Willingness to utilize a mobile treatment unit in five counties at the epicenter of the US rural opioid epidemic

Abstract

Introduction:

US federal policies are evolving to expand the provision of mobile treatment units (MTUs) offering medications for opioid use disorder (MOUD). Mobile MOUD services are critical for rural areas with poor geographic access to fixed-site treatment providers. This study explored willingness to utilize an MTU among a sample of people who use opioids in rural Eastern Kentucky counties at the epicenter of the US opioid epidemic.

Methods:

The study analyzed Cross-sectional survey data from the Kentucky Communities and Researchers Engaging to Halt the Opioid Epidemic (CARE2HOPE) study covering five rural counties in the state. Logistic regression models investigated the association between willingness to utilize an MTU providing buprenorphine and naltrexone and potential correlates of willingness, identified using the Behavioral Model for Vulnerable Populations.

Results:

The analytic sample comprised 174 people who used opioids within the past six months. Willingness to utilize an MTU was high; 76.5% of participants endorsed being willing. Those who had recently received MOUD treatment, compared to those who had not received any form of treatment or recovery support services, had six-fold higher odds of willingness to use an MTU. However, odds of being willing to utilize an MTU were 73% lower among those who were under community supervision (e.g., parole, probation) and 81% lower among participants who experienced an overdose within the past six months.

Conclusions:

There was high acceptability of MTUs offering buprenorphine and naltrexone within this sample, highlighting the potential for MTUs to alleviate opioid-related harms in underserved rural areas. However, the finding that people who were recently under community supervision or had overdosed were significantly less willing to seek mobile MOUD treatment suggest barriers (e.g., stigma) to mobile MOUD at individual and systemic levels, which may prevent improving opioid-related outcomes in these rural communities given their high rates of criminal-legal involvement and overdose.

1. INTRODUCTION

In 2021, more than 107,000 US residents died of drug overdose, with approximately 80,000 of these deaths involving opioids (National Vital Statistics System, 2022). Medications for opioid use disorder (MOUD), such as methadone, buprenorphine, and extended-release naltrexone, are effective in treating opioid use disorder (OUD) (Hser et al., 2016; Lim et al., 2022; Weiss et al., 2015). Methadone and buprenorphine specifically reduce the risk of fatal opioid overdose (Krawczyk, Mojtabai, et al., 2020; Wakeman et al., 2020) and the spread of infectious diseases like HIV by decreasing injection drug use and increasing access to preventive measures and adherence to antiretroviral medications (Metzger et al., 2010). MOUD are vital interventions to reduce the burden of opioid-related harms, but they have not expanded adequately to rural areas in the US during the opioid epidemic. Rural areas have more limited geographic access to certified opioid treatment programs (OTPs) compared to urban areas: driving time to the nearest OTP is approximately 49 minutes greater in rural versus urban areas (Kiang et al., 2021). Poor geographic access remains even after accounting for willingness to travel to fixed-site (i.e., non-mobile) OTPs (Mitchell et al., 2022).

In response to continued increases in opioid-related harms and poor treatment access in underserved areas, the US Substance Abuse and Mental Health Services Administration (SAMHSA) issued a Notice for Proposed Rulemaking in December 2022 advancing several changes to MOUD treatment policies currently outlined in Title 42 Part 8 of the Code of Federal Regulations (42 CFR Part 8) (SAMHSA, 2022). Changes include increasing the geographic reach of OTPs by promoting their operation of mobile treatment units (MTUs). MTUs are a potentially effective strategy to increase geographic access to MOUD in rural areas because they can travel to places where people who use opioids (PWUO) live and work. They typically operate through vans staffed with OTP-affiliated providers who prescribe and dispense MOUD and also offer counseling alongside a range of other harm reduction services (e.g., naloxone, sterile syringes).

To date, almost all research on mobile MOUD programs in the US has been limited to cities. These studies have demonstrated that MTUs are associated with higher treatment retention compared to fixed treatment sites (Greenfield et al., 1996; Krawczyk et al., 2019; Stewart et al., 2021) and have high acceptability among PWUO (Regis et al., 2020; Wenzel & Fishman, 2021). Studies in urban areas found that people who seek MOUD from MTUs were more likely to be Black or African American, Latine/-x, or unhoused and have a history of riskier substance use behaviors (e.g., injection drug use) and inconsistent engagement with fixed-site treatment (Hall et al., 2014; Iyengar et al., 2019; Krawczyk et al., 2019; O’Gurek et al., 2021). These findings indicate that MTUs reach particularly marginalized subgroups, but studies have largely neglected PWUO in rural, underserved areas.

