(4) MOUD Treatment Receipt and Retention by Sociodemographics at a Safety-Net Health-System

Background & Introduction: The opioid epidemic in the United States remains a public health crisis with nearly 82,000 overdose deaths in 2022 involving opioids. In 2022, 3.7% of U.S. adults, over 9 million individuals, needed opioid use disorder (OUD) treatment. Medications for OUD (MOUD), especially buprenorphine and methadone, are highly effective in reducing the risk of death for those with OUD. However, only one in four individuals needing treatment receive MOUD.

As states and cities attempt to scale interventions to reduce opioid-related morbidity and mortality, there is a growing concern that the potential benefits of these prevention and treatment efforts may not be shared equitably. Indeed, overdose death rates for non-Hispanic (NH) Blacks and American Indian/Alaska Natives are now higher than for NH Whites. NH Blacks with OUD are less likely than NH Whites to receive buprenorphine treatment. There is a need to understand where in the OUD continuum of care these inequities may arise and their extent. We evaluated how sociodemographic and other health disparities were associated with MOUD receipt and retention at an urban safety-net integrated healthcare system.

Methods: This study took place at Denver Health (DH), an integrated public safety-net healthcare system in Denver, Colorado that includes a 500-bed hospital, outpatient primary and specialty care, 19 school-based clinics, and specialty substance treatment services. We conducted a retrospective cross-sectional analysis of patients with a validated marker of OUD in the centralized electronic health record (EHR) from January 1, 2023 to September 30, 2024. The two outcomes were 1) receipt of non-pain formulations of methadone or buprenorphine, and 2) retention on medication at 90 days (i.e. no 30-day breaks in medication receipt). Sociodemographic factors included patient-reported race/ethnicity, sex, age, insurance (public, commercial, uninsured, or other/unknown insurance), housing status, justice-involvement, co-occurring substance use disorders (SUD) (including stimulant use disorder [StUD], alcohol use disorder [AUD], and other SUD), clinical comorbidities measured by the Charlson comorbidity index, and co-occurring mental health conditions. We used multivariable log binomial regressions to estimate the association between covariates and both MOUD receipt and retention at 90 days. Results were reported as risk ratios (RR) with 95% confidence intervals (CI). We estimated predicted percent treated and retained at 90 days from the final adjusted models using marginal standardization.

Results: Of 8,378 patients with OUD in the study period, 40% received MOUD and 41% of those were retained on MOUD for at least 90 days.

Sociodemographic results from multivariable regression suggest that compared to 22-34-year-olds, younger (12-21) and older (55+) adults had a lower likelihood of treatment (12-21 RR: 0.78; 95% CI: 0.68-0.91; 55-64 RR: 0.82; 95% CI: 0.74-0.91; 65+ RR: 0.73; 95% CI: 0.63-0.85). However, older age was associated with higher likelihood of retention (55-64 RR: 1.17; 95% CI: 1.04-1.32; 65+ OR: 1.19; 95% CI: 1.03-1.36). Sex was not associated with likelihood of treatment or retention. Compared to NH Whites, NH Blacks and those with Other/Unknown race/ethnicity had lower likelihood of treatment (NH Black RR: 0.69; 95% CI: 0.61-0.77; Other/Unknown race/ethnicity RR: 0.85; 95% CI: 0.76, 0.95), but not significantly different likelihood of retention. Hispanics did not have different likelihoods of treatment or retention compared to NH Whites. Uninsured individuals were less likely to receive MOUD (RR: 0.80; 95% CI: 0.71-0.89) but had a similar likelihood of retention. Those that experienced homelessness were less likely to be treated (RR: 0.94; 95% CI: 0.89-0.99) or retained (RR: 0.82; 95% CI: 0.76-0.89). Those that were justice involved did not have different likelihood of treatment compared to those not involved in the justice system but were less likely to be retained in treatment (RR: 0.74; 95% CI: 0.68-0.81).

