Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Feb 1.
Published in final edited form as: J Subst Abuse Treat. 2021 Jun 11;133:108533. doi: 10.1016/j.jsat.2021.108533

Black clients in expansion states who used opioids were more likely to access medication for opioid use disorder after ACA implementation

Natrina L Johnson 1,*, Sugy Choi 1, Carolina-Nicole Herrera 1
PMCID: PMC8664894  NIHMSID: NIHMS1720935  PMID: 34218991

Abstract

Introduction:

Black people in the United States who use opioids receive less treatment and die from overdoses at higher rates than White people. Medication for opioid use disorder (MOUD) decreases overdose risk. Implementation of the Affordable Care Act (ACA) in the United States was associated with an increase in MOUD. To what extent racial disparity exists in MOUD following ACA implementation remains unclear. Using a national sample of people seeking treatment for opioids (clients), we compared changes in MOUD after the ACA to determine whether implementation was associated with increased MOUD for Black clients relative to White clients.

Methods:

We identified 878,110 first episodes for clients with opioids as primary concern from SAMHDA’s Treatment Episodes Dataset-Admissions (TEDS-A; 2007–2018). We performed descriptive and logistic regression analyses to estimate odds of MOUD for Black and White clients by Medicaid expansion status. We interacted ACA implementation with racial group and performed subpopulation analyses for Medicaid enrollees and criminal justice–referred clients.

Results:

In expansion states post-ACA, MOUD increased from 33.6% to 51.3% for White clients and from 36.2% to 61.7% for Black clients. Pre-ACA, Black clients were less likely than White clients to use MOUD (adjusted odds ratio (aOR) = 0.88, 99th Confidence Interval (CI) = [0.85, 0.91]), and post-ACA, the change in odds of MOUD did not differ. Criminal justice–referred clients experienced less of a change in odds of MOUD among Black clients than among White clients (aOR = 0.74, CI = [0.62, 0.89]). Among Medicaid-insured clients, the change in odds of MOUD among Black clients was larger (aOR = 1.16, CI = [1.03, 1.30]). In the non–expansion states before 2014, Black clients were less likely to receive MOUD (aOR = 0.86, CI = [0.77, 0.95]) than White clients. After 2014, the change in odds of MOUD increased more for Black clients relative to White clients (aOR = 1.24, CI = [1.07, 1.44]). We did not find significant changes in MOUD for clients referred through the criminal justice system or with Medicaid.

Conclusion:

The ACA was associated with increased use of MOUD among Black clients and reduction in treatment disparity between Black and White clients. For criminal justice–referred Black clients, disparities in MOUD persist. Black clients with Medicaid in expansion states had the greatest reduction in disparities.

Keywords: Medication for addiction treatment (MAT), Medicaid expansion, Race, Substance use disorder (SUD), Policy evaluation

1. Introduction

The United States is still facing an opioid epidemic. Although the number of deaths related to the use of prescription opioids has decreased in recent years, the number of deaths resulting from the consumption of synthetic opioids has increased dramatically since 2015 (National Institute on Drug Abuse, 2020). As of 2016, close to 3 million Americans met the criteria for having an opioid use disorder (OUD; SAMHSA, 2016). In 2018, nearly 50,000 people died due to opioid use (National Institute on Drug Abuse, 2020). Medications for OUD (MOUD) are efficacious and evidence-based treatments, but many people with OUD do not utilize MOUD (Bagley et al., 2020; Hadland et al., 2017; Larochelle et al., 2018; Volkow et al., 2014a).

Research has attributed the current opioid epidemic to the increase in prescription opioids in the early 2000s (Madras, 2017; Paulozzi et al., 2006), which largely affected White people in the United States (Anderson et al., 2009; Singhal et al., 2016). Based on 2017 U.S. national estimates, the rate of past-year opioid misuse was slightly higher among White people compared to Black people (4.7% vs 3.5%) (Center for Behavioral Health Statistics, 2018), and the opioid overdose rate was higher as well (19.4 vs 12.9 per 100,000 in the population) (The Henry J. Kaiser Family Foundation, 2016). The rate of death due to opioids has increased dramatically for Black people in recent years, as much as 818% between 2014 and 2017 (Kiechle & Gonzalez, 2020), and this rate is now outpacing opioid-related deaths among White people (Furr-Holden et al., 2020). Further, a recent study of commercially insured patients showed that Black patients are less likely to receive follow-up after an emergency room visit for opioid overdose (Kilaru et al., 2020). Given the recent trend in opioid overdose deaths, and that Black people in the United States have historically underutilized treatment for substance use, including treatment for opioid use, (Marsh et al., 2009; Mennis & Stahler, 2016; Wu et al., 2016), understanding how recent efforts to expand access to treatment has changed rates of MOUD for Black people who use opioids is important.

1.1. The affordable care act and treatment for substance use disorders

One of the goals of the Affordable Care Act (ACA) was to expand health coverage to mitigate barriers in access to care by mandating that health insurance plans include coverage of mental health and treatment of substance use disorders (SUD) (Geissler & Evans, 2020). In the years prior to the ACA, Black and Hispanic clients were less likely to complete SUD treatment compared to White clients (Saloner & Cook, 2013). After adjusting for need, Black and Hispanic clients had a completion rate of 44.3% and 46.1%, respectively, compared to 52.4% completion rate among White clients for non–alcohol substance use treatment (Saloner & Cook, 2013). Lower completion rates were associated with socioeconomic factors, such as lack of health insurance, unemployment, and fewer years of formal education. Since the implementation of the ACA, the number of people who are uninsured has declined and SUD treatment has increased (Andrews et al., 2019; Cohen & Martinez, 2015). The use of MOUD increased to a greater extent in expansion states compared to non–expansion states during the years following the implementation of the ACA (Meinhofer & Witman, 2018; Mojtabai et al., 2019). An early study of SUD treatment by racial-ethnic groups did not show significant increases in addiction treatment after the ACA (Creedon & Lê Cook, 2016). However, Krawczyk et al. examined outcomes for 2014 only and found that the odds of MOUD were significantly higher among Black and Hispanic clients, compared to White clients (Krawczyk, Feder, et al., 2017).

This study addresses the mixed evidence associating the ACA with changes in OUD treatment. We sought to extend the growing body of research on racial disparities by comparing the changes in MOUD associated with ACA implementation among Black and White clients at federally subsidized treatment facilities in expansion states. As a secondary objective, we examined changes in MOUD by race before and after 2014 in the non–expansion states.

2. Methods

2.1. Data source and sample

We used the Treatment Episode Dataset-Admissions (TEDS-A) collected by the Center for Behavioral Health Statistics and Quality within the Substance Abuse and Mental Health Services Administration (SAMHDA) (Substance Abuse and Mental Health Data Archive, 2020). TEDS is a national dataset of substance treatment episodes only from federally subsidized clinics and, therefore, does not include clients seen in the private office-based addiction treatment setting. Our study period ran from 2007 to 2018 (SAMHSA, 2019, 2020). We excluded data from 2014 due to national transitions associated with ACA implementation. Our analysis included eighteen (18) states that did not expand Medicaid before 2019 (“non-expansion”, Appendix); and the nineteen (19) states and District of Columbia that expanded Medicaid by January 1, 2014 (“expansion”, Appendix). While several states including the District of Columbia expanded some Medicaid access prior to January 2014, states did not implement most of the regulatory changes resulting from the ACA until 2014, including Medicaid access for the broader population of nonelderly adults. We excluded states with (1) approved Section 1115 waivers allowing them to operate their Medicaid expansion programs in ways that limit benefits and eligibility, (2) states that enacted Medicaid expansion after January 1, 2014, and before January 1, 2019, (3) Puerto Rico, (4) states that did not report MOUD status, and (5) states that did not report covariates (Appendix). We did not require states to have data for all years of the analysis. TEDS-A reports episode-level data, representing admissions rather than individuals. We limited our analysis to the first annual episodes of non-Hispanic White and non-Hispanic Black clients for OUD treatment. We required all episodes to have nonmissing data for all covariates and outcomes, except for insurance coverage, given that many states did not report insurance status to SAMHDA. Our final analysis sample was 878,110 episodes that included sociodemographic, treatment, and state residency data (Fig. 1). This research did not involve human subjects and was exempt from IRB review at Boston University Medical Center.

