Multidimensional assessment of access to medications for opioid use disorder across urban and rural communities: A scoping review

Abstract

Background:

The majority of patients with opioid use disorder do not receive medications for opioid use disorder (MOUD), especially in rural areas. The patient-centered access to healthcare framework posits access as a multidimensional phenomenon impacted by five healthcare system and five patient ability dimensions. Interventions to improve local MOUD treatment outcomes require an understanding of how these dimensions differ across urban and rural communities. This scoping review sought to systematically appraise the literature on MOUD access across urban and rural communities (i.e., rurality) in the US using the patient-centered access framework.

Methods:

We performed a scoping review of 1) electronic databases, 2) grey literature, and 3) correspondence with content experts (March 2021). We included articles specifying the study sample by rurality and examining at least one dimension of access to MOUD. The analysis and qualitative synthesis of study results examined study characteristics and categorized key findings by access dimensions.

Results:

The search produced 3963 unique articles, of which 147 met inclusion criteria. Among included studies 96% (142/147) examined healthcare system dimensions of access while less than 20% (25/147) examined any of the five dimensions of patient ability. Additionally, 49% (72/147) of studies compared access dimensions by rurality. Across studies, increasing rurality was associated with fewer available MOUD services, but little was known about geographic variation in other critical dimensions of access.

Conclusions:

The vast majority of studies examined healthcare system dimensions of MOUD access and few studies made comparisons by rurality or prioritized the patient’s perspective, limiting our understanding of how access differs by rurality in the US. As COVID-19 spurs novel changes in MOUD delivery, this inadequate multidimen sional understanding of MOUD access may impede the tailoring of interventions to local needs. There is an urgent need for mixed-methods and community-engaged research prioritizing the patient’s perspective of MOUD access by rurality.

Registration:

Open Science Framework ( https://osf.io/wk6b9/).

Introduction

Opioid overdose deaths, especially involving synthetic opioids, continue to increase across urban and rural communities in the US, reaching an all-time high in 2020 (Hedegaard and Warner, 2020). Food and Drug Administration-approved medications for opioid use disorder (MOUD), including buprenorphine, methadone, and extended-release naltrexone (ER-naltrexone), improve treatment outcomes among individuals with OUD (Sordo et al., 2017). These medications are different pharmacologically and their use is subject to different regulatory requirements which in turn impacts accessibility within the US. Extended-release naltrexone is an opioid antagonist that is given as a monthly injectable and is administered by the prescriber. Methadone is a full opioid agonist and is only available through federally certified opioid treatment programs (OTPs). Buprenorphine is a partial opioid agonist available within general medical settings but requires a Drug Addiction Treatment Act of 2000 (DATA 2000) waiver to prescribe to more than 30 patients. Unfortunately, despite their effectiveness, only 11% of patients with OUD receive MOUD and rural residents are less likely to receive MOUD than their urban counterparts (Volkow, 2022; Williams et al., 2019).

While rural residents are less likely to receive MOUD, less is known about urban and rural differences in the factors impacting MOUD access in the US. Healthcare access refers to the ease with which individuals and communities can use appropriate services for their needs. The patient-centered access to healthcare framework organizes factors impacting access into five pairs of healthcare system accessibility and patient ability dimensions (Fig. 1 and Table 1) (Levesque et al., 2013). Urban and rural communities differ by patient characteristics, health beliefs, social factors, and health delivery systems, which may cause local variation in access to MOUD services (Kolak et al., 2020). A lack of knowledge of multidimensional variation in access to MOUD across urban and rural communities may impede the tailoring of interventions and policies to local needs and frustrate efforts to improve outcomes across the OUD treatment cascade (treatment engagement, MOUD initiation, MOUD retention, and remission) (Williams et al., 2019).

Fig. 1.

Patient-centered access to healthcare framework

Levesque, Harris, & Russell’s patient-centered access to healthcare framework includes five stages of access including the opportunity 1) to identify healthcare needs, 2) to seek healthcare services, 3) to reach, 4) to obtain, and 5) to use the appropriate health care services. Progression through each of these five stages is determined by the interaction between five paired healthcare system (top) and patient ability dimensions (bottom). Figure taken from: Levesque JF, Harris MF, Russell G. Int J Equity Health. 2013;12:18.

Table 1.

Definitions of dimensions in the patient-centered access to healthcare framework.

