested within one of 54 social strata based on all possible combinations of sex (i.e., male, female), age (i.e., 12-17, 18-25, 26+), race/ethnicity (i.e., non-Hispanic white, non-Hispanic Black, Hispanic), and rurality (i.e., urban, suburban, rural). A null model was computed for unmet treatment need with random intercepts for social strata, producing two intersectional parameters. The first parameter was variance partition coefficient (VPC) which reflects the proportion of variance observed between (vs. within) strata where the higher VPC indicates more significant differences between strata. The second parameter was stratum-specific predicted outcome prevalence and 95% credible intervals (CrIs).

Results: The analytic sample (N=253,389) included individuals (ages 12+) who needed SUD treatment; 47.5% were men; 20.7% were Hispanic, 14.2% were non-Hispanic Black, and 65.1% were non-Hispanic White. Respondents dwelling in urban, suburban, and rural counties represented 44.8%, 35.3%, and 19.9% of the sample, respectively. We observed a complex social pattern of unmet SUD treatment need. The highest estimated unmet need was among Non-Hispanic Black female adolescents (ages 12-17) in suburban counties (96.7%; 95% CrI. [96.0%-97.4%]), followed by those in urban counties (96.4%; 95% CrI. [95.7%-97.2%]) and those in rural counties (96.4%; 95% CrI. [94.5%-97.9%]).

Conclusions: Our findings underscore the heightened unmet SUD treatment need among multiple marginalized groups. These results offer guidance for tailored service delivery by clinicians, targeted interventions by community organizations, and implementation strategies aimed at reducing unmet SUD treatment need through a health equity approach with an intersectionality framework.