Fig. 1.

Fig. 1.

Open in a new tab

Strobe diagram detailing the identification of the sample population derived from the TEDS-A dataset (2007–2018).

2.2. Measures

We used planned MOUD (as indicated in the treatment plan upon admission) as our dependent variable. We compared changes in MOUD in the 7 years before ACA implementation (2007–2013) and the four years after ACA implementation (2015–2018) for Black and White clients by state Medicaid expansion status. We performed subpopulation analysis on two populations where Black clients are overrepresented, the criminal justice–referred clients and Medicaid enrollees. We used referral source to identify the criminal justice population and health insurance at admission to identify the Medicaid population. We limited the subanalysis of Medicaid-insured episodes to the states that reported health insurance information.

2.3. Covariates

Guided by the Kilbourne disparities model (Kilbourne et al., 2006), we identified covariates that included demographics, treatment-related data, and geography. The demographic variables were age groups (18–20, 21–24, 25–29, 30–34, 35–39, 40–44, 45–49, 50–54, and 55–64); sex (female, male); education status (<8th grade, 9th to 11th grade, 12th grade or GED, 1–3 years of college or vocational school, 4+ years of college); and homelessness (yes, no). The treatment-related variables were client admission setting (inpatient rehabilitation, detoxification, and ambulatory); heroin use; and polysubstance use. Admission setting has been associated with different MOUD preferences and outcomes (Korownyk et al., 2019; Ward et al., 2018). We identified clients living and treated in rural locations using CBSA data.

2.4. Data analysis

We performed descriptive and multivariate logistic regression analyses to understand how planned MOUD changed after the ACA. We also visualized unadjusted trends in the primary outcome of MOUD use separately in expansion and non–expansion states. For our descriptive analyses, we summarized the sociodemographic characteristics and MOUD of clients by race for the pre-post ACA periods, and for the subgroups of clients referred through the criminal justice system and with Medicaid. We graphed the trends for MOUD over time by race and expansion status for the full sample and the two subpopulations. To estimate the change in odds associated with post-ACA period, the study team performed episode-level multivariate logistic regression analyses to examine if the odds of MOUD changed in the expansion states. We included fixed effects for the client U.S. Census division, which yielded a better model fit than including state fixed effects. The research team did not perform a difference-in-differences analysis as our analysis did not find parallel trends between White and Black clients before the ACA. We tried alternative specifications involving matching between White and Black clients, but the trends remained distinctly different. Therefore, the team performed associative analyses with interactions to allow comparisons between racial groups. As a secondary analysis, we examined changes in odds of MOUD in non–expansion states. The variables of interest were racial group interacted with the ACA period. The team performed post-hoc analysis and interacted several theoretically linked covariates (gender and age; rural and census division; homelessness and referral source; admission type and polysubstance use) to avoid misspecification.

We estimated the adjusted odds of MOUD access for Black clients, compared to White clients, before the ACA; and the difference in the odds of MOUD for Black clients post-ACA relative to White clients. The research team performed the analysis on the total sample, the criminal justice–referred sample, and the Medicaid enrolled sample. We set our threshold for significance as the 99th confidence interval due to the large sample size. The team performed sensitivity and specificity testing and all regressions reported an area under the curve >0.70. We conducted all analyses using STATA 14.2 (StataCorp, 2015).

3. Results

3.1. Pre- and post-ACA episode characteristics by expansion status

3.1.1. Primary analyses: expansion states

Tables 1a, 1b, and 1c show treatment episode characteristics in years before and after the ACA by race and subgroup. In expansion states, 59.62% of White clients and 65.80% of Black clients were male. Most White clients before and after the ACA were younger than 35 years, and most of the Black clients were over 40 years of age. More than half of clients in expansion states self-referred to treatment. The proportion of self-referred episodes increased after the ACA in both racial groups. Black clients had a higher proportion of referrals from “other” health care providers (not alcohol/drug use care provider) compared to White clients (11.79% vs. 6.11%). Black clients in expansion states were more likely to report heroin use (88.36%), and Black clients had higher rates of homelessness at 17.26% pre-ACA and 11.91% post-ACA in expansion states. White clients (overall and in subgroups) had higher rates of educational attainment than Black clients before and after the ACA. Most episodes in the sample came from ambulatory care admissions, and the share of ambulatory episodes increased from 53.52% to 70.00% for White clients and from 59.42% to 80.85% for Black clients after the ACA in expansion states. Among criminal justice–referred episodes, ambulatory episodes increased among White clients and Black clients after the ACA (71.28% to 73.22%, 63.59% to 75.00%). Among episodes paid by Medicaid, ambulatory episodes also increased among White clients and Black clients after the ACA (54.10% to 85.13%; 80.23% to 89.17%). Self-referrals increased among Medicaid-insured Black clients after the ACA (37.96% to 70.93%).

Table 1a.

Pre- and post-ACA descriptive statistics among White and Black clients by expansion status (2007–2013 and 2015–2018).