Although policy changes allowing expanded MTUs are intended to address the MOUD treatment gap in rural areas, researchers know little about the acceptability of these programs, if established, among rural PWUO. Research on MTUs based in urban settings may not generalize to rural areas due to underlying differences in geographic access to treatment (Kiang et al., 2021; Mitchell et al., 2022) and individual- and community-level stigma (Beachler et al., 2021; Ibragimov et al., 2021; Richard et al., 2020). Evidence for the effectiveness of MTUs providing MOUD in rural areas is nascent yet promising: one study based in rural Maryland found that an MTU administering buprenorphine reduced opioid use among participants by approximately 33% compared to baseline and reduced average driving distance to a provider by 10 minutes (Weintraub et al., 2021).

Importantly, opioid-related harms (e.g., overdoses) and barriers to MOUD access are compounded for individuals—perhaps including those in rural areas—who have a history of criminal-legal system (CLS) involvement (Krawczyk, Schneider, et al., 2020). Reported barriers to MOUD uptake among CLS-involved individuals in urban areas have included a fear of law enforcement surveillance and harassment and concerns with increased visibility at fixed-site OTPs (Grella et al., 2020; Vail et al., 2021). Addressing these barriers is important, as studies have demonstrated that MOUD reduce the risk of overdose deaths among CLS-involved PWUO (Mace et al., 2020). In England, a corrections-based buprenorphine and methadone treatment program decreased deaths by 85% (Marsden et al., 2017) and the expansion of MOUD access within the Rhode Island Department of Corrections in the US decreased deaths by 61% among PWUO after release (Green et al., 2018). Promisingly, studies have highlighted the acceptability of MTUs offering buprenorphine among CLS-involved individuals in urban settings in the US (Krawczyk et al., 2019; Regis et al., 2020). Investigating acceptability among rural PWUO is especially critical as they face even higher rates of CLS involvement than their counterparts in urban areas (Weiss Riley et al., 2018).

To understand whether the expansion of MTUs offering MOUD in rural areas as promoted by SAMHSA might be acceptable, this study aims to explore willingness to utilize an MTU among a sample of PWUO in rural Eastern Kentucky, an Appalachian region at the epicenter of the US rural opioid epidemic (Schalkoff et al., 2020). Correlates of willingness to use an MTU, including a recent history of CLS involvement, are explored along with other factors that may shape treatment access in practice, as guided by Penchansky and Thomas (1981) who posited that treatment acceptability is a critical dimension of access. The Behavioral Model for Vulnerable Populations (BMVP) informed the selection of correlates (Gelberg et al., 2000). BMVP, derived from Andersen’s Model of Healthcare Utilization (Andersen & Newman, 2005), is a valuable heuristic for identifying factors that influence treatment access among particularly marginalized individuals (Gelberg et al., 2000). Several studies have used BMVP to investigate the correlates of treatment access specific to rural settings (Romo et al., 2018; Victor et al., 2018) and CLS-involved PWUO (Oser et al., 2016).

2. METHODS

2.1. Study Design and Procedures

The current study is a cross-sectional analysis of data from one wave of the longitudinal Gateway2Health project, a part of the Kentucky Communities and Researchers Engaging to Halt the Opioid Epidemic (CARE2HOPE) study that covers five counties in rural Eastern Kentucky. CARE2HOPE is part of the National Institute on Drug Abuse’s Rural Opioid Initiative, a consortium of eight studies across 10 states developed to 1) investigate the opioid epidemic and its impacts at the local level in rural settings and 2) inform interventions addressing opioid- and injection drug use-related harms (Jenkins et al., 2022).