Healthcare system dimensions	Patient ability dimensions
Approachability refers to whether an individual can identify that MOUD services exist in an area.	Ability to perceive pertains to whether an individual can recognize a need for treatment.
Acceptability relates to the sociocultural factors surrounding MOUD that make it possible for an individual to accept or provider to prescribe MOUD.	Ability to seek refers to an individual’s autonomy and capacity to choose to seek MOUD and the type of MOUD.
Availability and accommodation refer to whether services exist in a location and can be reached both physically and in a timely manner.	Ability to reach pertains to a patient’s ability to physically reach services and may be impacted by personal mobility, availability of transportation, and occupational flexibility.
Affordability refers to the economic capacity for individuals to spend resources and time to use appropriate services and for providers to offer appropriate services.	Ability to pay describes the economic capacity of individuals to generate resources, including income or health insurance, to pay for services.
Appropriateness refers to the fit between services and patient’s needs, and may include the types of services offered and their quality.	Ability to engage refers to whether an individual can participate meaningfully in their care.

The patient-centered access to healthcare framework includes five stages of access including the opportunity 1) to identify healthcare needs, 2) to seek healthcare services, 3) to reach, 4) to obtain, and 5) to use the appropriate health care services. Progression through each of these five stages is determined by the interaction between five paired healthcare system and five patient ability dimensions. Definitions for each of the access dimensions are included above.

Studies investigating variation in access to MOUD across urban and rural communities usually only measure one dimension of access, such as availability of services, or incorporate a single perspective, such as MOUD providers (Andrilla et al., 2017; Andrilla et al., 2019; Barnett et al., 2019; Franz et al., 2021). In isolation, these studies cannot provide a comprehensive understanding of how the interaction between dimensions of MOUD access impact local access. A prior systematic review of 18 studies examined rural barriers to MOUD access from the patient and provider perspective; however, this review only included studies with a key term specifying rurality and did not include grey literature (i.e. non-peer reviewed studies) or elicit studies from content experts (Lister et al., 2020). To our knowledge, no studies have comprehensively reviewed and synthesized research on MOUD access across urban and rural communities in the US, nor have they included studies on all dimensions of access from both the healthcare system and patient perspective. Therefore, we performed a scoping review using the patient-centered access to healthcare framework to inform novel opportunities to address geographic disparities in MOUD access.

Methods

Study overview

We conducted a scoping review in accordance with the PRISMA Extension for Scoping Reviews and the Joanna Briggs Institute Manual for Evidence Synthesis (Peters et al., 2020; Peters et al., 2021; Tricco et al., 2018). Prior to beginning the search, the review protocol was registered with Open Science Framework (https://osf.io/wk6b9/).

Patient-centered access framework

The patient-centered access framework places the patient at the center of the treatment-seeking process. The framework includes five stages of access including the opportunity to 1) identify healthcare needs, 2) seek healthcare services, 3) reach, 4) obtain, and 5) use the appropriate health care services. Progression through each of these five stages is determined by the interaction between five paired healthcare system and patient ability dimensions (Levesque et al., 2013). This framework was previously used to examine access to maternal and child care, mental health, and infectious disease treatment services and to compare access across urban and rural communities (Cu et al., 2021; Cyr et al., 2019).

Study eligibility

We included peer-reviewed and grey literature reporting observational, randomized controlled trial, qualitative, and mixed methods studies that specified the study population by rurality and examined at least one of the five healthcare system or five patient ability dimensions of access to one or more MOUD. We included all measures of rurality including multi-level categorical (e.g., rural, suburban, and urban) and continuous measures. We also included studies of receipt of MOUD, even if they did not assess a specific dimension of access. We excluded studies 1) not in English or outside the US; 2) examining MOUD for only withdrawal management; or 3) only examining retention in treatment. We limited our review to the US because the variation in MOUD regulatory policies and healthcare systems across nations make direct comparison and interpretation difficult.

Data sources and search strategy

Our search strategy included electronic databases, grey literature, and correspondence with content experts and was developed by an experienced research librarian (MCF). On January 6, 2022 we searched: MEDLINE, Embase, and APA Psycinfo on the Ovid database; Cochrane Library; Web of Science (Core Collection); CINAHL Complete (EBSCO); and PubMed Central (PMC). To maximize sensitivity, our formal search used controlled vocabulary terms and synonymous free-text words to capture the concepts of “MOUD” and “rural or urban settings in the United States” (eMethods 1). Our search strategy was peer reviewed by a second independent librarian. Searches were limited to the English language and no date restriction was applied.

