Advertisement

Disparities in Access to Opioid Treatment Programs and Office-Based Buprenorphine Treatment Across the Rural-Urban and Area Deprivation Continua: A US Nationwide Small Area Analysis

Open ArchivePublished:October 10, 2020DOI:https://doi.org/10.1016/j.jval.2020.08.2098

      Abstract

      Objectives

      To measure access to opioid treatment programs (OTPs) and office-based buprenorphine treatment (OBBTs) at the smallest geographic unit for which the Census Bureau publishes demographic and socioeconomic data (ie, block group) and to explore disparities in access to treatment across the rural-urban and area deprivation continua across the United States.

      Methods

      Access to OTPs and OBBTs at the block group in 2019 was quantified using an innovative 2-step floating catchment area technique that accounts for the supply of treatment facilities relative to the population size, proximity of facilities relative to the location of population in block groups, and time as a barrier within catchments. Block groups were stratified into tertiles based on the rural-urban continuum codes (metropolitan, micropolitan, small town, or rural) and area deprivation index (least-deprived, middle-deprived, most-deprived). The Integrated Nested Laplace Approximation approach was used for statistical analysis.

      Results

      Across the United States, 3329 block groups corresponding to 2 915 949 adults lacked access to OTPs within a 2-hour drive of their community and 130 block groups corresponding to 86 605 adults did not have access to OBBTs. Disparities in access to treatment were observed across the urban-rural and area deprivation continua including (1) lowest mean access score to OBBTs were found among most-deprived small towns, and (2) lower mean access score to OTPs were found among micropolitan and small towns.

      Conclusions

      The results of this study revealed disparities in access to medication-assisted treatment. The findings call for creative initiatives and local and regional policies to develop to mitigate access problems.