From 2018 to 2020, CARE2HOPE staff used respondent-driven sampling (RDS) to recruit participants from the five counties into a cohort study. Staff recruited seeds through community outreach (e.g., cookouts, neighborhood walks), flyer distribution at sites frequented by PWUO, and invitation of participants from a previous study who had consented to be contacted regarding future research (Young et al., 2020). Seeds could refer up to three peers ($10 per eligible peer that enrolled), who could in turn refer three peers, and so on until the study reached the desired sample size. Eligibility criteria for the CARE2HOPE cohort were: (1) being 18 years of age or older; (2) residing in one of the five counties in the study area; and (3) injecting any drug or using any opioid to get high through any administration route in the past 30 days. The baseline CARE2HOPE cohort comprised 338 participants, 81.4% of whom reported being given a coupon or code by another participant in an RDS chain, 12.1% were referred by someone but not given a coupon or code, 6.5% through advertisement or outreach at cookouts and community venues, 2.1% by invitation from staff based on the participants’ consent in prior studies to be contacted about future research, and 0.3% through social media.

After enrollment, study staff administered computer-assisted surveys at baseline and every six months thereafter for eight data collection waves in total. At the 12-month follow-up wave, staff added survey questions about the acceptability of and staffing and operations preferences for MTUs in response to increased community stakeholder interest in MTUs. In total, 241 participants completed this wave between February 2019 and January 2021. Because this analysis concerned MOUD, the analytic sample (n=174) included only those who reported using opioids in the past six months to get high at the 12-month wave. The sample excluded participants who had exclusively used drugs other than opioids, had not used drugs at the time that the mobile unit questions were administered, or did not complete this survey.

2.2. Measures

The outcome of interest was willingness to utilize an MTU providing buprenorphine and naltrexone, as the federal government did not authorize programs to provide methadone through MTUs until June 2021. Surveys asked participants, “If a new treatment program offering medication for opioid addiction was going to be created here and operated out of a mobile unit that made visits throughout the community, would you consider using it? This program would offer buprenorphine…naltrexone shots…or buprenorphine shots…offer brief counseling and referrals to other forms of drug treatment, and provide other health services.” Response options were ‘No,’ ‘Yes,’ ‘Maybe,’ and ‘Don’t Know.’ Responses were dichotomized, with ‘Yes’ (74.9%) coded as explicitly willing and other responses (‘No’: 13.7%, ‘Maybe’: 8.6%, ‘Don’t Know’: 0.6%) coded as not willing. Surveys then prompted those who had answered ‘No’ to select one or more of the following reasons why they would not use an MTU: ‘It would not give me enough privacy,’ ‘Mobile units don’t provide enough security,’ ‘A mobile unit would offer too few services,’ ‘I would be afraid that law enforcement would see me going in and out of it,’ ‘I would not want other people in the community to see me going in and out,’ and ‘Other.’ The CARE2HOPE study team selected response options based on barriers to accessing other harm reduction services (Lancaster et al., 2020) and revised options as needed with input from the field staff. In addition, surveys asked participants about preferences for MTU operations (e.g., parking locations, hours of operation) and staffing.

This study used BMVP to select correlates of willingness to use MTUs. BMVP constructs include predisposing characteristics (i.e., demographic and other preexisting factors), enabling characteristics (i.e., factors that enable access to and utilization of care), and needs characteristics (i.e., perceived and actual need for care). With the exception of perceived severity of opioid dependence, which surveys assessed only at baseline, all data on predisposing, enabling, and needs characteristics were obtained in the same follow-up survey at the 12-month timepoint during which the participants answered the MTU questions.

This study classified CLS involvement, the primary correlate of interest, as a predisposing characteristic using three separate yet non-mutually-exclusive dichotomous variables: a past-six-month history of arrest; incarceration; and probation, parole, supervised release, or other form of community supervision. Guided by BMVP, other predisposing characteristics included age, gender, and monthly income. Analyses did not include race and ethnicity measures, as 97.9% of participants identified as White non-Hispanic.

Enabling characteristics included the following dichotomous variables: access to transportation to medical appointments, health insurance status, and past-six-month history of syringe service program (SSP) utilization. A categorical measure of past-six-month history of substance use disorder (SUD) treatment and recovery support services captured (1) an absence of treatment or recovery support services of any kind, (2) treatment or services involving self-help groups (e.g., Narcotics Anonymous), outpatient counseling, inpatient or residential treatment, supervised detoxification, and staying in a sober house, and (3) methadone, buprenorphine, or naltrexone treatment, as well as any treatment modalities or services described in (2).