All clients
Expansion states
 Non-expansion states
NH-White Clients
 NH-Black Clients
 NH-White Clients
 NH-Black Clients
Pre-ACA
 Post-ACA
 Pre-ACA
 Post-ACA
 Pre-ACA
 Post-ACA
 Pre-ACA
 Post-ACA
N = 288,657 N = 219,846 N = 52,935 N = 47,796 N = 157,254 N = 94,141 N = 9583 N = 7934
Gender Male 59.62% 57.05% 65.80% 63.93% 51.64% 56.12% 58.06% 63.49%
Age at admission 18–20 yr 10.36% 3.25% 1.30% 1.02% 6.95% 2.85% 2.98% 1.39%
21–24 yr 23.04% 13.49% 3.56% 5.60% 20.19% 13.29% 10.16% 9.63%
25–29 yr 23.64% 25.69% 5.62% 11.29% 27.38% 27.31% 17.77% 22.98%
30–34 yr 14.61% 20.48% 7.34% 9.27% 17.83% 22.78% 16.74% 19.25%
35–39 yr 8.62% 13.34% 12.18% 7.94% 10.13% 14.46% 12.22% 13.45%
40–44 yr 6.67% 7.39% 18.45% 9.04% 6.66% 7.59% 10.78% 10.01%
45–49 yr 5.50% 5.89% 19.80% 15.89% 5.15% 5.06% 9.52% 7.65%
50–54 yr 4.23% 4.86% 17.01% 17.98% 3.58% 3.51% 9.80% 6.58%
55–64 yr 3.34% 5.62% 14.74% 21.97% 2.13% 3.14% 10.03% 9.07%
Referral source Individual (self-referral) 61.57% 65.38% 55.40% 68.98% 56.68% 59.38% 52.18% 60.84%
Alcohol/drug use care provider 6.29% 5.79% 7.70% 5.16% 5.53% 2.83% 6.26% 2.50%
Other health care provider 6.11% 4.27% 11.79% 3.66% 10.84% 12.08% 8.84% 11.75%
School 0.07% 0.03% 0.02% 0.01% 0.07% 0.06% 0.14% 0.10%
Employer 0.56% 0.24% 0.18% 0.10% 0.24% 0.16% 0.27% 0.14%
Other community referral 7.78% 9.24% 7.77% 11.66% 9.82% 6.96% 12.77% 5.26%
Criminal justice referral 17.61% 15.05% 17.15% 10.43% 16.81% 18.55% 19.55% 19.42%
Heroin reported at admission 56.67% 72.95% 88.36% 82.74% 18.99% 45.78% 48.49% 53.14%
Polysubstance use 14.03% 10.68% 12.95% 8.85% 19.71% 19.21% 19.52% 18.25%
Homeless 10.00% 11.31% 17.26% 11.91% 7.58% 10.74% 9.64% 10.26%
Admission setting Inpatient rehab 14.88% 11.87% 12.91% 8.76% 14.43% 13.61% 15.35% 14.42%
Detox 31.60% 18.13% 27.66% 10.38% 36.09% 37.92% 35.40% 33.56%
Ambulatory 53.52% 70.00% 59.42% 80.85% 49.48% 48.46% 49.25% 52.02%
Education Less than high school 4.65% 4.01% 4.80% 4.09% 4.51% 3.86% 4.26% 3.83%
Some high school education 18.06% 16.90% 33.95% 30.20% 21.00% 19.49% 28.59% 25.20%
High school education (or GED) 48.92% 54.40% 43.78% 52.01% 45.58% 52.86% 44.05% 49.84%
Some college 22.65% 17.37% 15.12% 7.68% 25.04% 19.89% 20.44% 18.35%
College graduate + 5.71% 7.32% 2.35% 6.01% 3.87% 3.90% 2.66% 2.79%
Rural 10.60% 32.75% 5.80% 57.35% 16.55% 35.30% 10.94% 26.30%
Census division New England 12.20% 8.98% 3.66% 2.85% 6.80% 4.93% 1.28% 0.71%
Middle Atlantic 21.39% 17.52% 27.55% 18.62% N/A N/A N/A N/A
East North Central 15.59% 5.54% 36.06% 2.46% 0.97% 1.72% 0.77% 1.01%
West North Central 2.63% 1.99% 1.30% 1.47% 7.40% 7.82% 20.10% 22.40%
South Atlantic 1.34% 20.82% 4.85% 54.34% 62.04% 56.55% 45.30% 44.86%
East South Central 7.22% 6.96% 1.10% 1.60% 7.17% 15.18% 6.69% 18.84%
West South Central 3.94% 3.80% 0.95% 0.94% 13.69% 11.17% 25.66% 11.39%
Mountain 35.69% 34.39% 24.53% 17.73% 1.94% 2.63% 0.20% 0.79%
Open in a new tab
Table 1b.

Pre- and post-ACA descriptive statistics among White and Black clients referred through criminal justice system, by expansion status (2007–2013 and 2015–2018).

Criminal justice referral
Expansion
 Non-expansion
NH-White clients
 NH-Black clients
 NH-White clients
 NH-Black clients
Pre-ACA
 Post-ACA
 Pre-ACA
 Post-ACA
 Pre-ACA
 Post-ACA
 Pre-ACA
 Post-ACA
N = 50,840 N = 33,081 N = 9075 N = 4984 N = 26,428 N = 17,458 N = 1873 N = 1541
Gender Male 62.79% 62.33% 77.82% 81.22% 55.29% 56.30% 66.84% 76.44%
Age at admission 18–20 yr 10.00% 3.70% 1.91% 2.85% 9.01% 3.44% 4.97% 2.21%
21–24 yr 23.98% 16.26% 5.32% 12.68% 23.30% 15.01% 13.45% 12.72%
25–29 yr 24.71% 29.52% 7.38% 18.72% 27.18% 28.48% 22.90% 26.93%
30–34 yr 15.03% 21.28% 9.60% 12.74% 16.84% 22.91% 17.14% 20.77%
35–39 yr 8.75% 12.90% 12.67% 9.25% 9.52% 13.86% 11.53% 13.04%
40–44 yr 6.53% 6.47% 18.42% 8.11% 5.98% 7.38% 8.86% 7.98%
45–49 yr 5.29% 4.47% 18.60% 13.06% 3.94% 4.43% 7.31% 6.36%
50–54 yr 3.61% 3.04% 15.02% 11.72% 2.81% 2.80% 6.73% 4.80%
55–64 yr 2.10% 2.35% 11.07% 10.87% 1.42% 1.69% 7.10% 5.19%
Heroin use 57.34% 74.05% 85.76% 67.36% 16.39% 40.72% 45.01% 44.78%
Polysubstance use 18.92% 15.65% 15.64% 15.63% 22.10% 19.97% 19.75% 20.70%
Homeless 8.23% 11.21% 13.79% 10.71% 4.03% 6.43% 3.58% 5.91%
Admission setting Inpatient rehab 22.93% 21.51% 26.63% 21.67% 19.99% 16.87% 16.34% 16.03%
Detox 5.78% 5.27% 9.77% 3.33% 15.68% 15.15% 11.21% 12.07%
Ambulatory 71.28% 73.22% 63.59% 75.00% 64.33% 67.97% 72.45% 71.90%
Education Less than high school 4.32% 3.90% 4.41% 3.31% 4.68% 3.42% 4.27% 2.53%
Some high school education 24.12% 19.99% 38.02% 32.34% 24.30% 22.55% 35.88% 31.47%
High School education (or GED) 48.18% 54.58% 43.74% 51.81% 45.08% 51.50% 40.20% 46.98%
Some college 19.61% 17.01% 12.14% 9.59% 22.82% 19.34% 17.73% 17.26%
College graduate + 3.76% 4.52% 1.70% 2.95% 3.12% 3.19% 1.92% 1.75%
Rural 15.12% 30.27% 7.66% 38.38% 21.16% 40.57% 12.92% 28.62%
Census division New England 4.31% 4.49% 2.55% 3.61% 4.16% 3.07% 0.85% 0.58%
Middle Atlantic 19.72% 25.19% 33.08% 40.79% N/A N/A N/A N/A
East North Central 22.85% 11.50% 25.77% 6.54% 1.63% 3.82% 0.64% 1.56%
West North Central 1.07% 1.22% 0.55% 2.03% 9.84% 8.00% 29.90% 25.70%
South Atlantic 0.90% 11.21% 5.18% 28.59% 59.56% 45.30% 41.11% 33.48%
East South Central 11.72% 13.03% 0.78% 2.25% 14.18% 27.34% 14.15% 30.24%
West South Central 4.50% 5.57% 0.95% 1.79% 6.89% 7.36% 12.97% 6.75%
Mountain 34.93% 27.80% 31.14% 14.41% 3.74% 5.11% 0.37% 1.69%
Open in a new tab
Table 1c.

Pre- and post-ACA descriptive statistics among White and Black clients with Medicaid, by expansion status (2007–2013 and 2015–2018).