      Keywords

      Introduction

      The United States is experiencing a steady increase in the rates of drug overdose deaths, with opioids driving this crisis.
      National Institutes of Health
      Overdose death rates. National Institute on Drug Abuse, US Dept of Health and Human Services.
      In 2018, approximately 10.3 million Americans ages 12 or older had reported misusing opioids and 2.0 million were opioid dependent.
      Substance Abuse and Mental Health Services Administration, US Department of Health and Human Services
      Key substance use and mental health indicators in the United States: results from the 2018 National Survey on Drug Use and Health (HHS Publication No. PEP19-5068, NSDUH Series H-54).
      Medication-assisted treatment is an evidence-based practice that combines the use of medications and behavioral therapies, which is effective in the treatment of opioid use disorders and may sustain recovery.
      • Nielsen S.
      • Larance B.
      • Lintzeris N.
      Opioid agonist treatment for patients with dependence on prescription opioids.
      ,
      Substance Abuse and Mental Health Services Administration
      National Survey of Substance Abuse Treatment Services (N-SSATS): 2016. Data on Substance Abuse Treatment Facilities.
      Methadone, buprenorphine, and naltrexone are approved by the US Food and Drug Administration for the treatment of opioid use disorders.
      • Nielsen S.
      • Larance B.
      • Lintzeris N.
      Opioid agonist treatment for patients with dependence on prescription opioids.
      ,
      Substance Abuse and Mental Health Services Administration
      National Survey of Substance Abuse Treatment Services (N-SSATS): 2016. Data on Substance Abuse Treatment Facilities.
      A large literature on methadone accumulated over decades substantiates its relative safety and effectiveness in treating opioid use disorders.
      • Wolff K.
      Substance misuse: substitution drugs (methadone and buprenorphine): methadone and buprenorphine.
      ,
      • Whelan P.J.
      • Remski K.
      Buprenorphine vs methadone treatment: a review of evidence in both developed and developing worlds.
      Buprenorphine, a partial opioid agonist, reduces respiratory depression and the potential of overdose,
      • Wolff K.
      Substance misuse: substitution drugs (methadone and buprenorphine): methadone and buprenorphine.
      but because of weaker efficacy, divergence risk, and the risk of withdrawal, it is only recommended for clients with mild to moderate drug dependence and for those on low doses of methadone.
      • Whelan P.J.
      • Remski K.
      Buprenorphine vs methadone treatment: a review of evidence in both developed and developing worlds.
      There is limited evidence on the effectiveness of naltrexone for the treatment of opioid use disorders,
      • Minozzi S.
      • Amato L.
      • Vecchi S.
      • Davoli M.
      • Kirchmayer U.
      • Verster A.
      Oral naltrexone maintenance treatment for opioid dependence.
      and for this reason access to naltrexone programs was not part of this study.
      Methadone is only offered at opioid treatment programs (OTPs), where other forms of treatment for detoxification or maintenance may also be offered.
      • Nielsen S.
      • Larance B.
      • Lintzeris N.
      Opioid agonist treatment for patients with dependence on prescription opioids.
      ,
      Substance Abuse and Mental Health Services Administration
      National Survey of Substance Abuse Treatment Services (N-SSATS): 2016. Data on Substance Abuse Treatment Facilities.
      Buprenorphine can be offered in OTPs, facilities that do not have OTPs (non-OTP facilities), and office-based settings. According to the 2018 National Survey of Substance Abuse Treatment Services, 1095 of 1519 OTPs and 3854 of the 13 290 non-OTP facilities offered buprenorphine.
      Substance Abuse and Mental Health Services Administration
      National Survey of Substance Abuse Treatment Services (N-SSATS): 2016. Data on Substance Abuse Treatment Facilities.
      The Drug Addiction Treatment Act of 2000, enabled physicians who have received training to request a waiver from the Substance Abuse and Mental Health Services Administration and offer office-based buprenorphine treatment (OBBT).
      Substance Abuse and Mental Health Services Administration
      Statutes, regulations, and guidelines.
      Currently, in addition to physicians, nurse practitioners, physician assistants, clinical nurse specialists, certified registered nurse anesthetists, and certified nurse-midwives may offer OBBT.
      Substance Abuse and Mental Health Services Administration
      Statutes, regulations, and guidelines.
      Within the first year of obtaining a waiver, eligible physicians are permitted to treat up to 30 patients, which can be increased to 100 and 275 in the following years subject to approval by the Substance Abuse and Mental Health Services Administration.
      Substance Abuse and Mental Health Services Administration
      Understanding the final rule for a patient limit of 275.
      States have varying prescribing policies for other practitioners. In less restrictive states, nurse practitioners do not need a physician’s oversight for medication prescription and physician assistants’ scope of practice are determined by practices rather than states.
      • Spetz J.
      • Toretsky C.
      • Chapman S.
      • Phoenix B.
      • Tierney M.
      Nurse practitioner and physician assistant waivers to prescribe buprenorphine and state scope of practice restrictions.
      In 2018, approximately 400 000 (19.7%) individuals with an opioid use disorder received any substance use treatment at a specialty facility.
      Substance Abuse and Mental Health Services Administration
      Key substance use and mental health indicators in the United States: results from the 2018 National Survey on Drug Use and Health.
      Of the population treated, approximately 383 000 individuals were treated with methadone.
      • McBournie A.
      • Duncan A.
      • Connolly E.
      • Rising J.
      Methadone barriers persist, despite decades of evidence. Health Affairs blog..
      Thus, OTPs remain critical settings for the treatment of opioid use disorder, especially for clients with severe drug dependence. As of 2018, up to 3 million individuals could receive OBBT if providers were to prescribe up to their patient-limit capacity.
      • Grimm C.A.
      Geographic disparities affect access to buprenorphine services for opioid use disorder. Office of Inspector General, US Department of Health and Human Services.
      Providers are not offering treatment at or near their capacity, which may be attributed to stigma, low Medicaid insurance reimbursement rates, and/or prescribing requirements, among other factors.
      • Jones C.M.
      • McCance-Katz E.F.
      Characteristics and prescribing practices of clinicians recently waivered to prescribe buprenorphine for the treatment of opioid use disorder.
      ,
      • Thomas C.P.
      • Doyle E.
      • Kreiner P.W.
      • et al.