Needs characteristics were perceived severity of opioid dependence, missing an appointment due to lack of transportation in the past six months, having a current HCV diagnosis, and experiencing an overdose in the past six months. All measures of need, with the exception of perceived severity of opioid dependence, were dichotomous. Perceived severity of opioid dependence, collected only at baseline, was a continuous measure summing item scores of the Severity of Dependence Scale (Gossop et al., 1995). Scores ranged from 0–15, with a score of 5 or more indicating opioid dependence and higher scores indicating higher perceived severity of dependence (Gossop et al., 1995; Iraurgi Castillo et al., 2010).

2.3. Analysis

Descriptive statistics characterized central tendencies and dispersions for all variables. A multivariable logistic regression model with the three CLS variables (i.e., arrest, incarceration, and community supervision), adjusting for the remaining predisposing, enabling, and needs covariates, generated adjusted odds ratios. The multivariable model used robust standard errors to account for non-independent observations due to RDS. Because perceived severity of opioid dependence, recent SSP and SUD treatment utilization, HCV diagnosis, and experiencing an overdose in the past six months may be potential mediators between the predisposing characteristics and outcome, we conducted a sensitivity analysis by omitting these variables and comparing results. Results did not substantially change, and the final adjusted model retained these covariates. Approximately 10% of participants had missing data on one or more correlates selected for analysis. For all models, multiple imputations by chained equations (m=100 imputations) accounted for bias due to potential missingness.

3. RESULTS

3.1. Sample Characteristics

After including only participants who reported using opioids to get high in the past six months at the 12-month time point, the sample comprised 174 participants. This sample did not differ from the CARE2HOPE cohort at baseline with respect to covariates included in the models. Participants had an average age of 36.4 years (SD: 8.4 years), and 55.8% identified as male while the remaining 44.3% identified as female (Table 1). The average monthly income was $702.44 (SD: $1,238.59). Most participants (58.1%) reported having been arrested, 42.8% had been incarcerated, and 21.4% had been under community supervision in the past six months. With regards to enabling characteristics, 61.3% had access to transportation to medical appointments, 83.0% had health insurance or other coverage, and the majority (69.8%) had not received any SUD treatment while only 20.9% had taken MOUD in the past six months. Lastly, within the needs characteristics domain of BMVP, the average score for perceived severity of opioid dependence (range: 0–15) was 6.8 (SD: 4.8), indicating that the average participant was dependent on opioids. Table 1 displays descriptive data for the remaining correlates.

Table 1.

Sample characteristics (n=174)

Variablea	Mean (SD)	n (%)
Outcome
Willing to use mobile treatment unit	130 (76.5)
Predisposing characteristics
Age	36.4 (8.4)
Gender
Male		97 (55.8)
Female		77 (44.3)
Monthly income ($)	702.44 (1238.59)
Criminal legal system involvementb
Arrest		100 (58.1)
Incarceration		74 (42.8)
Community supervision (e.g., parole)		37 (21.4)
Enabling characteristics
Has transportation to medical appointments		106 (61.3)
Has health insurance or other coverage		142 (83.0)
SSP visitb		85 (49.1)
SUD treatment or recovery supportb
None		120 (69.8)
Treatment or recovery support without MOUDc		16 (9.3)
MOUD +/− other treatment or recovery supportc		36 (20.9)
Needs characteristics
Perceived severity of opioid dependenced	6.8 (4.8)
Missed appointment due to transportationb		22 (12.9)
Positive HCV diagnosis		109 (62.6)
Overdoseb		18 (10.3)

SD: standard deviation, SSP: syringe exchange program, SUD: substance use disorder, MOUD: medications for opioid use disorder, HCV: hepatitis C virus

^aWith the exception of perceived severity of opioid dependence (see footnote d), all measures were collected during the 12-month data collection wave of the CARE2HOPE study. Perceived severity of opioid dependence was assessed only at baseline.

^bPast-six-month history

^cIncludes self-help groups (e.g., Narcotics Anonymous), outpatient counseling, inpatient or residential treatment, supervised detoxification, and staying in a sober house

^dPerceived severity was assessed at baseline using a 5-item scale scored on a 4-point Likert scale and the final perceived severity measure (range: 0–15) was calculated by summing the scores of the five items.