Next, two authors (MJ and TB) conducted the grey literature search, which consisted of both a structured Google search and a hand search of relevant websites identified by the research team. For the structured Google search, standardized search terms (eMethods 2) were used, and the review of articles stopped after 20 consecutive results did not meet inclusion criteria. The first author (TB) also completed a hand search of websites (eMethods 2).

Next, we contacted 21 subject matter experts to identify additional articles. Experts were either commonly cited during the database search or were identified by the research team as experts. Finally, additional articles were selected through citation chaining by reviewing the reference lists of included studies.

Search results were pooled in EndNote (Clarivate Analytics. EndNote X9) and duplicates removed. This set was uploaded to Covidence (Veritas Health Innovation. Covidence Systematic Review Software) for screening.

Study selection

Two reviewers (TB and MJ) independently screened all titles and abstracts and removed articles not meeting inclusion criteria. The same two reviewers screened the full text of remaining articles. Each record was screened by both reviewers separately. Disagreements at both stages were resolved by a third author (PJ).

Data extraction

Two authors (TB and MJ) completed all data extraction using a tool created in Qualtrics XM and reviewed by all authors (eMethods 3). Extracted items included study characteristics (year, design, location, sample size, sample characteristics), definition of rurality, medication types, data source, and key findings related to dimensions of access. To ensure reliable extraction, two authors (TB and MJ) first extracted data from a common set of 20 articles and reviewed agreement. Then, the two authors extracted data from half of the articles each, and subsequently reviewed the data of the other half to ensure agreement. Consistent with scoping review methods, we did not use a standardized appraisal tool to assess the quality of studies (Peters et al., 2021).

Synthesis of results

First, we performed descriptive analyses of the quantitative study characteristics. Next, we qualitatively synthesized results across studies to appraise the scope of the literature. We organized our synthesis first by describing studies that only examined receipt of MOUD followed by the two paired dimensions (one healthcare system and one patient ability dimension) for each of the five stages of access.

Results

Characteristics of studies

Our search produced 3963 unique articles, of which 147 met the inclusion and exclusion criteria (Fig. 2). Studies were published between 1975 and 2022 (Table 2), with 2019 the median year of publication (e Fig. 1). Studies were most frequently cross-sectional (73%; 108/147), secondary quantitative analyses (52%; 77/147) among individuals (60%; 88/147) or counties (16%; 24/147). Approximately half of studies were conducted in rural (28%; 42/147) or urban settings only (20%; 30/147) and nearly half (49%; 72/147) made access comparisons by rurality. Most studies examined buprenorphine treatment access (88%; 129/147), and a minority of studies (18%; 27/147) examined all three medications. Only three studies examined access to MOUD for racial/ethnic minorities across rurality. The complete characteristics of all included studies are listed in eTables 1 and 2.

Fig. 2.

PRISMA flow diagram for scoping review of medications for opioid use disorder access across urban and rural communities

From: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Table 2.

Characteristics of eligible studies of access to medications for opioid use disorder across urban and rural communities.

Study characteristic		Eligible Studies (n = 147)
Study year (median, IQR)		2019 (2018–2020)
Study design, n (%)
	Cross sectional	108 (73%)
	Qualitative	19 (13%)
	Retrospective cohort/Case control	16(11%)
	Prospective cohort	0
	Randomized controlled trial	2 (1%)
	Mixed methods	1 (1%)
	Quasi-experimental	1 (1%)
Data source, n (%)
	Primary quantitative	48 (33%)
	Secondary quantitative	77 (52%)
	Primary qualitative	18 (12%)
	Mixed methods	4 (3%)
Primary unit of analysis, n (%)
	Individual	88 (60%)
	Healthcare facility	18 (12%)
	Zip code	2 (1%)
	City/Town	1 (1%)
	County	24 (16%)
	Census tract/Census block group	7 (5%)
	State	1 (1%)
	Other	7 (5%)
Setting, n (%)
▒Rural only		42 (28%)
▒Urban only		30 (20%)
▒Urban-rural binary		14 (10%)
▒Multilevel		61 (41%)
Compare access across rurality		72 (49%)
Definition of rurality, n (%)
	Rural-urban continuum codes (RUCC)	24 (16%)
	Rural-urban commuting areas (RUCA)	14 (9%)
	Urban influence codes (UIC)	9 (6%)
	US census bureau	10 (7%)
	OMB for counties	6 (3%)
	NCHS Urban-Rural Classification Scheme	16 (11%)
	Geographic isolation scale	0
	Self-report	4 (3%)
	Not given	62 (42%)
	Other	2 (1%)
Medications, n (%)	Buprenorphine	129 (88%)
	Methadone	77 (52%)
	ER-Naltrexone	33 (22%)
	All three medications	27 (18%)