      Prescribing patterns of buprenorphine waivered physicians.
      Thus, the buprenorphine waiver program has the potential to significantly increase access to medication-assisted treatment.
      Gaps exist in availability and accessibility of OTPs and OBBTs across the United States.
      Substance Abuse and Mental Health Services Administration
      National Survey of Substance Abuse Treatment Services (N-SSATS): 2016. Data on Substance Abuse Treatment Facilities.
      ,
      • Rosenblum A.
      • Cleland C.M.
      • Fong C.
      • Kayman D.J.
      • Tempalski B.
      • Parrino M.
      Distance traveled and cross-state commuting to opioid treatment programs in the United States.
      ,
      • Jones C.W.
      • Christman Z.
      • Smith C.M.
      • et al.
      Comparison between buprenorphine provider availability and opioid deaths among US counties.
      Long distance to OTPs is associated with decreased length of stay in treatment, poorer treatment adherence, and reduced treatment retention.
      • Amiri S.
      • Lutz R.B.
      • McDonell M.G.
      • Roll J.M.
      • Amram O.
      Spatial access to opioid treatment program and alcohol and cannabis outlets: analysis of missed doses of methadone during the first, second, and third 90 days of treatment.
      • Amiri S.
      • Lutz R.
      • Socías E.
      • et al.
      Increased distance was associated with lower daily attendance to an opioid treatment program in Spokane County Washington.
      • Beardsley K.
      • Wish E.D.
      • Fitzelle D.B.
      • O’Grady K.
      • Arria A.M.
      Distance traveled to outpatient drug treatment and client retention.
      Results of a US nationwide study showed that approximately 25% of clients receiving methadone commuted more than 15 miles and 8% traveled more than 50 miles to receive treatment.
      • Rosenblum A.
      • Cleland C.M.
      • Fong C.
      • Kayman D.J.
      • Tempalski B.
      • Parrino M.
      Distance traveled and cross-state commuting to opioid treatment programs in the United States.
      Unlike methadone that requires intensive engagement with OTPs, buprenorphine can be prescribed and dispensed in office-based settings. Nonetheless, another study showed that approximately 26% of counties in the United States have fewer than 4 office-based buprenorphine providers and 7% do not have any.
      • Jones C.W.
      • Christman Z.
      • Smith C.M.
      • et al.
      Comparison between buprenorphine provider availability and opioid deaths among US counties.
      Of particular concern, rural counties, which experience similar burdens of opioid-related mortality as their urban counterparts, are far less likely to house substance use treatment facilities and office-based buprenorphine providers.
      • Look K.A.
      • Kile M.
      • Morgan K.
      • Roberts A.
      Community pharmacies as access points for addiction treatment.
      ,
      • Barnett M.L.
      • Lee D.
      • Frank R.G.
      In rural areas, buprenorphine waiver adoption since 2017 driven by nurse practitioners and physician assistants.
      Additionally, only 1 in 10 people with severe substance use disorders receive the specialty treatment they need in a given year.
      Substance Abuse and Mental Health Services Administration, US Department of Health and Human Services
      Facing addiction in America: the surgeon general’s report on alcohol, drugs, and health.
      These findings illustrate remarkable disparities in availability and accessibility of OTPs and OBBTs in the United States that may result in substantially worse treatment outcomes.
      Various techniques have been used to measure the spatial availability or accessibility of healthcare facilities.
      • Guagliardo M.F.
      Spatial accessibility of primary care: concepts, methods and challenges.
      The facility-to-population ratio and distance to the closest facility are among the most commonly used techniques.
      • Jones C.W.
      • Christman Z.
      • Smith C.M.
      • et al.
      Comparison between buprenorphine provider availability and opioid deaths among US counties.
      ,
      • Shi L.
      • Macinko J.
      • Starfield B.
      • Politzer R.
      • Wulu J.
      • Xu J.
      Primary care, social inequalities and all-cause, heart disease and cancer mortality in US counties: a comparison between urban and non-urban areas.
      • Skillman S.
      • Dahal A.
      Washington State’s Physician Workforce in 2016.
      Association of American Medical Colleges
      2017 state physician workforce data report.
      • Basu S.
      • Berkowitz S.A.
      • Phillips R.L.
      • Bitton A.
      • Landon B.E.
      • Phillips R.S.
      Association of primary care physician supply with population mortality in the United States, 2005-2015.
      These techniques are subject to several limitations. The facility-to-population ratio does not account for travel distance to facilities. Distance to the closest facility lacks consideration for supply of facilities and population that may have demand for such services.
      • Guagliardo M.F.
      Spatial accessibility of primary care: concepts, methods and challenges.
      To overcome these limitations, we applied a 2-step floating catchment area technique to measure access to OTPs and OBBTs using data from the 2018 US Census and the 2019 opioid treatment program directory and buprenorphine practitioner locator by the Substance Abuse and Mental Health Services Administration.
      Substance Abuse and Mental Health Services Administration
      Buprenorphine practitioner locator.
      ,
      Substance Abuse and Mental Health Services Administration
      Opioid treatment program directory.
      This technique accounts for 3 components of access: the supply of treatment facilities relative to the population size, proximity of facilities relative to the location of the population in block groups, and time as a barrier within catchments. The purpose of this study was to measure access to OTPs and OBBTs across the United States at the smallest geographic unit for which the US Census Bureau publishes demographic and socioeconomic data, and to explore variations in access to OTPs and OBBTs across the rural-urban and area-level deprivation continua. This approach is highly informative, because it provides avenues for health departments to develop and implement new policy solutions to mitigate access problems based on specific characteristics of individual communities.

      Methods

      This cross-sectional study was conducted at the level of census block group level. Census block groups generally contain between 600 and 3000 people, with an optimum size of 1500 people.
      United States Census Bureau
      Standard hierarchy of census geographic entities.
      Out of the 217 197 block groups in the United States, 216 537 (99.7%) were included in our analyses; the remaining 660 (0.3%) were excluded because of missing demographic or socioeconomic data. This study did not require institutional review board review because it used publicly available sociodemographic and medication-assisted treatment facility data.