Overall, willingness to utilize an MTU was high, with 76.5% of participants reporting being willing. Of those who were not willing to use an MTU (n=44), 34.1% were concerned about a lack of privacy, 27.2% did not want people in the community to see them near the MTU, 25.0% did not want law enforcement to see them near the MTU, 13.7% were concerned about a lack of security, 0.02% did not believe an MTU would offer enough services, and 29.5% would not use an MTU for other reasons, including wanting to abstain from any opioid, not liking buprenorphine, and a lack of perceived need for MOUD.

Table S1 includes findings on preferences for MTU operations and staffing. The two most preferred locations for MTU parking were hospital or clinic parking lots (73.0%) or health department parking lots (71.7%). Most preferred the MTU operate on Fridays (74.1%) in the afternoons between 1–3 PM (52.3%) or in the evenings between 5–9 PM (55.8%) hours, but over a quarter of participants preferred after 9 PM (27.8%). Most indicated that they would want to receive services from nurses (80.5%), CARE2HOPE staff (79.3%), substance use counselors (76.4%), health professionals other than nurses (71.8%), or people in recovery (60.3%).

3.2. Model Results

Table 2 displays results of the adjusted logistic regression model, as well as unadjusted bivariate models specific to each of the BMVP correlates. After multiple imputation, the analytic sample (i.e., sample with no missing data for any covariate) comprised 174 participants. The odds of being willing to use an MTU were 73% lower among those who had recently been under community supervision compared to those who were not exposed to this level of CLS involvement (aOR: 0.27; 95% CI: 0.09–0.80). However, participants with a recent history of arrest (aOR: 0.73, 95% CI: 0.26–2.04) or incarceration (aOR: 2.79, 95% CI: 0.78–10.03) were not significantly more or less willing to use an MTU than those who did not experience these CLS exposures. None of the other selected predisposing characteristics (e.g., age, income) were associated with willingness to use an MTU.

Table 2.

Logistic regressions for the associations between willingness to utilize a mobile treatment unit and correlates, crude and adjusted (n=174)

Variablea	Crude OR (95% CI)	Adjusted OR (95% CI)
Predisposing characteristics
Age	0.99 (0.95–1.03)	1.00 (0.96–1.05)
Gender
Male	Ref	Ref
Female	1.25 (0.61–2.57)	1.55 (0.65–3.69)
Monthly income ($)	1.00 (0.97–1.02)	1.00 (0.97–1.03)
Criminal legal system involvementb
Arrest	1.12 (0.56–2.24)	0.73 (0.26–2.04)
Incarceration	1.29 (0.62–2.68)	2.79 (0.78–10.03)
Community supervision (e.g., parole)	0.51 (0.23–1.17)	0.27 (0.09–0.80)*
Enabling characteristics
Has transportation to medical appointments	0.67 (0.31–1.41)	0.52 (0.21–1.28)
Has health insurance or other coverage	0.97 (0.36–2.63)	0.72 (0.24–2.17)
SSP visitb	1.56 (0.75–3.21)	2.24 (0.89–5.68)
SUD treatment or recovery supportb
None	Ref	Ref
Treatment or recovery support without MOUDc	1.76 (0.50–6.19)	2.38 (0.35–16.33)
MOUD +/− other treatment or recovery supportc	4.61 (1.31–16.27)*	6.77 (1.91–23.96)**
Needs characteristics
Perceived severity of opioid dependenced	1.00 (0.94–1.05)	1.03 (0.94–1.13)
Missed appointment due to transportationb	0.81 (0.29–2.30)	0.78 (0.22–2.76)
Positive HCV diagnosis	1.08 (0.50–2.36)	1.14 (0.42–3.04)
Overdoseb	0.37 (0.13–1.05)	0.21 (0.06–0.79)*

OR: odds ratio, CI: confidence interval, SSP: syringe exchange program, SUD: substance use disorder, MOUD: medications for opioid use disorder, HCV: hepatitis C virus

^*p-value<0.05,

^**p-value<0.01

^aWith the exception of perceived severity of opioid dependence (see footnote d), all measures were collected during the 12-month data collection wave of the CARE2HOPE study. Perceived severity of opioid dependence was assessed only at baseline.