Medicaid Insured Clients
Expansion states
 Non-Expansion states
NH-White clients
 NH-Black clients
 NH-White clients
 NH-Black clients
Pre-ACA
 Post-ACA
 Pre-ACA
 Post-ACA
 Pre-ACA
 Post-ACA
 Pre-ACA
 Post-ACA
N = 18,946
 N = 66,049
 N = 10,555
 N = 28,386
 N = 12,235
 N = 8346
 N = 1033
 N = 1014
Gender Male 43.59% 49.92% 54.45% 62.35% 31.72% 28.66% 30.59% 31.76%
Age at admission 18–20 yr 8.15% 2.15% 0.58% 0.76% 8.44% 4.29% 4.07% 1.48%
21–24 yr 21.10% 10.23% 1.53% 4.25% 20.02% 11.61% 13.07% 8.68%
25–29 yr 26.27% 24.65% 2.94% 9.56% 27.79% 24.53% 22.46% 22.68%
30–34 yr 17.00% 22.24% 4.28% 7.96% 17.81% 22.54% 17.72% 22.39%
35–39 yr 9.51% 14.92% 11.62% 6.13% 9.65% 15.17% 11.42% 14.50%
40–44 yr 7.06% 8.53% 20.67% 8.86% 6.03% 8.39% 7.55% 9.57%
45–49 yr 4.98% 6.81% 23.10% 18.46% 4.91% 4.88% 5.03% 5.33%
50–54 yr 3.69% 5.46% 19.26% 20.92% 3.08% 4.07% 7.65% 5.33%
55–64 yr 2.22% 5.00% 16.01% 23.11% 2.26% 4.53% 11.04% 10.06%
Referral source Individual (self-referral) 67.93% 66.74% 37.96% 70.93% 59.10% 57.90% 60.41% 65.38%
Alcohol/drug use care provider 5.86% 3.95% 5.89% 4.21% 6.02% 4.49% 6.10% 2.37%
Other health care provider 7.64% 3.83% 43.69% 3.47% 14.16% 11.54% 8.03% 12.23%
School 0.08% 0.02% 0.01% 0.01% 0.07% 0.10% 0.10% 0.10%
Employer 0.03% 0.06% 0.01% 0.04% 0.07% 0.18% 0.10% 0.10%
Other community referral 9.71% 14.25% 7.65% 14.68% 9.89% 9.06% 12.88% 7.79%
Criminal justice referral 8.75% 11.16% 4.79% 6.67% 10.68% 16.74% 12.39% 12.03%
Heroin reported at admission 61.21% 69.98% 95.12% 85.51% 22.05% 39.95% 56.63% 54.73%
Polysubstance use 14.41% 10.01% 12.00% 8.17% 18.81% 20.61% 17.91% 16.77%
Homeless 7.61% 6.04% 6.58% 7.66% 4.48% 7.06% 7.26% 7.50%
Admission setting Inpatient rehab 7.84% 5.93% 7.66% 5.50% 13.67% 19.08% 20.52% 18.34%
Detox 38.06% 8.94% 12.12% 5.33% 16.18% 10.26% 20.23% 19.23%
Ambulatory 54.10% 85.13% 80.23% 89.17% 70.15% 70.67% 59.24% 62.43%
Education Less than high school 15.68% 5.14% 6.69% 4.15% 6.13% 4.80% 5.13% 4.44%
Some high school education 12.89% 19.70% 37.60% 33.03% 28.57% 22.93% 34.66% 28.80%
High school education (or GED) 48.24% 59.88% 39.93% 53.21% 45.50% 47.99% 40.56% 42.31%
Some college 19.63% 6.31% 13.91% 2.18% 17.33% 21.29% 18.01% 21.60%
College graduate + 3.55% 8.97% 1.87% 7.43% 2.47% 2.98% 1.65% 2.86%
Rural 10.16% 68.18% 3.28% 81.66% 23.23% 47.20% 10.07% 32.35%
Census division New England 49.23% 8.35% 4.11% 1.55% 49.07% 22.80% 6.58% 2.17%
Middle Atlantic 18.27% 10.66% 19.45% 12.39% N/A N/A N/A N/A
East North Central 18.95% 0.19% 66.73% 0.44% 1.84% 5.34% 0.97% 2.27%
West North Central 0.14% 0.02% 0.02% 0.00% 21.53% 22.60% 55.47% 55.13%
South Atlantic 4.33% 66.12% 8.92% 84.17% 9.78% 0.08% 9.58% 0.00%
East South Central 4.19% 11.26% 0.27% 0.98% 5.66% 26.66% 5.81% 23.57%
West South Central 3.16% 3.04% 0.43% 0.44% 10.13% 16.75% 21.49% 15.78%
Mountain 1.73% 0.36% 0.08% 0.02% 1.99% 5.76% 0.10% 1.08%
Open in a new tab

3.1.2. Secondary analyses: non–expansion states

In non–expansion states, 51.64% of White clients and 58.06% of Black clients were male (Tables 1a–1c). Like the expansion states, most of the White clients before and after 2014 were younger than 35 years, and most of the Black clients were over 40 years of age. More than half of clients in non–expansion states also self-referred to treatment. White clients had a lower proportion of criminal justice–referred episodes compared to Black clients (16.81% vs. 19.55%) before 2014, but the proportion of criminal justice–referred episodes increased after 2014 among White clients (18.55%). About one-fifth of White and Black clients reported using more than one drug (19.71% vs. 19.52%) and Black clients were more likely to report heroin use (18.99% vs. 48.49%). The rate of homelessness was also higher among the Black client population at 9.64% before 2014 and 10.26% after 2014 in non–expansion states. The proportion of ambulatory episodes did not change in non–expansion states. The criminal justice–referred ambulatory visits increased among White clients (64.33% to 67.97%) but decreased for Black clients after 2014 (72.45% to 71.90%) in non–expansion states. Most of the visits paid by Medicaid after 2014 for White and Black clients were ambulatory visits; the study did not observe a change in the number of episodes paid by Medicaid for White clients (70.15% to 70.67%), but we did observe a modest increase in the number of visits paid for among Black clients (59.24% to 62.43%).

3.2. Trends and changes in MOUD utilization by expansion status

3.2.1. Primary analyses: expansion states

Table 2 and Fig. 2 report the unadjusted changes in the proportion of clients with planned MOUD over time. In expansion states, the proportion of White clients with MOUD significantly increased from an average of 33.57% in the pre-ACA period to an average of 51.34% in the post-ACA period. The proportion of Black clients with MOUD significantly increased from an average of 36.16% in the pre-ACA period to an average of 61.65% in the post-ACA period. For criminal justice referrals, the proportion of White clients with MOUD significantly increased from 4.53% to 12.24%, and the proportion of Black clients with MOUD also increased from 6.94% to 12.12% (Fig. 3). MOUD among the Medicaid population in expansion state increased from 32.83% to 59.84% for White clients and from 27.04% to 64.87% for Black clients (Fig. 4).

Table 2.

Differences in MOUD use before (2008–2013) and after (2015–2018) ACA implementation by subgroup and race.

All Clients
Expansion
 Non-expansion
NH-White clients
 NH-Black clients
 NH-White clients
 NH-Black Clients
Pre-ACA Post-ACA Pre-ACA Post-ACA Pre-ACA Post-ACA Pre-ACA Post-ACA
Total episodes 288,657 219,846 52,926 47,793 157,236 94,135 9583 7934
Total MOUD episodes 96,903 112,876 19,135 29,462 22,676 9950 1135 868
% with MOUD 33.57% 51.34%*** 36.16% 61.65%*** 14.42% 10.57%*** 11.84% 10.94%
Criminal justice referral
NH-White clients
 NH-Black clients
 NH-White clients
 NH-Black clients
Pre-ACA Post-ACA Pre-ACA Post-ACA Pre-ACA Post-ACA Pre-ACA Post-ACA
Total episodes 50,840 33,081 9075 4984 26,428 17,458 1873 1541
Total MOUD episodes 2636 3506 630 604 2636 3506 16 14
% with MOUD 4.53% 12.24%*** 6.94% 12.12%*** 2.70% 2.70% 0.85% 0.91%
Medicaid
NH-White clients
 NH-Black clients
 NH-White clients
 NH-Black clients
Pre-ACA Post-ACA Pre-ACA Post-ACA Pre-ACA Post-ACA Pre-ACA Post-ACA
Total episodes 18,946 66,049 10,555 28,386 12,235 8346 1033 1014
Total MOUD episodes 6220 39,524 2854 18,415 4403 2318 270 257
% with MOUD 32.83% 59.84%*** 27.04% 64.87%*** 35.99% 27.77%*** 26.14% 25.35%
Open in a new tab

Notes.