      Measures

      Outcome variables

      Two continuous outcome variables, access to OTPs and access to OBBTs, were quantified at the block group level. To measure potential access to OTPs and OBBTs, we used a 2-step floating catchment area method with a distance decay function.
      • Schuurman N.
      • Bérubé M.
      • Crooks V.A.
      Measuring potential spatial access to primary health care physicians using a modified gravity model.
      This technique accounts for 3 components of access: the supply of treatment facilities relative to the population size, proximity of facilities relative to the location of population in block groups, and time as a barrier within catchments.
      • McGrail M.R.
      • Humphreys J.S.
      • Ward B.
      Accessing doctors at times of need–measuring the distance tolerance of rural residents for health-related travel.
      Facilities are represented by their geocoded address (ie, latitude and longitude). Block groups are represented by the latitude and longitude of population-weighted centroids that capture how populations are distributed in block groups. The 2-step floating catchment area method uses floating, overlapping catchment areas. The size of the catchment is determined by maximum travel time where all facilities within that catchment are accessible and those outside of the catchment are not accessible.
      • McGrail M.R.
      • Humphreys J.S.
      • Ward B.
      Accessing doctors at times of need–measuring the distance tolerance of rural residents for health-related travel.
      This method includes a distance decay function equivalent to recognizing that population groups are not equally likely to be accessing services from both nearby and up to the catchment boundaries, which denotes that the interaction between population and facilities decreases as distance between them increases.
      • Schuurman N.
      • Bérubé M.
      • Crooks V.A.
      Measuring potential spatial access to primary health care physicians using a modified gravity model.
      ,
      • Luo W.
      • Wang F.
      Measures of spatial accessibility to health care in a GIS environment: synthesis and a case study in the Chicago region.
      Data on the location of treatment facilities were assembled using the opioid treatment program directory and buprenorphine practitioner locator by the Substance Abuse and Mental Health Services Administration in 2019.
      Substance Abuse and Mental Health Services Administration
      Buprenorphine practitioner locator.
      ,
      Substance Abuse and Mental Health Services Administration
      Opioid treatment program directory.
      These data provided information on 1643 OTPs and 38 014 OBBTs in 50 states and the District of Columbia. ESRI ArcGIS Business Analyst USA Local Composite geocoder and R software were used for geocoding the addresses. Firstly, we calculated travel time between the longitude and latitude of population-weighted block group centroids and OTPs and OBBTs using the ESRI ArcGIS Network analyst origin destination matrix.
      Secondly, for each OTP and OBBT, we searched all block group centroids that were within a 120-minute drive from each treatment facility. No decay was applied to travel times less than 10 minutes.
      • Schuurman N.
      • Bérubé M.
      • Crooks V.A.
      Measuring potential spatial access to primary health care physicians using a modified gravity model.
      The decay for travel times between 10 and 120 minutes increased linearly and proportionally to the travel time.
      • Schuurman N.
      • Bérubé M.
      • Crooks V.A.
      Measuring potential spatial access to primary health care physicians using a modified gravity model.
      Travel times further than 120 minutes were considered inaccessible. The 120-minute travel time cutoff has been previously used and was therefore chosen to accommodate differences in population and healthcare service distribution.
      • Schuurman N.
      • Bérubé M.
      • Crooks V.A.
      Measuring potential spatial access to primary health care physicians using a modified gravity model.
      Thirdly, we calculated a demand ratio for OTPs and OBBTs by accounting for the population of block groups served and travel time impedance between treatment facilities and block groups. Our analysis included population of people 18 and over, because this population is mainly admitted to treatment programs. Lastly, for each block group centroid, we searched all treatment facilities that were within a 120-minute drive from each block group. An access score for block groups was calculated by summing up the demand score for OTPs or OBBTs that serve each block group. Access scores represent a ratio of OTPs or OBBTs (supply) to the population of block groups served (demand) with only selected treatment facilities and block group population entering the numerator and denominator.
      • Schuurman N.
      • Bérubé M.
      • Crooks V.A.
      Measuring potential spatial access to primary health care physicians using a modified gravity model.

      Explanatory variables

      We used the Rural-Urban Commuting Area (RUCA) codes for delineating among metropolitan, micropolitan, or small town or rural areas.
      WWAMI Rural Health Research Center
      Rural-urban commuting area codes (RUCAs).
      ,
      United States Department of Agriculture
      Rural-urban commuting area codes.
      RUCA codes use work commuting information, population data, and measures of urbanization to classify neighborhoods into urban and rural areas.
      WWAMI Rural Health Research Center
      Rural-urban commuting area codes (RUCAs).
      ,
      United States Department of Agriculture
      Rural-urban commuting area codes.
      Block groups with RUCA primary codes of 1 to 3 were classified as metropolitan areas (primary flow within or to urbanized areas), those with RUCA primary codes of 4 to 6 were classified as micropolitan areas (primary flow within or to urban clusters), and those with RUCA primary codes of 7-10 were classified as small towns or rural areas (primary flow within or to small urban clusters).
      WWAMI Rural Health Research Center
      Rural-urban commuting area codes (RUCAs).
      ,
      United States Department of Agriculture
      Rural-urban commuting area codes.
      We used area deprivation index (ADI), a validated composite score of socioeconomic disadvantage, to represent the social and economic characteristics of block groups.
      • Singh G.K.
      Area deprivation and widening inequalities in US mortality, 1969-1998.
      ,
      • Knighton A.J.
      • Savitz L.
      • Belnap T.
      • Stephenson B.
      • VanDerslice J.
      Introduction of an area deprivation index measuring patient socioeconomic status in an integrated health system: implications for population health.
      The ADI was developed based on 17 Census variables in 4 domains of poverty, housing, employment, and education.
      • Singh G.K.
      Area deprivation and widening inequalities in US mortality, 1969-1998.
      ,
      • Knighton A.J.
      • Savitz L.
      • Belnap T.
      • Stephenson B.
      • VanDerslice J.
      Introduction of an area deprivation index measuring patient socioeconomic status in an integrated health system: implications for population health.
      ADI scores range from 1 to 100 with higher scores indicating greater deprivation.
      • Kind A.J.H.
      • Buckingham W.R.
      Making neighborhood-disadvantage metrics accessible – the neighborhood atlas.
      We divided census block groups into tertiles of deprivation (1 = least deprived, 2 = middle deprived, 3 = most deprived).

      Control variables

      We obtained data on percent of black or African American population and percent of Hispanic population at the block group level from the 2017 American Community Survey 5-year estimates.
      United States Census Bureau
      American Community Survey 5-year estimates.
      These data were the most recent Census data available at the time of this analysis.

