Understanding Women’s Enrollment in Alcohol Use Disorder Treatment: Testing Andersen’s Behavioral Model and the Role of Medicaid

Background: Alcohol use disorder (AUD) is a growing public health concern among women, with rising rates of binge drinking and AUD narrowing the gender gap in alcohol use. Despite the chronic nature of AUD, it remains a treatable condition when appropriate care is accessed. Nevertheless, fewer than 5% of women with AUD access treatment annually, often due to significant barriers such as stigma, affordability, and lack of support. Medicaid, a public insurance program for low-income individuals, may be a potential facilitator of treatment access. This study applies Andersen’s Behavioral Model of Health Service Use to examine factors that influence women’s AUD treatment enrollment and evaluates the role of Medicaid as a moderating factor.

Methods: Using pooled data from the 2015–2023 National Survey on Drug Use and Health (NSDUH), the study includes women aged 18 and older who met criteria for AUD (n = 28,729; weighted N = 14,383,241). Variables were grouped into predisposing factors, enabling resources, and health needs, consistent with Andersen’s Model. Confirmatory factor analysis (CFA) tested whether the theoretical model adequately explained treatment enrollment. Structural equation modeling (SEM) and logistic regression were used to compare Andersen’s Model with an empirically derived factor structure and to assess the moderating role of Medicaid.

Results: Findings revealed that, while Andersen’s Model formed distinct latent constructs, model fit was poor. Exploratory factor analysis (EFA) identified a refined two-factor model consisting of facilitators (income, education, government assistance, and physical health status) and barriers (mental health status, AUD severity, disability status, and physical health status), which significantly improved model fit and explained 70.4% of the variance in AUD treatment enrollment. Notable, physical health status loaded onto both constructs, suggesting its dual role as both a barrier and a facilitator. Facilitators were positively associated with treatment enrollment overall (β = 1.37, p < .001), but this effect was significantly weaker for women enrolled in Medicaid (β = -0.84, p = .008). In contrast, while barriers were negatively associated with treatment enrollment in general (β = -8.95, p < .001), this relationship reversed for Medicaid-enrolled women (β = 1.45, p = .034), suggesting that Medicaid may buffer the negative impact of barriers.

Discussion: These findings highlight the complexity of treatment access for women with AUD and suggest that Medicaid may play a dual role: dampening the effect of facilitators while simultaneously enabling access for women facing more substantial health-related barriers. Potential explanations include Medicaid's administrative limitations, structural barriers such as transportation, and stigma, particularly among women with intersecting marginalized identities. Additionally, Medicaid’s disability coverage and Section 1115 Demonstration waivers may uniquely support treatment access for women with greater clinical needs. This study creates a more nuanced understanding of treatment access for women with AUD and demonstrates that, while Andersen’s Model offers a useful foundation, empirical refinements reveal more accurate and actionable insights, particularly regarding the complex and sometimes contradictory role of Medicaid in enabling care. Findings underscore the importance of policies that address structural barriers and support women with greater needs.