IQR = interquartile range; OMB = Office of Management and Budget; NCHS = National Center for Health Statistics

Of the five healthcare system dimensions (Table 3), availability and accommodation (54%; 79/147) was examined most frequently, followed by acceptability (24%; 36/147). The patient ability dimensions were less frequently examined, with ability to engage (7%; 10/147) being the most examined. No studies assessed ability to seek.

Table 3.

Dimensions of access examined among eligible studies (n = 147).

	Eligible studies (n = 147)
Dimensions of access, n (%)
Global measure of access *
Health system accessibility	19 (13%)
Approachability	4 (3%)
Acceptability	36 (24%)
Availability and Accommodation	79 (54%)
Affordability	11 (7%)
Appropriateness	12 (8%)
Patient ability
Ability to perceive	4 (3%)
Ability to seek	0
Ability to reach	5 (3%)
Ability to pay	6 (4%)
Ability to engage	10 (7%)
Number of dimensions measured by each study, n (%)
Measure of receipt *	19 (13%)
1	107 (73%)
2	14 (9%)
3	9 (6%)
4 or greater	3 (2%)

^*Global measure of receipt indicates studies that only assessed whether individuals received medications for opioid use disorder and did not measure a dimension of access

Qualitative synthesis

Studies measuring receipt of MOUD

Receipt of MOUD was commonly assessed using administrative or insurance claims data. These studies found that across rurality, urban residents, compared to rural residents, were more likely to receive MOUD in both outpatient (Finlay et al., 2020) and inpatient settings (Soares et al., 2020). Studies consistently found that rurality impacted the type of MOUD received; rural residents were more likely to receive buprenorphine versus methadone, relative to urban residents (Manhapra & Quinones, 2016; Stein et al., 2012).

Opportunity to identify needs: approachability and ability to perceive

Four studies examined the approachability (e.g., whether services are identifiable) of MOUD services but only one made comparisons by rurality. This study found DATA 2000 waivered buprenorphine providers in rural communities were more likely to opt into appearing within the publicly-available SAMHSA database than urban providers (Nguyen et al., 2021).

Four studies reported on patients’ ability to perceive a need for OUD treatment, including two that compared individuals by rurality. However, none of these studies measured need for medication treatment specifically. These studies used data from the National Survey on Drug Use and Health (NSDUH) and found that rates of self-reported treatment need among individuals with OUD were low (10.5–13.4%) and did not differ by rurality (Borders, 2018; Saini et al., 2022).

Opportunity to seek services: acceptability and ability to seek

Acceptability (e.g., sociocultural fit) of MOUD services was a frequently studied dimension. Acceptability studies examined patient and provider attitudes about MOUD, provider reported barriers to prescribing MOUD, and community attitudes and beliefs about MOUD. Acceptability was assessed primarily through surveys and qualitative interviews with patients and providers. A minority of studies (17%; 6/36) made comparisons by rurality.

Surveys of provider attitudes about buprenorphine in emergency department and office-based settings frequently measured perceptions of effectiveness and interest in adoption; these studies found no differences by rurality (Lundgren et al., 2011; Zuckerman et al., 2020). One cross-sectional survey assessing physician bias found higher levels of unfavorable attitudes towards patients with OUD among rural versus urban physicians (Franz et al., 2021).

Several qualitative studies assessed office-based providers’ perceived barriers to prescribing buprenorphine (Andrilla et al., 2017; DeFlavio et al., 2015). One survey found concerns about MOUD diversion, lack of experience with MOUD, and lack of mental health providers and specialty support to be more common among providers in rural communities (Andrilla et al., 2020).