      Statistical Analysis

      Univariate analysis described measures of central tendency and variability for the explanatory and outcome variables. Statistical analyses relied on a Bayesian spatial modeling, techniques that are very flexible in capturing complexities of high-dimensional spatial data.
      • Blangiardo M.
      • Cameletti M.
      • Baio G.
      • Rue H.
      Spatial and spatio-temporal models with R-INLA.
      • Blangiardo M.
      Spatial and Spatio-temporal Bayesian Models with R-INLA.
      • Schrödle B.
      • Held L.
      Spatio-temporal disease mapping using INLA.
      These techniques adjust for spatial autocorrelation or whether access to treatment and mortality are clustered, randomly distributed, or dispersed. This is important because areas closer to each other tend to have the similar characteristics than areas far apart. Mean for access to OTPs and OBBTs and their 95% credible intervals across the rural-urban and area-level deprivation continua were estimated using Gaussian models. The spatially structured residual was modeled using an intrinsic conditional autoregressive model and the unstructured residual using exchangeability among block groups. The neighborhood adjacency matrix was defined using a US block group shapefile. The adjacency matrix presented the pattern of neighborhood structure and identified isolated and central areas. Models included fixed effects for RUCA or ADI or joint effect of RUCA and ADI. Models were adjusted for percent of black or African American population and percent of Hispanic population.
      Bayesian models present posterior estimates of the effect of explanatory variables expressed as mean, 95% credible intervals for the mean, and relative rates (calculated as exponential [raw coefficient]). Credible intervals that exclude 0 for the mean (or 1 for relative rates) are considered significant.
      • Blangiardo M.
      • Cameletti M.
      • Baio G.
      • Rue H.
      Spatial and spatio-temporal models with R-INLA.
      • Blangiardo M.
      Spatial and Spatio-temporal Bayesian Models with R-INLA.
      • Schrödle B.
      • Held L.
      Spatio-temporal disease mapping using INLA.
      The Integrated Nested Laplace Approximation package in R was used for generating Bayesian statistics.

      Results

      The 2-step floating catchment area technique returned 247.2 million possible routes of travel between OBBTs and block groups and 11.5 million routes between OTPs and block groups. Appendices show descriptive statistics for access measures summarized for the 50 states and the District of Columbia (see Appendices in Supplemental Materials found at https://doi.org/10.1016/j.jval.2020.08.2098). Across the United States, 3 329 block groups corresponding to 2 915 949 adults lacked access to OTPs within a 2-hour drive of their community and 130 block groups corresponding to 86 605 adults did not have access to OBBTs.
      Three of the 5 states with the lowest population densities in the United States (Alaska, Wyoming, and South Dakota) had the largest percentage of population with no access to OBBTs or OTPs. Alaska had the largest percentage of the population with no access to OBBTs, followed by Idaho and Nevada. Wyoming, South Dakota, and Alaska had the largest percentage of population with no access to OTPs. Figure 1 shows access to OTPs and OBBTs at the block group level in the United States.
      Figure thumbnail gr1
      Figure 1Location of office-based buprenorphine treatments (OBBTs) (top left), access to OBBTs (top right), location of opioid treatment programs (OTPs) (bottom left), and access to OTPs (bottom right) in the United States. Locations of OTPs and OBBTs were obtained from the Opioid Treatment Program Directory and Buprenorphine Practitioner Locator (Substance Abuse and Mental Health Services Administration). Access to OTPs and OBBTs was calculated using a 2-step floating catchment area method with a distance decay function. This technique accounts for 3 components of access: the supply of treatment facilities relative to the population size, proximity of facilities relative to the location of the population in block groups, and time as a barrier within catchments. Larger values represent higher access to treatment accounting for the 3 components of access.
      Source: Authors’ analysis of data for 2019 from the Opioid Treatment Program Directory and Buprenorphine Practitioner Locator (Substance Abuse and Mental Health Services Administration).
      The effect of RUCA or ADI on access to OBBTs and OTPs from the Integrated Nested Laplace Approximation models is shown in Table 1. Disparities in access to treatment were observed across the urban-rural continuum. Micropolitan areas and small towns had significantly lower mean access scores to OBBTs compared to metropolitan areas. The mean access score to OBBTs was lower, respectively, by 22% (mean = −0.24, SD = 0.005) and 25% (mean = −0.29, SD = 0.004) for micropolitan areas and small towns compared to metropolitan block groups. Area-level deprivation was also significantly and negatively associated with access score. Mean access score to OBBTs was lower, respectively, by 17% (mean = −0.18, SD = 0.002) and 20% (mean = −0.23, SD = 0.003) for middle-deprived and most-deprived areas compared to least-deprived block groups.
      Table 1Posterior estimates of the effect of RUCA and ADI on access to OBBTs and OTPs from the INLA models.
      Treatment programVariablesMeanSD2.5% CI97.5% CIRate ratio
      OBBTRUCAMetropolitan (reference)
      Micropolitan−0.240.005−0.25−0.240.78
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Small town or rural−0.290.004−0.30−0.290.75
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Spatial structure (iCAR)23.940.6722.5625.29---
      Spatial unstructured (exchangeable)13.550.1013.3613.73---
      ADILeast-deprived (reference)
      Middle-deprived−0.180.002−0.19−0.180.83
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Most-deprived−0.230.003−0.23−0.230.80
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Spatial structure (iCAR)24.860.6623.4726.25---
      Spatial unstructured (exchangeable)12.690.0912.5212.86---
      OTPRUCAMetropolitan (reference)
      Micropolitan−0.140.002−0.15−0.140.87
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Small town or rural−0.160.002−0.16−0.160.85
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Spatial structure (iCAR)234.577.04220.62249.65---
      Spatial unstructured (exchangeable)75.300.4274.4776.14---
      ADILeast-deprived (reference)
      Middle-deprived−0.050.001−0.05−0.050.95
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Most-deprived−0.060.001−0.06−0.060.94
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Spatial structure (iCAR)228.416.80214.86242.98---
      Spatial unstructured (exchangeable)71.820.4071.0472.60---
      Note. Models are adjusted for percent of black or African American population, percent of Hispanic population, and percent of families with income below 185% of federal poverty line.
      ADI indicates area deprivation index; CI, credible interval; INLA, integrated nested Laplace approximation; OBBTs, office-based buprenorphine treatments; OTPs, opioid treatment programs; RUCA, rural-urban commuting area.
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Similar results were observed for access to OTPs. Micropolitan areas and small towns had lower mean access score to OTPs compared to metropolitan block groups. The mean access score to OTPs was lower, respectively, by 13% (mean = −0.14, SD = 0.002) and 15% (mean = −0.16, SD = 0.002) for micropolitan areas and small towns compared to metropolitan block groups. Area-level deprivation was negatively associated with access score. The mean access score was lower respectively, by 5% (mean = −0.05, SD = 0.001) and 6% (mean = −0.06, SD = 0.001) for middle-deprived and most-deprived areas compared to least-deprived block groups.
      The joint effect of RUCA and ADI on access to OBBTs and OTPs is shown in Figure 2 and Table 2. The interaction of rurality and deprivation had a negative effect on the mean access score to OBBTs. The least-deprived micropolitan areas and small towns had a lower mean access score to OBBTs compared to least-, middle-, and most-deprived metropolitan areas. The most-deprived small towns had the lowest mean access score to OBBTs compared to the least-deprived metropolitan areas (mean = −0.47, SD = 0.005). In access to OTPs, micropolitan areas and small towns had a lower mean access score to OTPs compared to their metropolitan counterparts. Access to OTPs was not greatly decreased by increasing level of deprivation.
      Figure thumbnail gr2
      Figure 2The joint effect of rural-urban commuting area (RUCA) and area deprivation index (ADI) on access to office-based buprenorphine treatments (OBBTs) (top) and opioid treatment programs (OTPs) (bottom).
      Source: Authors’ analysis of data for 2019 from the Opioid Treatment Program Directory and Buprenorphine Practitioner Locator (Substance Abuse and Mental Health Services Administration).
      Table 2Posterior estimates of the joint effect of RUCA and ADI on access to OBBTs and OTPs from the INLA models.
      Treatment programVariablesMeanSD2.5% CI97.5% CIRate ratio
      OBBTMetropolitan and least-deprived (reference)
      Metropolitan and middle-deprived−0.200.008−0.21−0.180.82
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Metropolitan and most-deprived−0.230.010−0.25−0.210.80
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Micropolitan and least-deprived−0.170.002−0.18−0.170.84
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Micropolitan and middle-deprived−0.370.006−0.38−0.360.69
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Micropolitan and least-deprived−0.420.006−0.43−0.400.66
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Small town or rural and least-deprived−0.200.003−0.21−0.200.82
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Small town or rural and middle-deprived−0.430.005−0.44−0.420.65
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Small town or rural and least-deprived−0.470.005−0.48−0.460.63
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Spatial structure (iCAR)25.190.71723.7226.69---
      Spatial unstructured (exchangeable)13.20.09213.0213.38---
      OTPMetropolitan and least-deprived (reference)
      Metropolitan and middle-deprived−0.190.003−0.20−0.190.82
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Metropolitan and most-deprived−0.230.003−0.24−0.230.79
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Micropolitan and least-deprived−0.050.001−0.05−0.050.95
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Micropolitan and middle-deprived−0.170.002−0.17−0.170.84
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Micropolitan and least-deprived−0.190.002−0.19−0.190.83
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Small town or rural and least-deprived−0.060.001−0.06−0.060.94
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Small town or rural and middle-deprived−0.180.002−0.19−0.180.83
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Small town or rural and least-deprived−0.200.002−0.20−0.190.82
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.
      Spatial structure (iCAR)243.917.436229.32259.85---
      Spatial unstructured (exchangeable)74.460.41773.6475.28---
      Note. Models are adjusted for percent of black or African American population, percent of Hispanic population, and percent of families with income below 185% of federal poverty line.
      ADI indicates area deprivation index; CI, credible interval; INLA, integrated nested Laplace approximation; OBBTs, office-based buprenorphine treatments; OTPs, opioid treatment programs; RUCA, rural-urban commuting area.
      Indicates findings that are well-supported by the data as evidenced by CIs that exclude 0 for the mean.