***

Indicates significant differences at p < 0.001.

Fig. 2.

Fig. 2.

Open in a new tab

Unadjusted Proportion of Black and White clients with MOUD use in expansion and non-expansion states (2007–2013, 2015–2018).

Fig. 3.

Fig. 3.

Open in a new tab

Unadjusted proportion of Black and White clients referred through the criminal justice system with MOUD use in expansion and non-expansion states (2007–2013, 2015–2018).

Fig. 4.

Fig. 4.

Open in a new tab

Unadjusted Proportion of Black and White clients with Medicaid who had MOUD use in expansion and non-expansion states (2007–2013, 2015–2018).

3.2.2. Secondary analyses: non–expansion states

In non–expansion states, the proportion of White clients with MOUD increased in 2010, steadily decreased until 2017, and then increased again in 2018 (Fig. 2). The proportion of Black clients with MOUD followed a decreasing trend until it went up in 2018. On average, the proportion of White clients with MOUD decreased (14.42% to 10.57%), and MOUD for Black clients did not differ from before 2014 (11.84% to 10.94%; Table 2). On average, less than 3% of criminal justice–referred episodes for White clients and less than 1% of episodes for Black clients in non–expansion states had MOUD (Fig. 3). MOUD among the Medicaid population decreased from 35.99% to 27.77% for White clients and from 26.14% to 25.35% for Black clients (Fig. 4).

3.3. Multivariate results

Table 3 presents the odds of MOUD after the ACA by subgroup in the expansion states, and for the same period in the non–expansion states, controlling for all covariates. Pre-ACA, Black clients in expansion states were less likely than White clients to receive MOUD (adjusted odds ratio (aOR) = 0.88, 99th Confidence Interval (CI) = [0.85,0.91]). Post- ACA, no difference existed in the increase in MOUD by racial group in the expansion states, meaning that the odds of MOUD increased by a similar amount for Black and White clients. Among criminal justice–referred clients, Black clients were significantly more likely to receive MOUD than White clients before the ACA (aOR = 1.21, CI = [1.06, 1.38]); and after the ACA, Black clients had a smaller increase in odds of MOUD compared to White clients (aOR = 0.74, CI = [0.62, 0.89]). Among Medicaid covered clients, Black clients were significantly less likely to receive MOUD than White clients pre-ACA (aOR = 0.56, CI = [0.50, 0.63]), and Black clients had a larger increase in odds of MOUD compared to White clients with Medicaid coverage (aOR = 1.16, CI = [1.03, 1.30]).

Table 3.

Difference in odds of MOUD use between White and Black clients in expansion vs. non-expansion states by subgroup.

Reference population: White clients, pre-ACA All
 Criminal justice referral
 Medicaid
Expansion states
 Non-expansion states
 Expansion states
 Non-expan states
 Expansion states
 Non-expansion states
aOR 99% CI aOR 99% CI aOR 99% CI aOR 99% CI aOR 99% CI aOR 99% CI
Black clients, pre-ACA 0.88*** 0.85, 0.91 0.86*** 0.77, 0.95 1 21*** 1.06, 1.38 0.38*** 0.19, 0.74 0.56*** 0.50, 0.63 1 79*** 1.38, 2.33
White clients, post-ACA 1.33*** 1.29, 1.39 1.07 0.98, 1.16 1.89*** 1.64, 2.17 0.93 0.65, 1.33 1.13** 1.01, 1.27 0.91 0.72, 1.15
Black clients, post-ACA 1.16*** 1.10, 1.22 1.13 0.99, 1.30 1.69*** 1.41, 2.04 0.43*** 0.20, 0.93 0 73*** 0.64, 0.83 1.51** 1.09, 2.10
Difference in odds ratios (Black#post-ACA) 0.98 0.94, 1.03 1 24*** 1.07, 1.44 0.74*** 0.62, 0.89 1.21 0.46, 3.19 1.16*** 1.03, 1.30 0.93 0.72, 1.15
Open in a new tab

Notes: We adjusted regressions for census region-fixed effects and episode level characteristics like gender, age group, homelessness, educational attainment, service setting, polysubstance use, heroin use, and referral source.

***

Indicates significant change at p < 0.001.

**

Indicates significant change at p < 0.01.

In non–expansion states before 2014, Black clients were less likely to receive MOUD (aOR = 0.86, CI = [0.77, 0.95]). After 2014, MOUD among Black clients rose by 1.24 times the increase among White clients (aOR = 1.24, CI = [1.07, 1.44]). Before 2014, Black clients who were referred through the criminal justice system were significantly less likely to receive MOUD than White clients (aOR = 0.38, CI = [0.19, 0.74]), and after 2014, no significant changes occurred in the odds of MOUD among White or Black clients. Black clients with Medicaid were significantly more likely to receive MOUD than White clients before 2014 (aOR = 1.79, CI = [1.38, 2.33]), and no significant changes occurred in the odds of MOUD among White or Black clients with Medicaid in non–expansion states after 2014.

4. Discussion

This study builds on prior work investigating the association between the ACA and MOUD at federally subsidized treatment centers. We found that in the general sample of client treatment episodes, the ACA in expansion states was associated with significantly increased odds of MOUD for both Black and White clients relative to White clients pre- ACA, and no difference occurred in the level of increase among the groups. Clients in expansion states referred through the criminal justice system had remarkably lower odds of MOUD compared to clients referred through other sources. Among clients with Medicaid, the ACA was associated with increased odds of MOUD for both White and Black clients in expansion states, with a greater relative increase among Black clients.

4.1. Increase in utilization of MOUD over time

We found a steady increase in the proportion of first-time episodes with MOUD over time in the Medicaid expansion states. This finding is promising given the association of MOUD with reduced opioid overdose and mortality (Volkow et al., 2014b; Wakeman et al., 2020). Rising MOUD utilization could reflect awareness of the rising number of opioid-related overdose and deaths that occurred during the study period (Jones et al., 2020), although larger cohort studies have not shown large increases in OUD treatment postoverdose (Frazier et al., 2017; Larochelle et al., 2018). Given the significant increases observed in expansion states, this outcome likely reflects the success of federal insurance reforms, changes to prescribing limitations, and health promotion programs aimed to increase the adoption of MOUD (Clemans-Cope et al., 2017; Kozhimannil et al., 2019; McKenna, 2017; Wen et al., 2017). The context and quality of MOUD treatment tempers these encouraging results.