Few studies examined the acceptability of MOUD from the patient and community perspective (Beachler et al., 2021). In one survey of adults in rural states, 59% of respondents reported supporting MOUD while 50% supported incarceration for people who use drugs (Durantini, Grid for the Reduction of & Albarracin, 2021). Other qualitative studies assessed patient attitudes toward methadone in largely urban places (Rubio, 2013; Stancliff et al., 2002), finding that concerns of being labeled an “addict” and lack of community support for MOUD were seen as barriers to starting and continuing methadone (Scorsone et al., 2020). A few, largely rural, studies focused on acceptability in special populations, including pregnant individuals (Mattocks et al., 2017; Mullins et al., 2019) and individuals involved with the criminal justice system (Bunting et al., 2018; Matusow et al., 2013). Notably, no studies examined specific barriers related to acceptability within Native American or Indigenous populations, such as the presence of treatment options that align with their medical beliefs/practices. In addition, no studies compared patient or community acceptability by rurality or examined the acceptability of ER-naltrexone. In contrast to acceptability, our search did not produce any studies of an individuals’ ability to seek MOUD services (e.g., autonomy and capacity to choose to seek services).

Opportunity to reach: availability and accommodation and ability to reach

Availability and accommodation (e.g., location and timeliness) was the most studied dimension of access, and over half (52%; 41/79) made comparisons by rurality. Availability and accommodation was the only dimension frequently assessed using a multilevel measure of rurality, rather than dichotomous. The majority (61%; 48/79) of studies specifically looked at the geographic availability of MOUD services using one of three methods: 1) the density of providers/facilities per geographic unit (Abraham et al., 2019; Andrilla et al., 2019; Barnett et al., 2019; Haffajee et al., 2019), 2) the travel cost (distance, drive/travel time to nearest MOUD services) (Joudrey et al., 2019; Kleinman, 2020; Langabeer et al., 2020), and 3) gravity models (measures the overlap of spatial demand for and supply of services) (Amiri et al., 2021; Cao et al., 2019; Hyder et al., 2021). The density method was the most commonly used (67%; 32/48), followed by travel cost (27%; 13/48), and gravity models (6%; 3/48).

Waivered buprenorphine providers and OTPs (i.e., federally certified methadone clinics) were both more likely to be located in urban versus rural communities (Abraham et al., 2019; Andrilla & Patterson, 2021; Andrilla et al., 2019; Barnett et al., 2019; Haffajee et al., 2019). Similarly, studies consistently found travel costs to buprenorphine and methadone services increased with increasing rurality (Joudrey, Edelman et al., 2019; Kleinman, 2020; Langabeer et al., 2020). Few studies examined the availability of ER-naltrexone.

Other components of availability and accommodation were less frequently assessed. An audit study found rural buprenorphine providers were less likely to be accepting new patients relative to urban providers (Beetham, Saloner, Wakeman, Gaye & Barnett, 2019). A study examining the timeliness of MOUD services (i.e. the time from initial engagement to the receipt of the medication) in the Veterans Health Administration found rural veterans were less likely to receive MOUD within 30 days of a new treatment episode (Wyse et al., 2019). In a study of publicly-funded urban OTPs, females (compared to males) and Black patients (compared to White patients) spent longer on the treatment wait list (Marsh et al., 2021). Other studies found a large percentage of waivered buprenorphine providers, over half in one study (Andrilla et al., 2018), were not currently prescribing buprenorphine, with no difference by rurality (Jones & McCance-Katz, 2019). Finally, two recent audit studies found that the majority of pharmacies were not immediately able to dispense buprenorphine, with no difference by rurality (Hill et al., 2021; Kazerouni et al., 2021).

Comparatively few studies examined a patient’s ability to reach MOUD services (Huhn et al., 2017). One study found that rural Medicaid beneficiaries with OUD traveled a median of 49 miles to reach the closest buprenorphine or ER-naltrexone provider and longer distances were associated with lower rates of initiation and retention in treatment (Cole et al., 2019). Three qualitative studies found rural individuals with OUD reported a lack of personal or public transportation as significant barriers to MOUD (Bunting et al., 2018; Kane et al., 2020; Kramlich et al., 2018). No studies compared patient’s ability to reach MOUD services by rurality, including variation in personal mobility, availability of public or private transportation, occupational flexibility, or availability of child care.