      Discussion

      We applied a fine-grained and sophisticated technique to measure access to OTPs and OBBTs instead of relying on conventional techniques that do not adequately illustrate nuances of potential spatial access or assume uniform spatial access to treatment facilities.
      • Schuurman N.
      • Bérubé M.
      • Crooks V.A.
      Measuring potential spatial access to primary health care physicians using a modified gravity model.
      Additionally, our analyses were conducted at the block group level, the smallest geographic unit for which the US Census Bureau publishes demographic and socioeconomic data, rather than at larger spatial units such as the county, which might obscure important observations. In this regard, counties can range in size from 26 (Arlington County, Virginia) to 20 105 square miles (San Bernardino County, California), with populations ranging from 74 (Loving County, Texas) to 10 105 722 (Los Angeles, California).
      Our results contribute to an ongoing challenge on spatial accessibility of medication-assisted treatment facilities. Across the United States, 3329 block groups corresponding to 2 915 949 adults did not have access to OTPs and 130 block groups corresponding to 86 605 adults did not have access to OBBTs within a 2-hour drive of their community. Three of the 5 states with the lowest population densities in the United States (Alaska, Wyoming, and South Dakota) had the largest percentage of the population with no access to OBBTs or OTPs. Some states, such as South Dakota and Wyoming, intend to opt out of the Medicaid expansion,
      • Cummings J.R.
      • Allen L.
      • Clennon J.
      • Ji X.
      • Druss B.G.
      Geographic access to specialty mental health care across high- and low-income US cmmunities.
      which may reduce access or make them ineligible for a federal Substance Abuse Prevention and Treatment Block Grant. This lack of funding, in turn, threatens to worsen disparities in access to medication-assisted treatment in these states.
      • Cummings J.R.
      • Allen L.
      • Clennon J.
      • Ji X.
      • Druss B.G.
      Geographic access to specialty mental health care across high- and low-income US cmmunities.
      Disparities in access to treatment were observed across the urban-rural and area-deprivation continua. Of primary interest and a unique aspect of our study was the demonstration of the combined effect of RUCA and ADI on access to OBBTs and OTPs. The least-deprived micropolitan areas and small towns had a lower mean access score to OBBTs compared to the least-, middle-, and most-deprived metropolitan areas. The most-deprived small towns had the lowest mean access score to OBBTs. In access to OTPs, micropolitan areas and small towns had a lower mean access score to OTPs compared to their metropolitan counterparts. In contrast, access to OTPs did not greatly decrease by increasing the level of deprivation. These disparities in access to OBBT align well with limited availability and accessibility of healthcare services in small towns and rural areas,
      • Shi L.
      • Macinko J.
      • Starfield B.
      • Politzer R.
      • Wulu J.
      • Xu J.
      Primary care, social inequalities and all-cause, heart disease and cancer mortality in US counties: a comparison between urban and non-urban areas.
      ,
      • Shi L.
      • Macinko J.
      • Starfield B.
      • Politzer R.
      • Xu J.
      Primary care, race, and mortality in US states.
      ,
      • Shi L.
      • Macinko J.
      • Starfield B.
      • Politzer R.
      • Wulu J.
      • Xu J.
      Primary care, social inequalities, and all-cause, heart disease, and cancer mortality in US counties, 1990.
      specifically buprenorphine prescribers
      • Andrilla C.H.A.
      • Coulthard C.
      • Patterson D.G.
      Prescribing practices of rural physicians waivered to prescribe buprenorphine.
      and mental health professionals.
      • Cummings J.R.
      • Allen L.
      • Clennon J.
      • Ji X.
      • Druss B.G.
      Geographic access to specialty mental health care across high- and low-income US cmmunities.
      Several features of methadone and buprenorphine treatment in rural and underserved settings deserve comment. According to the Substance Abuse and Mental Health Services Administration’s Buprenorphine Practitioner Locator, 67% of buprenorphine prescribers were allopathic and osteopathic physicians, 18% were nurse practitioners, and less than 5% were physician’s assistants. Although the number of buprenorphine prescribers has increased across the United States, disparities in access to buprenorphine are a major problem.
      • Andrilla C.H.A.
      • Coulthard C.
      • Patterson D.G.
      Prescribing practices of rural physicians waivered to prescribe buprenorphine.
      ,
      • Abraham A.J.
      • Andrews C.M.
      • Harris S.J.
      • Friedmann P.D.
      Availability of medications for the treatment of alcohol and opioid use disorder in the USA.
      Most prescribers are located in urban areas while rural areas encompass 97% of the nation’s land area and contain 20% of the US population.
      United States Department of Agriculture
      Population & migration.
      Additionally, fewer percentages of nurse practitioners obtained waivers in more restrictive states where nurse practitioners needed physician oversight to prescribe medication.
      • Spetz J.
      • Toretsky C.
      • Chapman S.
      • Phoenix B.
      • Tierney M.
      Nurse practitioner and physician assistant waivers to prescribe buprenorphine and state scope of practice restrictions.
      Although figures regarding the supply of providers who can subscribe buprenorphine may be underestimated or overestimated, a census of buprenorphine prescribers is necessary for making informed decisions about solutions to the unavailability of OBBT in specific areas.
      Rural and disadvantaged individuals may not seek healthcare not only because of limited availability and accessibility of providers, but also owing to concerns about stigma, confidentiality, and being disproportionately uninsured or underinsured.
      • Kelly J.
      • Rane D.
      • Huylebroeck B.
      • Wortley P.
      • Drenzek C.
      Case study: Georgia’s rural versus non-rural populations.
      • Gong G.
      • Phillips S.G.
      • Hudson C.
      • Curti D.
      • Philips B.U.
      Higher US rural mortality rates linked to socioeconomic status, physician shortages, and lack of health insurance.
      • Warner T.D.
      • Monaghan-Geernaert P.
      • Battaglia J.
      • Brems C.
      • Johnson M.E.
      • Roberts L.W.
      Ethical considerations in rural health care: a pilot study of clinicians in Alaska and New Mexico.
      • Handley T.E.
      • Kay-Lambkin F.J.
      • Inder K.J.
      • et al.
      Self-reported contacts for mental health problems by rural residents: predicted service needs, facilitators and barriers.
      Buprenorphine treatment is covered by Medicaid programs in all states,
      • Askari M.S.
      • Martins S.S.
      • Mauro P.M.
      Medication for opioid use disorder treatment and specialty outpatient substance use treatment outcomes: differences in retention and completion among opioid-related discharges in 2016.
      while methadone as medication-assisted treatment is reimbursed by 42 states only.
      Substance Abuse and Mental Health Services Administration
      Medicaid Coverage of Medication-Assisted Treatment for Alcohol and Opioid Use Disorders and of Medication for the Reversal of Opioid Overdose.
      According to the 2017 National Survey of Substance Abuse Treatment Services, only 35% of OTPs accepted Medicare,
      Substance Abuse and Mental Health Services Administration
      National Survey of Substance Abuse Treatment Services (N-SSATS): 2017. Data on Substance Abuse Treatment Facilities.
      and it was not until January 2020 that Medicare started covering the cost of medication-assisted treatment at OTPs.
      Centers for Medicare & Medicaid Services, US Department of Health and Human Services
      Revisions to payment policies under the physician fee schedule and other changes to part B payment policies.
      Collectively, these factors may result in late diagnoses and treatment of health conditions, and associated greater morbidity and mortality, especially for chronic physical and behavioral conditions among rural residents.
      • Henning-Smith C.
      • Kozhimannil K.
      • Casey M.
      • Prasad S.
      • Moscovice I.
      Rural-urban differences in Medicare quality outcomes and the impact of risk adjustment.
      ,
      • Anderson T.J.
      • Saman D.M.
      • Lipsky M.S.
      • Lutfiyya M.N.
      A cross-sectional study on health differences between rural and non-rural U.S. counties using the County Health Rankings.
      Educating the entire healthcare workforce about the benefits of medication treatment, reducing stigma around treatment-seeking, and improving health insurance coverage of medication treatment are ways to expand availability and accessibility of treatment.
      • Abraham A.J.
      • Andrews C.M.
      • Harris S.J.
      • Friedmann P.D.
      Availability of medications for the treatment of alcohol and opioid use disorder in the USA.
      Identifying and encouraging providers to obtain buprenorphine waivers in high-need areas with low waivered providers can also increase access to treatment.
      • Grimm C.A.
      Geographic disparities affect access to buprenorphine services for opioid use disorder. Office of Inspector General, US Department of Health and Human Services.
      Recruiting students with rural backgrounds, offering curricula and rotations on rural practice, and providing financial incentives such as federal loan repayment programs may help retain providers in rural communities.
      • Brooks R.G.
      • Walsh M.
      • Mardon R.E.
      • Lewis M.
      • Clawson A.
      The roles of nature and nurture in the recruitment and retention of primary care physicians in rural areas: a review of the literature.
      Expanding medication treatment in Federally Qualified Health Centers may expand availability of medication treatment in underserved areas. In addition, telepsychiatry holds promise to improve accessibility of medication-assisted treatment services in rural and disadvantaged poulations.
      • Cummings J.R.
      • Allen L.
      • Clennon J.
      • Ji X.
      • Druss B.G.
      Geographic access to specialty mental health care across high- and low-income US cmmunities.
      Our analysis is not without limitations. First, we measured access to OTPs and OBBTs only. Our access scores do not include specialty non-opioid treatment programs or care provided in other settings such as mental health centers, inpatient and outpatient hospital settings, or detention centers. Thus, the degree of disparities in access to buprenorphine treatment may be overstated. Second, we used a 120-minute cutoff to measure potential access to OBBTs and OTPs. This cutoff was selected to accommodate urban and rural populations and their potential drive time to seek treatment. Individuals may use public transportation to commute to treatment facilities, and our analysis did not account for different transport modalities. Third, our method measures potential spatial access and not actual ability to use healthcare services. Fourth, our analysis included the population of people 18 and over in block groups, because this population is admitted to treatment programs. Finally, our analysis does not account for treatment capacity, wait time, or type of insurance coverage. OTPs are understaffed and underfunded, and patients may have to wait for months to be admitted.
      Nonetheless, this study is the first national examination of access to OBBTs and OTPs taking into account area-level population, supply of OTPs or OBBTs, and travel time between treatment facilities and area-level population. The disparities to medication-assisted treatment in the United States call for immediate action, especially in deprived and rural communities. The opioid use disorder public health crisis calls for targeted policy responses to ensure broad access to evidence-based prevention, treatment, and recovery services for all Americans.
      US Department of Health and Human Services
      Facing Addiction in America: The Surgeon General’s Report on Alcohol, Drugs, and Health.
      Author Contributions: Concept and design: Amiri, Amram
      Acquisition of data: Amiri
      Analysis and interpretation of data: Amiri, McDonell, Denney, Buchwald, Amram
      Drafting of the manuscript: Amiri, McDonell, Denney, Buchwald
      Critical revision of paper for important intellectual content: Amiri, McDonell, Denney, Amram
      Statistical analysis: Amiri
      Obtaining funding: Amiri
      Supervision: Buchwald
      Conflict of Interest Disclosures: Dr Amiri reported receiving grants from Washington State University during the conduct of the study. Dr McDonell reported receiving grants from the National Institute on Alcohol Abuse and Alcoholism (R01 AA022070, R01 AA020248, R21 AA027045 and R41 AA026793), Substance Abuse and Mental Health Services Administration (H79 TI 092557), Health Resources and Services Administration, the National Institute on Drug Abuse (R44 DA049629) outside the submitted work. Drs Denney and Amram reported receiving grants from Washington State University during the conduct of the study. No other disclosures were reported.
      Funding/Support: This investigation was supported in part by funds provided for medical and biological research by the State of Washington Initiative Measure No.171.
      Role of Funder/Sponsor: The funder has no role in the design, data collection, analysis, and interpretation of the results; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