4.2. Disparities in access to MOUD

While it is encouraging that the ACA was associated with increased access to MOUD for Black clients, readers should interpret these findings within the context of accessing specific MOUDs. The TEDS-A data do not distinguish among buprenorphine, methadone, and naltrexone, so we could not determine the differential use of each medication. Although the number of clients utilizing buprenorphine at federal treatment facilities increased dramatically over the study period, in 2015, roughly 50,000 clients used buprenorphine versus nearly 330,000 clients utilizing methadone (Alderks, 2013). Thus, many clients seeking treatment at federally subsidized treatment centers are accessing methadone. Methadone and buprenorphine are two of the most effective treatments for OUD (Wakeman et al., 2020), yet clinicians may prescribe one medication more than the other. Studies have demonstrated that methadone is more effective for people who inject drugs and who use heroin, and is associated with higher retention in care compared to buprenorphine (Connock et al., 2007; Hser et al., 2014). However, research does not recommend that methadone be used with patients at risk of methadone toxicity and overdose due to heavy alcohol use or other sedating medications (Lee et al., 2014). Studies have shown that buprenorphine has a lower risk of overdose, especially during the initial weeks of treatment and dose titration (Bell et al., 2009). Patients may also have a preference for one of the medications due to their prior experience with side effects and provider treatment, or due to practical concerns (Ridge et al., 2009; Uebelacker et al., 2016). For example, patients may prefer buprenorphine to avoid daily visits to a methadone clinic, which can cause work conflicts and the burden of travel time and cost.

Affordability and access are critical factors for initiating and continuing MOUD treatment. The expansion of Medicaid was associated with a 70% increase in buprenorphine prescriptions covered by Medicaid and by 2018, Medicaid and commercial plans covered buprenorphine in all 50 states (Meinhofer & Witman, 2018; Reif et al., 2017; Wen et al., 2017). In contrast, as of 2016, Medicaid did not cover methadone in at least 17 states, and many commercial plans required prior authorization and costly copays, limiting access for low-income clients (Polsky et al., 2020; Saloner, Stoller, & Barry, 2016). Evidence indicates that racial disparities exist in access to buprenorphine (Goedel et al., 2020; Hansen et al., 2016; Lagisetty et al., 2019). To what extent those disparities may have affected the findings from this study is unclear. Given the factors involved in deciding which medication may be most effective for each patient, providers should understand whether ongoing systemic barriers to any one of the MOUDs related to insurance, income, or geographical area may limit access. We suggest researchers consider how these barriers contribute to disparities in OUD treatment outcomes.

4.3. Utilization of MOUD among clients referred through the criminal justice system

Prior literature has shown that people involved with the criminal justice system have the least access to MOUD (Krawczyk, Picher, et al., 2017; Winkelman et al., 2020; Yang et al., 2019). This study found that among criminal justice–referred clients, less than 12% in the expansion states and less than 3% in the non–expansion states had access to MOUD. Two 2016 studies found that the initial ACA insurance reforms improved the odds of health insurance coverage among clients involved with the criminal justice system (Saloner, Bandara, et al., 2016; Winkelman et al., 2016). Although Saloner and colleagues did not find a corresponding increase in treatment utilization in 2014, both studies found that people involved with the criminal justice system and utilizing SUD treatment were more likely to have Medicaid compared to other insurance or no insurance. A more recent study focused on MOUD access among clients referred through the criminal justice system also found that their utilization of MOUD has increased but has remained low relative to the general population (Khatri et al., 2021). The literature discusses less the ongoing disparity in MOUD between White and Black clients involved with the criminal justice system. This study found that Black clients were 60% less likely to have MOUD than White clients referred through the criminal justice system in non–expansion states. Readers should interpret this finding carefully, as only 30 out of 3414 Black clients had MOUD. Nevertheless, this study is one of the first that demonstrates the extent of the disparity that exists between these two marginalized groups. Thus, future work should seek to understand why MOUD utilization remains low among clients involved with the criminal justice system, especially for Black clients.

4.4. Changes in utilization of MOUD among the Medicaid-insured population

The greater rise in MOUD for Medicaid clients in expansion states is consistent with other studies that have evaluated recent policy changes on access to substance use treatment, and specifically for the treatment for OUD. Geissler and Evans found that the implementation of the Mental Health Parity and Addiction Equity Act (MHPAEA) was associated with a 4.6 percentage point increase in the probability of an SUD treatment facility accepting Medicaid before the ACA (Geissler & Evans, 2020). Private and Medicaid coverage and payments increased after the ACA for clients receiving SUD treatment, and admissions with medications used to treat OUD more than doubled after Medicaid expansion (Maclean & Saloner, 2019; Meinhofer & Witman, 2018). Mojtabai and colleagues also found larger increases in the use of MOUD in expansion states compared to non–expansion states in their comparison of treatment in the first couple of years post-ACA compared to pre-ACA (Mojtabai et al., 2019). Our findings confirm and extend their results.

4.5. Limitations

Readers should consider several limitations when interpreting the results of this study. The parallel trend assumption did not hold for a difference-in-differences study design; Black and White clients in the non–expansion states had different MOUD trends than the Black and White clients in the expansion states and, therefore, could not function as comparison groups. We included first-episode data only, which allowed us to study individuals but did not allow us to determine any change in MOUD through the course of treatment, or the length of time that opioid use had been a problem upon admission; although we did control for the type of substance used. TEDS-A only includes data from federally subsidized treatment programs; it does not include waivered primary care providers. TEDS-A includes an administrative report of race and OUD, which may be inconsistent between states. TEDS-A has limited insurance data and much of that data are missing; thus, readers should interpret the Medicaid subgroup analyses with some caution. We only included episodes that indicated opioids as a primary concern. Our analysis may be missing patients whose primary substance at admission was not opioids.

4.6. Conclusion

Federal policies designed to increase access to substance use treatment increased MOUD and reduced disparities for Black clients who used opioids. In the years after ACA implementation, the proportion of Black clients that utilized MOUD increased by twenty-five percentage points, such that nearly 60% of Black clients who reported opioids as their primary substance of concern had access to MOUD in expansion states. The greatest reduction in an MOUD treatment disparity between Black and White clients occurred for clients with Medicaid in expansion states. Nonetheless, racial disparities in OUD treatment and underutilization of MOUD persist. In non–expansion states, 11% of Black clients received MOUD. Fewer than 1 in 6 Black clients referred by the criminal justice system in the United States had access to MOUD between 2007 and 2018. Policy-makers and health care leaders seeking to improve access to OUD treatment across the country should focus their work on marginalized populations that disproportionately lack access to MOUD.

Supplementary Material

mmc1

Acknowledgements

The authors thank Dr. Ziming Zuan for his comments on an earlier draft of the manuscript. This research was supported in part by a grant from the National Institute on Drug Abuse [T32DA041898].

Footnotes

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jsat.2021.108533.