^bPast-six-month history

^cIncludes self-help groups (e.g., Narcotics Anonymous), outpatient counseling, inpatient or residential treatment, supervised detoxification, and staying in a sober house

^dPerceived severity was assessed at baseline using a 5-item scale scored on a 4-point Likert scale and the final perceived severity measure (range: 0–15) was calculated by summing the scores of the five items

Within the enabling characteristics domain, participants who had received MOUD treatment in the past six months had more than six times the odds of being willing to use an MTU compared to those who had not received any kind of SUD treatment (aOR: 6.77, 95% CI: 1.91–23.96). Access to transportation, insurance status, and history of recent SSP visit were not significant correlates.

Within the needs characteristics domain, those who had experienced an overdose in the past six months had approximately 81% lower odds of being willing to utilize an MTU (aOR: 0.21, 95% CI: 0.06–0.79). Other measures of need—perceived severity of opioid dependence, missing a medical appointment due to lack of transportation, and HCV status—were not significantly associated with willingness to utilize mobile MOUD services.

4. DISCUSSION

Recognizing the growing urgency of the rural opioid epidemic, PWUO, treatment providers, researchers, and other community members have issued calls for opioid treatment policy changes expanding access to MOUD and have expressed a need for tailored treatment approaches that address barriers specific to rural areas (Jenkins, 2021). SAMHSA’s proposed changes to 42 CFR Part 8 (SAMHSA, 2022) and subsequent local legal and regulatory changes allowing for the operation and expansion of MTUs have the potential to bridge the treatment gap in rural settings. However, few studies have explored perceptions of and interest in mobile programs among PWUO in these underserved areas.

This study makes a unique contribution to the literature by investigating willingness to use an MTU providing buprenorphine and naltrexone, if implemented, among PWUO in rural Appalachian Kentucky. Approximately three out of four participants expressed they would be willing to utilize a mobile program if offered, and those who were not willing primarily expressed concerns about a lack of privacy and law enforcement and others in the community seeing them near the MTU. These findings are consistent with qualitative studies in urban settings that demonstrate acceptability of mobile buprenorphine treatment among PWUO (Regis et al., 2020; Wenzel & Fishman, 2021). Acceptability of MTUs in rural areas is especially important given poorer geographic access to fixed-site treatment programs and increased driving distance to treatment providers (Kiang et al., 2021). A growing number of counties in rural Kentucky are administering SSPs and naloxone distribution programs through MTUs, and programs may wish to explore the possibility of implementing MOUD treatment services to expand MOUD access.

Importantly, a substantial proportion of the study’s participants were CLS-involved: 58.1% had been arrested, 42.8% incarcerated, and 21.4% under community supervision in the past six months. While the War on Drugs has historically disproportionately criminalized communities of color and reinforced the opioid epidemic and its consequences in urban areas, PWUO in rural America have increasingly borne its effects (Cooper et al., 2022). In particular, the War on Drugs has paved the way for a strengthened carceral state in rural communities, leading to increased drug-related arrests, incarcerations, and experiences with police violence (Gottschalk, 2020). Pretrial incarceration rates in rural areas have exponentially increased in recent years, and jails in rural areas have contributed more to the rise in mass incarceration in the US than those in urban areas (Kang-Brown & Subramanian, 2017).

Given high rates of CLS involvement in our population and in rural areas overall, our finding that PWUO under community supervision—compared to those not exposed to this level of CLS involvement—have 73% lower odds of willingness to seek mobile treatment may present a substantial impediment to improving opioid-related outcomes in rural communities. This finding is consistent with research at fixed sites (Grella et al., 2020). Individuals under community supervision may be unwilling to engage in mobile MOUD treatment because of fear that law enforcement officers or drug courts may find them in violation of their parole or probation if they (1) test positive for MOUD, even if prescribed, or (2) witness them interacting with individuals in possession of illicit drugs or firearms while in line outside of an MTU. Mobile programs may exacerbate concerns related to surveillance, as MOUD treatment-seekers may have to wait outside of MTUs prior to intake and for follow-up appointments. Barriers to MOUD treatment in this population reflect a discrepancy between policies expanding access to mortality-reducing MOUD and punitive CLS policies that prevent MOUD expansion. Protecting privacy while increasing MOUD access through MTUs is an important topic for future research. There was no association between other CLS measures and the outcome, perhaps because those who were recently arrested or incarcerated but not placed under community supervision after release do not need to meet the same requirements to remain in the community.