      Supplemental Material

      References

        • National Institutes of Health
        Overdose death rates. National Institute on Drug Abuse, US Dept of Health and Human Services.
        • Substance Abuse and Mental Health Services Administration, US Department of Health and Human Services
        Key substance use and mental health indicators in the United States: results from the 2018 National Survey on Drug Use and Health (HHS Publication No. PEP19-5068, NSDUH Series H-54).
        Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration, Rockville, MD2019
        • Nielsen S.
        • Larance B.
        • Lintzeris N.
        Opioid agonist treatment for patients with dependence on prescription opioids.
        JAMA. 2017; 317: 967-968
        • Substance Abuse and Mental Health Services Administration
        National Survey of Substance Abuse Treatment Services (N-SSATS): 2016. Data on Substance Abuse Treatment Facilities.
        • Wolff K.
        Substance misuse: substitution drugs (methadone and buprenorphine): methadone and buprenorphine.
        in: Payne-James J. Byard R.W. Corey T.S. Henderson C. Encyclopedia of Forensic & Legal Medicine. Elsevier B.V., Oxford, England2005
        • Whelan P.J.
        • Remski K.
        Buprenorphine vs methadone treatment: a review of evidence in both developed and developing worlds.
        J Neurosci Rural Pract. 2012; 3: 45-50
        • Minozzi S.
        • Amato L.
        • Vecchi S.
        • Davoli M.
        • Kirchmayer U.
        • Verster A.
        Oral naltrexone maintenance treatment for opioid dependence.
        Cochrane Database Syst Rev. 2011; 4CD001333
        • Substance Abuse and Mental Health Services Administration
        Statutes, regulations, and guidelines.
        • Substance Abuse and Mental Health Services Administration
        Understanding the final rule for a patient limit of 275.
        • Spetz J.
        • Toretsky C.
        • Chapman S.
        • Phoenix B.
        • Tierney M.
        Nurse practitioner and physician assistant waivers to prescribe buprenorphine and state scope of practice restrictions.
        JAMA. 2019; 321: 1407-1408
        • Substance Abuse and Mental Health Services Administration
        Key substance use and mental health indicators in the United States: results from the 2018 National Survey on Drug Use and Health.
        • McBournie A.
        • Duncan A.
        • Connolly E.
        • Rising J.
        Methadone barriers persist, despite decades of evidence. Health Affairs blog..
        • Grimm C.A.
        Geographic disparities affect access to buprenorphine services for opioid use disorder. Office of Inspector General, US Department of Health and Human Services.
        • Jones C.M.
        • McCance-Katz E.F.
        Characteristics and prescribing practices of clinicians recently waivered to prescribe buprenorphine for the treatment of opioid use disorder.
        Addiction. 2019; 114: 471-482
        • Thomas C.P.
        • Doyle E.
        • Kreiner P.W.
        • et al.
        Prescribing patterns of buprenorphine waivered physicians.
        Drug Alcohol Depend. 2017; 181: 213-218
        • Rosenblum A.
        • Cleland C.M.
        • Fong C.
        • Kayman D.J.
        • Tempalski B.
        • Parrino M.
        Distance traveled and cross-state commuting to opioid treatment programs in the United States.
        J Environ Public Health. 2011; 2011948789
        • Jones C.W.
        • Christman Z.
        • Smith C.M.
        • et al.
        Comparison between buprenorphine provider availability and opioid deaths among US counties.
        J Subst Abuse Treat. 2018; 93: 19-25
        • Amiri S.
        • Lutz R.B.
        • McDonell M.G.
        • Roll J.M.
        • Amram O.
        Spatial access to opioid treatment program and alcohol and cannabis outlets: analysis of missed doses of methadone during the first, second, and third 90 days of treatment.
        Am J Drug Alcohol Abuse. 2020; 46: 78-87
        • Amiri S.
        • Lutz R.
        • Socías E.
        • et al.
        Increased distance was associated with lower daily attendance to an opioid treatment program in Spokane County Washington.
        J Subst Abuse Treat. 2018; 93: 26-30
        • Beardsley K.
        • Wish E.D.
        • Fitzelle D.B.
        • O’Grady K.
        • Arria A.M.
        Distance traveled to outpatient drug treatment and client retention.
        J Subst Abuse Treat. 2003; 25: 279-285
        • Look K.A.
        • Kile M.
        • Morgan K.
        • Roberts A.
        Community pharmacies as access points for addiction treatment.
        Res Social Adm Pharm. 2019; 15: 404-409
        • Barnett M.L.
        • Lee D.
        • Frank R.G.
        In rural areas, buprenorphine waiver adoption since 2017 driven by nurse practitioners and physician assistants.
        Health Aff. 2019; 38: 2048-2056
        • Substance Abuse and Mental Health Services Administration, US Department of Health and Human Services
        Facing addiction in America: the surgeon general’s report on alcohol, drugs, and health.
        • Guagliardo M.F.
        Spatial accessibility of primary care: concepts, methods and challenges.
        Int J Health Geogr. 2004; 3: 3
        • Shi L.
        • Macinko J.
        • Starfield B.
        • Politzer R.
        • Wulu J.
        • Xu J.
        Primary care, social inequalities and all-cause, heart disease and cancer mortality in US counties: a comparison between urban and non-urban areas.
        Public Health. 2005; 119: 699-710
        • Skillman S.
        • Dahal A.
        Washington State’s Physician Workforce in 2016.
        Center for Health Workforce Studies, University of Washington, Seattle, WA2017
        • Association of American Medical Colleges
        2017 state physician workforce data report.
        2017
        • Basu S.
        • Berkowitz S.A.
        • Phillips R.L.
        • Bitton A.
        • Landon B.E.
        • Phillips R.S.
        Association of primary care physician supply with population mortality in the United States, 2005-2015.
        JAMA Intern Med. 2019; 179: 506-514
        • Substance Abuse and Mental Health Services Administration
        Buprenorphine practitioner locator.
        • Substance Abuse and Mental Health Services Administration
        Opioid treatment program directory.
        • United States Census Bureau
        Standard hierarchy of census geographic entities.
        • Schuurman N.
        • Bérubé M.
        • Crooks V.A.