By modeling both spatial demand for and supply of services, studies using gravity models attempt to account for both availability and patient’s ability to reach services. These studies have consistently found that MOUD treatment shortages are more likely within increasing rurality, but have declined for all urban-rural classifications in recent years (Cao et al., 2019; Dick et al., 2015; Grimm, 2020). No gravity model studies included patient-reported data on ability to reach services; rather, studies relied on general population data to model ability to reach (Dick et al., 2015; Iloglu et al., 2021).

Opportunity to obtain or use services: affordability and ability to pay

Affordability of services was most frequently assessed by phone audit or provider surveys and primarily examined the type of insurance accepted by providers, consistently finding that rural buprenorphine providers were more likely to accept Medicaid than urban providers (Andrilla et al., 2020; Bedrick et al., 2020; Parran et al., 2017; Sorrell et al., 2020).

A patient’s ability to pay was primarily assessed by whether an individual had health insurance and/or appropriate income. Studies found that obtaining Medicaid, especially through Medicaid expansion, facilitated obtaining MOUD services in rural communities (Cole et al., 2019; Kane et al., 2020; Marsh et al., 2021). No studies compared insurance coverage by rurality among individuals with OUD.

Opportunity to be offered services: appropriateness and ability to engage

Commonly-studied concepts related to appropriateness (e.g., fit between services and patient needs) included whether providers offered wrap-around services (e.g., medical and social services), mental health services, and specialty consultation. Studies consistently found MOUD was more likely to be delivered by general practitioners and advance practice providers in rural communities, relative to specialists in urban communities (Andrilla et al., 2017; Andrilla et al., 2020). A cross-sectional study of measures of appropriateness found no evidence that prescribing practices or quality of care (measured by induction practices, frequency of visits, and dosing patterns) were different in urban and rural communities (Lin & Knudsen, 2019). Two studies examined how the COVID-19 pandemic changed patients’ engagement patterns with MOUD, finding that tele-buprenorphine was being increasingly used (Hughes, Verrastro, Fusco, Wilson & Ostrach, 2021) and OTPs were increasingly offering new services (i.e. curbside treatment or telehealth) with no difference between rural and urban communities (Cantor & Laurito, 2021).

In regards to a patient’s ability to engage (e.g., meaningful participation in care) in MOUD services, qualitative studies identified several facilitators in both urban and rural communities, including feeling included in decision-making (Kramlich et al., 2018) and self-help groups/peer support systems (Kane et al., 2020; Scorsone, 2020). Similarly, barriers to engagement included lack of social support (Scorsone, 2020), daily dosing of methadone (Kane et al., 2020), and inability to meet basic needs such as housing (Hawk et al., 2020). However, no studies compared how measures of engagement or program requirements differ by rurality.

Discussion

This scoping review systematically appraised the literature on patient-centered dimensions of access to MOUD across urban and rural communities in the US, finding only two dimensions of access (acceptability and availability) were frequently examined and few studies compared access by rurality or prioritized the patient’s perspective of abilities. Across multiple studies, increasing rurality was associated with fewer available MOUD services and longer travel costs to services, but little was known about geographic variation in other critical dimensions of MOUD access. The current literature was reflective of a provider-centric approach, with 96% of studies examining healthcare system dimensions of access and less than 20% examining patient ability dimensions. Together, these results demonstrate significant knowledge gaps, hindering the tailoring of MOUD services to the local needs of patients in the context of the ongoing overdose epidemic.

Our results are consistent with a prior systematic review of rural-specific barriers to MOUD, which found that studies frequently examined availability of MOUD and few examined access to ER-naltrexone (Lister et al., 2020). Our review expands upon these findings by using a multidimensional framework of access examining dimensions across rurality, resulting in a much larger number of studies (147 vs. 18 studies). While we observed a consensus across studies that increasing rurality was associated with lower MOUD availability and lower rates of MOUD receipt, the extent of a causal relationship between availability and receipt of MOUD remains less clear because no studies examined availability and receipt while accounting for multiple dimensions of access. Our review found evidence that other dimensions of access vary by rurality and these may confound or modify the relationship between availability and receipt of MOUD. Evidence within the current literature, such as rural providers being less likely to accept new patients or provide timely buprenorphine initiation or the low acceptability of methadone within predominately urban communities (Beetham, Saloner, Wakeman, Gaye & Barnett, 2019; Stancliff et al., 2002; Wyse et al., 2019), suggests the dominant barriers to MOUD access vary locally.