References

  1. Alderks CE (2013). Trends in the use of methadone, buprenorphine, and extended-release naltrexone at substance Abuse treatment facilities: 2003–2015 (update). In The CBHSQ Report. [PubMed] [Google Scholar]
  2. Anderson KO, Green CR, & Payne R (2009). Racial and ethnic disparities in pain: Causes and consequences of unequal care. Journal of Pain. 10.1016/j.jpain.2009.10.002. [DOI] [PubMed] [Google Scholar]
  3. Andrews CM, Pollack HA, Abraham AJ, Grogan CM, Bersamira CS, D’Aunno T, & Friedmann PD (2019). Medicaid coverage in substance use disorder treatment after the affordable care act. Journal of Substance Abuse Treatment.. 10.1016/j.jsat.2019.04.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bagley SM, Larochelle MR, Xuan Z, Wang N, Patel A, Bernson D, … Walley AY (2020). Characteristics and receipt of medication treatment among young adults who experience a nonfatal opioid-related overdose. Annals of Emergency Medicine.. 10.1016/j.annemergmed.2019.07.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bell JR, Butler B, Lawrance A, Batey R, & Salmelainen P (2009). Comparing overdose mortality associated with methadone and buprenorphine treatment. Drug and Alcohol Dependence, 104(1–2), 73–77. 10.1016/j.drugalcdep.2009.03.020. [DOI] [PubMed] [Google Scholar]
  6. Center for Behavioral Health Statistics. (2018). Results from the 2017 National Survey on Drug Use And Health: Detailed tables. https://www.samhsa.gov/data/sites/default/files/cbhsq-reports/NSDUHDetailedTabs2017/NSDUHDetailedTabs2017.pdf.
  7. Clemans-Cope L, Epstein M, & Kenney GM (2017). Rapid growth in Medicaid spending on medications to treat opioid use disorder and overdose. Urban Institute. [Google Scholar]
  8. Cohen RA, & Martinez ME (2015). Health insurance coverage: Early release of estimates from the National Health Interview Survey, January-March 2015. Center for Disease Control and Prevention, 2015(March), 1–11. http://www.cdc.gov/nchs/data/nhis/earlyrelease/insur201508.pdf. [Google Scholar]
  9. Connock M, Juarez-Garcia A, Jowett S, Frew E, Liu Z, Taylor RJ, … Taylor RS (2007). Methadone and buprenorphine for the management of opioid dependence: A systematic review and economic evaluation. In, Vol. 11, Issue 9. Health Technology Assessment. National Co-Ordinating Centre for HTA. 10.3310/hta11090. [DOI] [PubMed] [Google Scholar]
  10. Creedon TB, & Lê Cook B (2016). Datawatch: Access to mental health care increased but not for substance use, while disparities remain. Health Affairs, 35(6), 1017–1021. 10.1377/hlthaff.2016.0098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Frazier W, Cochran G, Lo-Ciganic WH, Gellad WF, Gordon AJ, Chang CCH, & Donohue JM (2017). Medication-assisted treatment and opioid use before and after overdose in Pennsylvania Medicaid. JAMA - Journal of the American Medical Association.. 10.1001/jama.2017.7818. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Furr-Holden D, Milam AJ, Wang L, & Sadler R (2020). African Americans now outpace whites in opioid-involved overdose deaths: A comparison of temporal trends from 1999 to 2018. Addiction. 10.1111/add.15233. [DOI] [PubMed] [Google Scholar]
  13. Geissler KH, & Evans EA (2020). Changes in Medicaid acceptance by substance abuse treatment facilities after implementation of federal parity. Medical Care.. 10.1097/MLR.0000000000001242. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Goedel WC, Shapiro A, Cerdá M, Tsai JW, Hadland SE, & Marshall BDL ´ (2020). Association of racial/ethnic segregation with treatment capacity for opioid use disorder in counties in the United States. JAMA Network Open, 3(4), 203711. 10.1001/jamanetworkopen.2020.3711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hadland SE, Frank Wharam JW, Schuster MA, Zhang F, Samet JH, & Larochelle MR (2017). Trends in receipt of buprenorphine and naltrexone for opioid use disorder among adolescents and young adults, 2001–2014. JAMA Pediatrics. 10.1001/jamapediatrics.2017.0745. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hansen H, Siegel C, Wanderling J, & DiRocco D (2016). Buprenorphine and methadone treatment for opioid dependence by income, ethnicity and race of neighborhoods in New York City. Drug and Alcohol Dependence.. 10.1016/j.drugalcdep.2016.03.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hser Y-I, Saxon AJ, Huang D, Hasson A, Thomas C, Hillhouse M, … Ling W (2014). Treatment retention among patients randomized to buprenorphine/naloxone compared to methadone in a multi-site trial. Addiction, 109(1), 79–87. 10.1111/add.12333. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Jones JD, Campbell AN, Brandt L, Castillo F, Abbott R, & Comer SD (2020). Intervention in an opioid overdose event increases interest in treatment among individuals with opioid use disorder. Substance Abuse. 10.1080/08897077.2020.1809607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Khatri UG, Howell BA, & Winkelman TNA (2021). Medicaid expansion increased medications for opioid use disorder among adults referred by criminal justice agencies. Health Affairs, 40(4), 562–570. 10.1377/hlthaff.2020.01251. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kiechle JE, & Gonzalez CM (2020). The opioid crisis and the black/African American population: An urgent issue (Samhsa; ). [Google Scholar]
  21. Kilaru AS, Xiong A, Lowenstein M, Meisel ZF, Perrone J, Khatri U, … Delgado MK (2020). Incidence of treatment for opioid use disorder following nonfatal overdose in commercially insured patients...The Leonard Davis Institute of Health Economics at the University of Pennsylvania. JAMA Network Open, 3(5), Article e205852. 10.1001/jamanetworkopen.2020.5852. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kilbourne AM, Switzer G, Hyman K, Crowley-Matoka M, & Fine MJ (2006). Advancing health disparities research within the health care system: A conceptual framework. American Journal of Public Health, 96(12), 2113–2121. 10.2105/AJPH.2005.077628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Korownyk C, Perry D, Ton J, Kolber MR, Garrison S, Thomas B, … Lindblad AJ (2019). Opioid use disorder in primary care: PEER umbrella systematic review of systematic reviews (Canadian Family Physician Medecin de Famille Canadien; ). [PMC free article] [PubMed] [Google Scholar]
  24. Kozhimannil KB, Dowd WN, Ali MM, Novak P, & Chen J (2019). Substance use disorder treatment admissions and state-level prenatal substance use policies: Evidence from a national treatment database. Addictive Behaviors.. 10.1016/j.addbeh.2018.11.019. [DOI] [PubMed] [Google Scholar]
  25. Krawczyk N, Feder KA, Fingerhood MI, & Saloner B (2017). Racial and ethnic differences in opioid agonist treatment for opioid use disorder in a U.S. national sample. Drug and Alcohol Dependence. 10.1016/j.drugalcdep.2017.06.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Krawczyk N, Picher CE, Feder KA, & Saloner B (2017). Only one in twenty justice-referred adults in specialty treatment for opioid use receive methadone or buprenorphine. Health Affairs. 10.1377/hlthaff.2017.0890. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lagisetty R, Bohnert A, Clay M, Maust D, & R. PA (2019). Buprenorphine and the treatment divide between whites and blacks. Journal of General Internal Medicine. 10.1007/11606.1525-1497. [DOI] [Google Scholar]
  28. Larochelle MR, Bernson D, Land T, Stopka TJ, Wang N, Xuan Z, … Walley AY (2018). Medication for opioid use disorder after nonfatal opioid overdose and association with mortality: A cohort study. Annals of Internal Medicine. 10.7326/M17-3107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lee SC, Klein-Schwartz W, Doyon S, & Welsh C (2014). Comparison of toxicity associated with nonmedical use of benzodiazepines with buprenorphine or methadone. Drug and Alcohol Dependence, 138(1), 118–123. 10.1016/j.drugalcdep.2014.02.014. [DOI] [PubMed] [Google Scholar]