        Measuring potential spatial access to primary health care physicians using a modified gravity model.
        Can Geogr. 2010; 54: 29-45
        • McGrail M.R.
        • Humphreys J.S.
        • Ward B.
        Accessing doctors at times of need–measuring the distance tolerance of rural residents for health-related travel.
        BMC Health Serv Res. 2015; 15: 212
        • Luo W.
        • Wang F.
        Measures of spatial accessibility to health care in a GIS environment: synthesis and a case study in the Chicago region.
        Environ Plann B Plann Des. 2003; 30: 865-884
        • WWAMI Rural Health Research Center
        Rural-urban commuting area codes (RUCAs).
        http://depts.washington.edu/uwruca/
        Date accessed: September 15, 2018
        • United States Department of Agriculture
        Rural-urban commuting area codes.
        • Singh G.K.
        Area deprivation and widening inequalities in US mortality, 1969-1998.
        Am J Public Health. 2003; 93: 1137-1143
        • Knighton A.J.
        • Savitz L.
        • Belnap T.
        • Stephenson B.
        • VanDerslice J.
        Introduction of an area deprivation index measuring patient socioeconomic status in an integrated health system: implications for population health.
        EGEMS (Wash DC). 2016; 4: 1238
        • Kind A.J.H.
        • Buckingham W.R.
        Making neighborhood-disadvantage metrics accessible – the neighborhood atlas.
        N Engl J Med. 2018; 378: 2456-2458
        • United States Census Bureau
        American Community Survey 5-year estimates.
        2017
        http://factfinder.census.gov/
        Date accessed: February 15, 2019
        • Blangiardo M.
        • Cameletti M.
        • Baio G.
        • Rue H.
        Spatial and spatio-temporal models with R-INLA.
        Spat Spatiotemporal Epidemiol. 2013; 4: 39-55
        • Blangiardo M.
        Spatial and Spatio-temporal Bayesian Models with R-INLA.
        Wiley, Chichester, England2015
        • Schrödle B.
        • Held L.
        Spatio-temporal disease mapping using INLA.
        Environmetrics. 2011; 22: 725-734
        • Cummings J.R.
        • Allen L.
        • Clennon J.
        • Ji X.
        • Druss B.G.
        Geographic access to specialty mental health care across high- and low-income US cmmunities.
        JAMA Psychiatry. 2017; 74: 476-484
        • Shi L.
        • Macinko J.
        • Starfield B.
        • Politzer R.
        • Xu J.
        Primary care, race, and mortality in US states.
        Soc Sci Med. 2005; 61: 65-75
        • Shi L.
        • Macinko J.
        • Starfield B.
        • Politzer R.
        • Wulu J.
        • Xu J.
        Primary care, social inequalities, and all-cause, heart disease, and cancer mortality in US counties, 1990.
        Am J Public Health. 2005; 95: 674-680
        • Andrilla C.H.A.
        • Coulthard C.
        • Patterson D.G.
        Prescribing practices of rural physicians waivered to prescribe buprenorphine.
        Am J Prev Med. 2018; 54: S208-S214
        • Abraham A.J.
        • Andrews C.M.
        • Harris S.J.
        • Friedmann P.D.
        Availability of medications for the treatment of alcohol and opioid use disorder in the USA.
        Neurotherapeutics. 2020; 17: 55-69
        • United States Department of Agriculture
        Population & migration.
        • Kelly J.
        • Rane D.
        • Huylebroeck B.
        • Wortley P.
        • Drenzek C.
        Case study: Georgia’s rural versus non-rural populations.
        in: Parks F.M. Felzien G.S. Jue S. HIV/AIDS in Rural Communities: Research, Education, and Advocacy. Springer International Publishing, Cham2017: 3-20
        • Gong G.
        • Phillips S.G.
        • Hudson C.
        • Curti D.
        • Philips B.U.
        Higher US rural mortality rates linked to socioeconomic status, physician shortages, and lack of health insurance.
        Health Aff. 2019; 38: 2003-2010
        • Warner T.D.
        • Monaghan-Geernaert P.
        • Battaglia J.
        • Brems C.
        • Johnson M.E.
        • Roberts L.W.
        Ethical considerations in rural health care: a pilot study of clinicians in Alaska and New Mexico.
        Community Ment Health J. 2005; 41: 21-33
        • Handley T.E.
        • Kay-Lambkin F.J.
        • Inder K.J.
        • et al.
        Self-reported contacts for mental health problems by rural residents: predicted service needs, facilitators and barriers.
        BMC Psychiatry. 2014; 14: 249
        • Askari M.S.
        • Martins S.S.
        • Mauro P.M.
        Medication for opioid use disorder treatment and specialty outpatient substance use treatment outcomes: differences in retention and completion among opioid-related discharges in 2016.
        J Subst Abuse Treat. 2020; 114108028
        • Substance Abuse and Mental Health Services Administration
        Medicaid Coverage of Medication-Assisted Treatment for Alcohol and Opioid Use Disorders and of Medication for the Reversal of Opioid Overdose.
        Substance Abuse and Mental Health Services Administration, Rockville, MD2018
        • Substance Abuse and Mental Health Services Administration
        National Survey of Substance Abuse Treatment Services (N-SSATS): 2017. Data on Substance Abuse Treatment Facilities.
        Office of Applied Studies, Rockville, MD2018
        • Centers for Medicare & Medicaid Services, US Department of Health and Human Services
        Revisions to payment policies under the physician fee schedule and other changes to part B payment policies.
        • Henning-Smith C.
        • Kozhimannil K.
        • Casey M.
        • Prasad S.
        • Moscovice I.
        Rural-urban differences in Medicare quality outcomes and the impact of risk adjustment.
        Med Care. 2017; 55: 823-829
        • Anderson T.J.
        • Saman D.M.
        • Lipsky M.S.
        • Lutfiyya M.N.
        A cross-sectional study on health differences between rural and non-rural U.S. counties using the County Health Rankings.
        BMC Health Serv Res. 2015; 15: 441
        • Brooks R.G.
        • Walsh M.
        • Mardon R.E.
        • Lewis M.
        • Clawson A.
        The roles of nature and nurture in the recruitment and retention of primary care physicians in rural areas: a review of the literature.
        Acad Med. 2002; 77: 790-798
        • US Department of Health and Human Services
        Facing Addiction in America: The Surgeon General’s Report on Alcohol, Drugs, and Health.
        Office of the Surgeon General, Washington, DC2016