Results examining associations with other SUDs show those with StUD had higher likelihood of treatment than those without (RR: 1.60; 95% CI: 1.51-1.71) but potentially slightly lower retention (RR: 0.93; 95% CI: 0.86-1.01). Those with AUD were less likely to receive MOUD (RR: 0.84; 95% CI: 0.79-0.90) and be retained (RR: 0.86; 95% CI: 0.78-0.94) compared to those without. Presence of another SUD (beyond StUD and AUD) was not associated with different likelihoods for treatment or retention.

Results for other health characteristics show those with a mental health diagnosis were more likely to be treated (RR: 1.20; 95% CI: 1.13-1.27) and retained (RR: 1.43; 95% CI: 1.31-1.56) compared to those without. Compared to individuals without any medical comorbidities, those with a Charlson index score of 1-2 were more likely to be treated (RR: 1.20; 95% CI: 1.13-1.27). Those with a score of 3-4 and 5 or more did not have significantly different likelihoods of treatment or retention.

Conclusion & Discussion: This study reveals disparities in two stages of the OUD continuum. Notably, unhoused patients and those with AUD had lower likelihood of receiving and remaining in MOUD treatment. In contrast, those with mental health conditions were more likely to receive and be retained on treatment. Two factors (co-occurring StUD and age) showed an inverse relationship between treatment and retention. Youth, NH Black patients, and those without insurance were less likely to receive treatment.

This presentation will compare findings to literature. For example, the finding that those with co-occurring StUD had higher treatment rates contrasts with national trends; this may reflect more healthcare touchpoints among patients with StUD providing more opportunities to initiate, or DH-specific programming to engage patients with StUD. Still, these patients potentially experienced lower retention, indicating a need to address StUD-related barriers.

Limitations include the single-site (limiting generalizability) and the reliance on the EHR, which may not capture other important factors like OUD severity or willingness to engage in treatment. However, this is the first study of its kind to assess disparities at two stages of the continuum among the same population. Future research should evaluate programs and training aimed at reducing disparities in OUD treatment and retention.

References: Arifkhanova, A., Prieto, J. T., Davidson, A. J., Al-Tayyib, A., Hawkins, E., Kraus, E., McEwen, D., Podewils, L. J., Foldy, S., Gillespie, E., Taub, J., & Shlay, J. C. (2023). Defining Opioid-related Problems Using a Health Care Safety Net Institution's Inpatient Electronic Health Records: Limitations of Diagnosis-based Definitions. J Addict Med, 17(1), 79-84. https://doi.org/10.1097/ADM.0000000000001041

Dowell, D., Brown, S., Gyawali, S., Hoenig, J., Ko, J., Mikosz, C., Ussery, E., Baldwin, G., Jones, C. M., Olsen, Y., Tomoyasu, N., Han, B., Compton, W. M., & Volkow, N. D. (2024). Treatment for Opioid Use Disorder: Population Estimates - United States, 2022. MMWR Morb Mortal Wkly Rep, 73(25), 567-574. https://doi.org/10.15585/mmwr.mm7325a1

Dunphy, C. C., Zhang, K., Xu, L., & Guy, G. P., Jr. (2022). Racial‒Ethnic Disparities of Buprenorphine and Vivitrol Receipt in Medicaid. Am J Prev Med, 63(5), 717-725. https://doi.org/10.1016/j.amepre.2022.05.006

Kariisa, M., Davis, N. L., Kumar, S., Seth, P., Mattson, C. L., Chowdhury, F., & Jones, C. M. (2022). Vital Signs: Drug Overdose Deaths, by Selected Sociodemographic and Social Determinants of Health Characteristics - 25 States and the District of Columbia, 2019-2020. MMWR Morb Mortal Wkly Rep, 71(29), 940-947. https://doi.org/10.15585/mmwr.mm7129e2

National Institute on Drug Abuse. (2024). Drug Overdose Death Rates. Retrieved 6/28/2024 from https://nida.nih.gov/research-topics/trends-statistics/overdose-death-rates#:~:text=National%20Drug%20Overdose%20Deaths%E2%80%94Number%20Among%20All%20Ages%2C%201999%2D,overdose%20deaths%20reported%20in%202022.