Recent receipt of MOUD treatment—an enabling characteristic as per BMVP—was another significant positive correlate of willingness to utilize mobile MTUs, consistent with prior research showing that past treatment predicts future willingness to accept treatment (Bergman et al., 2020; Blanco et al., 2015; Tuten et al., 2018), including in rural populations (Oser et al., 2011). Also, studies have found that fixed-site SUD treatment programs play an important role in increasing awareness and utilization of MTUs (Krawczyk et al., 2019; Stewart et al., 2021). More traditional treatment programs may thus refer patients to mobile programs in circumstances where geographic access to treatment is limited, as is often the case in rural settings.

Surprisingly, findings from this study highlight that those who had experienced an overdose in the past six months had 81% lower odds of being willing to use an MTU. Macmadu and colleagues (2021) similarly identified a history of overdose as a significant predictor of non-enrollment in any MOUD treatment program in a cohort of 18,374 PWUO in Rhode Island (Macmadu et al., 2021). The reasons underlying the negative association between prior overdose and MOUD treatment acceptability are presently unclear. Future research should aim to better understand the context of overdose events, including any interventions administered by first responders or bystanders at the scene or offers to help engage in treatment and recovery support services.

There are several limitations in this study. There was a small sample size limiting power to detect significant associations between the outcome and potentially important correlates, as well as community supervision status and reasons for not being willing to use an MTU. Additionally, methadone had been available through MTUs operating prior to 2007 and the movement to lift the moratorium on Drug Enforcement Administration (DEA) approvals of new MTUs was burgeoning at the time of the study (Chan et al., 2021; McBournie et al., 2019; Vestal, 2018). In June 2021, the DEA issued a final rule allowing the operation of MTUs linked to fixed-site OTPs (Drug Enforcement Administration, 2021). At the time of data collection, Kentucky did not have state regulations allowing for mobile methadone and methadone-providing MTUs were not yet federally authorized to operate. The intent of the survey questions was to guide community selection and implementation of an intervention launching in the coming years. As a result, the study team opted to focus on examining willingness to use an MTU providing other types of MOUD. MTUs providing mobile methadone are urgently needed in rural areas that lack access to fixed-site OTPs, and future research should explore willingness to use methadone-providing MTUs and design preferences among rural PWUO. The survey also did not distinguish between willingness to use MTUs and willingness to use MOUD generally. The analytic models did, however, evaluate the influence of a recent history of MOUD treatment, which reflects acceptability of continued or future MOUD use (Bergman et al., 2020; Blanco et al., 2015; Tuten et al., 2018). This study did not collect data regarding the circumstances of participants’ overdoses and their appraisals of these experiences to further explore whether interventions at the scene or linkage to treatment and services may explain our finding that those with a recent history of overdose have lower odds of willingness to use an MTU. Lastly, willingness to seek mobile MOUD services, may not translate to actual utilization of services. It will be important to investigate actual implementation and utilization of mobile services and rates of retention in rural areas over time.

Overall, these findings advance the application of BMVP in investigating SUD-related services in rural areas and inform the implementation of changes to 42 CFR Part 8 expanding mobile MOUD programs and the operation of existing MTUs offering buprenorphine, naltrexone, and other services for PWUO. This study supports the acceptability of these services among PWUO in US rural areas but suggests particular challenges for those who are CLS-involved or have a recent history of overdose. The effective implementation of MTUs offering MOUD, informed by community perspectives and needs, may decrease the rate of fatal and nonfatal opioid overdoses and other harms (e.g., injection-related infections such as HCV, endocarditis, and skin and soft tissue infections) and improve the lives of PWUO in rural settings.

HIGHLIGHTS.

• Acceptability of mobile treatment units among rural people who use opioids is high

• People under community supervision may be less willing to use mobile services

• Findings inform recent policy changes to expand treatment access using mobile units