  30. Maclean JC, & Saloner B (2019). The effect of public insurance expansions on substance use disorder treatment: Evidence from the affordable care act. Journal of Policy Analysis and Management. 10.1002/pam.22112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Madras BK (2017). The surge of opioid use, addiction, and overdoses responsibility and response of the US health care. JAMA Psychiatry. 10.1001/jamapsychiatry.2017.0163. [DOI] [PubMed] [Google Scholar]
  32. Marsh JC, Cao D, Guerrero E, & Shin HC (2009). Need-service matching in substance abuse treatment: Racial/ethnic differences. Evaluation and Program Planning. 10.1016/j.evalprogplan.2008.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. McKenna RM (2017). Treatment use, sources of payment, and financial barriers to treatment among individuals with opioid use disorder following the national implementation of the ACA. Drug and Alcohol Dependence.. 10.1016/j.drugalcdep.2017.06.028. [DOI] [PubMed] [Google Scholar]
  34. Meinhofer A, & Witman AE (2018). The role of health insurance on treatment for opioid use disorders: Evidence from the affordable care act Medicaid expansion. Journal of Health Economics.. 10.1016/j.jhealeco.2018.06.004. [DOI] [PubMed] [Google Scholar]
  35. Mennis J, & Stahler GJ (2016). Racial and ethnic disparities in outpatient substance use disorder treatment episode completion for different substances. Journal of Substance Abuse Treatment.. 10.1016/j.jsat.2015.12.007. [DOI] [PubMed] [Google Scholar]
  36. Mojtabai R, Mauro C, Wall MM, Barry CL, & Olfson M (2019). The affordable care act and opioid agonist therapy for opioid use disorder. Psychiatric Services 10.1176/appi.ps.201900025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. National Institute on Drug Abuse. (2020). Overdose death rates. 2018. National Institutes of Health. https://www.drugabuse.gov/drug-topics/trends-statistics/overdose-death-rates. [Google Scholar]
  38. Paulozzi LJ, Budnitz DS, & Xi Y (2006). Increasing deaths from opioid analgesics in the United States. Pharmacoepidemiology and Drug Safety. 10.1002/pds.1276. [DOI] [PubMed] [Google Scholar]
  39. Polsky D, Arsenault S, & Azocar F (2020). Private coverage of methadone in outpatient treatment programs. Psychiatric Services. 10.1176/appi.ps.201900373. [DOI] [PubMed] [Google Scholar]
  40. Reif S, Creedon TB, Horgan CM, Stewart MT, & Garnick DW (2017). Commercial health plan coverage of selected treatments for opioid use disorders from 2003 to 2014. Journal of Psychoactive Drugs 10.1080/02791072.2017.1300360. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Ridge G, Gossop M, Lintzeris N, Witton J, & Strang J (2009). Factors associated with the prescribing of buprenorphine or methadone for treatment of opiate dependence. Journal of Substance Abuse Treatment 10.1016/j.jsat.2008.09.007. [DOI] [PubMed] [Google Scholar]
  42. Saloner B, Bandara SN, McGinty EE, & Barry CL (2016). Justice-involved adults with substance use disorders: Coverage increased but rates of treatment did not in 2014. Health Affairs. 10.1377/hlthaff.2016.0005. [DOI] [PubMed] [Google Scholar]
  43. Saloner B, & Cook BL (2013). Blacks and hispanics are less likely than whites to complete addiction treatment, largely due to socioeconomic factors. Health Affairs. 10.1377/hlthaff.2011.0983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Saloner B, Stoller KB, & Barry CL (2016). Medicaid coverage for methadone maintenance and use of opioid agonist therapy in specialty addiction treatment. Psychiatric Services. 10.1176/appi.ps.201500228. [DOI] [PubMed] [Google Scholar]
  45. SAMHSA. (2016). About population data - National Survey on Drug Use and Health. https://www.samhsa.gov/data/population-data-nsduh/about.
  46. SAMHSA. (2019). Treatment episode data set (TEDS-A-2000–2017). https://www.datafiles.samhsa.gov/study/treatment-episode-data-set-admissions-teds-2000-2017-nid18475.
  47. SAMHSA. (2020). Treatment episode data set (TEDS-A-2018). https://www.datafiles.samhsa.gov/study-dataset/treatment-episode-data-set-admissions-2018-teds-2018-ds0001-nid19020.
  48. Singhal A, Tien YY, & Hsia RY (2016). Racial-ethnic disparities in opioid prescriptions at emergency department visits for conditions commonly associated with prescription drug abuse. PLoS One. 10.1371/journal.pone.0159224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. StataCorp. (2015). Stata Statistical Software: Release 14. College Station, TX: StataCorp LP. [Google Scholar]
  50. Substance Abuse and Mental Health Data Archive. (2020). Treatment Episode Data Set Admissions (TEDS-A). https://www.datafiles.samhsa.gov/study-series/treatment-episode-data-set-admissions-teds-nid13518.
  51. The Henry J. Kaiser Family Foundation. (2016). Opioid overdose deaths by gender. In | The Henry J. Kaiser Family Foundation; (Issue Cdc). https://www.kff.org/other/state-indicator/opioid-overdose-deaths-by-raceethnicity/. [Google Scholar]
  52. Uebelacker LA, Bailey G, Herman D, Anderson B, & Stein M (2016). Patients’ beliefs about medications are associated with stated preference for methadone, buprenorphine, naltrexone, or no medication-assisted therapy following inpatient opioid detoxification. Journal of Substance Abuse Treatment 10.1016/j.jsat.2016.02.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Volkow ND, Frieden TR, Hyde PS, & Cha SS (2014a). Medication-assisted therapies — Tackling the opioid-overdose epidemic. New England Journal of Medicine, 370(22), 2063–2066. 10.1056/NEJMp1402780. [DOI] [PubMed] [Google Scholar]
  54. Volkow ND, Frieden TR, Hyde PS, & Cha SS (2014b). Medication-assisted therapies — Tackling the opioid-overdose epidemic. New England Journal of Medicine 10.1056/nejmp1402780. [DOI] [PubMed] [Google Scholar]
  55. Wakeman SE, Larochelle MR, Ameli O, Chaisson CE, McPheeters JT, Crown WH, … Sanghavi DM (2020). Comparative effectiveness of different treatment pathways for opioid use disorder. JAMA Network Open. 10.1001/jamanetworkopen.2019.20622. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Ward EN, Quaye ANA, & Wilens TE (2018). Opioid use disorders: Perioperative management of a special population. Anesthesia and Analgesia. 10.1213/ANE.0000000000003477. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Wen H, Hockenberry JM, Borders TF, & Druss BG (2017). Impact of medicaid expansion on medicaid-covered utilization of buprenorphine for opioid use disorder treatment. Medical Care.. 10.1097/MLR.0000000000000703. [DOI] [PubMed] [Google Scholar]
  58. Winkelman TNA, Ford BR, Shlafer RJ, McWilliams A, Admon LK, & Patrick SW (2020). Medications for opioid use disorder among pregnant women referred by criminal justice agencies before and after Medicaid expansion: A retrospective study of admissions to treatment centers in the United States. PLoS Medicine. 10.1371/journal.pmed.1003119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Winkelman TNA, Kieffer EC, Goold SD, Morenoff JD, Cross K, & Ayanian JZ (2016). Health insurance trends and access to behavioral healthcare among justice-involved individuals—United States, 2008–2014. Journal of General Internal Medicine. 10.1007/s11606-016-3845-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wu LT, Zhu H, & Swartz MS (2016). Treatment utilization among persons with opioid use disorder in the United States. Drug and Alcohol Dependence 10.1016/j.drugalcdep.2016.10.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Yang JC, Roman-Urrestarazu A, & Brayne C (2019). Differences in receipt of opioid agonist treatment and time to enter treatment for opioid use disorder among specialty addiction programs in the United States, 2014–17. PLoS One. 10.1371/journal.pone.0226349. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

mmc1
NIHMS1720935-supplement-mmc1.docx (16.9KB, docx)

RESOURCES