Our finding that the majority of studies (88%) examined access to buprenorphine while only 18% examined access to all three FDA approved medications may reflect the prioritization of buprenorphine given it is the only opioid agonist treatment available outside of specialty treatment settings within the US. Not all patients are retained in buprenorphine treatment (O’Connor et al., 2020), and it is important to note that access to buprenorphine alone is not sufficient to ensure quality OUD treatment services. Given differences in pharmacology, delivery, and patient preference, all three MOUDs should be accessible in communities to facilitate treatment individualization, respect for patient autonomy, and to maximize retention. Research on MOUD access should reflect and reinforce this standard (National Academies of Sciences, Engineering, and Medicine, 2019).

Incorporating knowledge of patient abilities has previously been used to improve mental health services for individuals with depression and anxiety (Packness et al., 2019; Roberge et al., 2014) and improve access to chronic disease care (Cu et al., 2021). Our results showing limited knowledge of patient abilities across communities suggests a missed opportunity to similarly address novel factors impeding MOUD access. Little was known about geographic differences in patients’ perceptions of need for services (ability to perceive), preference or choice of MOUD (ability to seek), ability to travel by private or public transportation (ability to reach), insurance coverage (ability to pay), or need for wrap around services to mitigate treatment barriers (ability to engage). This gap in knowledge prevents examination of potential interactions between healthcare services and patient reported abilities, such as local differences in availability of and patient preferences for different modes of transportation to reach MOUD services.

Our study attempted to go beyond the rural-urban dichotomy and examine dimensions of access across the spectrum of rurality. While several studies have used multilevel measures of rurality, particularly when examining availability and accommodation, there were few studies examining how access differs using more granular categorizations of rurality, such as suburban areas. Our results suggest this represents a significant limitation for the field. There is marked variation in community characteristics among rural or urban communities and research limited to a binary definition is likely insufficient to enable the tailoring of services to local needs.

Increasing community partnerships to improve dissemination and implementation of evidence-based research was part of the National Institute on Drug Abuse 2016–2020 strategic plan and our results demonstrate an ongoing urgent need for community-engaged research prioritizing the patient’s perspective of MOUD access to achieve this strategic goal (NIDA, 2015). Such partnerships would provide an opportunity for local small area data collection that would better position studies to understand more granular differences between geographic areas.

Telehealth and mobile methadone units are examples of innovations which may mitigate traditional barriers to MOUD across urban and rural communities (Wang et al., 2021). While our findings highlighted increasing telehealth utilization for MOUD, especially in rural communities (Cantor & Laurito, 2021; Hughes, Verrastro, Fusco, Wilson & Ostrach, 2021; Wang et al., 2021), it is critical that effectiveness and implementation research incorporate patient perspectives to optimize these innovations to local conditions and avoid exacerbating existing inequalities in MOUD access.

Several groups at high risk of overdose across urban and rural communities were infrequently represented within the literature. We found only three studies on Native American or Indigenous populations within rural areas, where overdose rates are particularly high (Mpofu et al., 2021). Similarly, few studies adequately addressed the intersection of geography and access to MOUD for racial/ethnic minorities, sexual and gender minorities, and individuals within the criminal justice system (Goedel et al., 2020; Joudrey, Khan, et al., 2019). Without inclusion or stratification of these groups in study samples, it is difficult to characterize demographic inequities across urban and rural communities.

Limitations

There are limitations to this scoping review. First, our search may not have identified all relevant studies; notably some ethnographic work may not have been captured by our search terms. However, we took several steps, including contacting experts and searching the grey literature to minimize this possibility. Second, because we excluded studies outside the US, our results do not generalize to international settings. Finally, studies that did not find rural-urban differences may have been less likely to be published due to publication bias.

Conclusions

Our systematic appraisal of the literature found only acceptability and availability of MOUD services were frequently studied dimensions of access across urban and rural communities and studies prioritizing patients’ perspectives were broadly lacking. Studies prioritizing a multidimensional and patient-centered approach to access are needed to inform local interventions for expanding MOUD access. As the COVID-19 pandemic spurs rapid development of innovative models of MOUD delivery, a focus on ameliorating these research gaps will help inform improvements in access across urban and rural communities.

Supplementary Material

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.drugpo.2022.103931.