Implementation of Jail and Prison-Based Medication Treatment for Opioid Use Disorder Programs: A Narrative Synthesis

Abstract

Provision of medications for opioid use disorder (MOUD) programs in carceral settings is critical to reducing overdose during the high-risk period following release from incarceration. Efforts to expand carceral MOUD programs have increased in recent years. We conducted a narrative review to synthesize evidence on the implementation of MOUD in U.S. carceral facilities. We analyzed 36 studies from 2019–2023 using the Exploration, Preparation, Implementation, Sustainment framework. Findings highlight that MOUD in carceral settings requires significant resources, infrastructure, and staffing. Concerns over MOUD diversion is common with variation in how programs respond. Stigma against MOUD remains a challenge, particularly when treating pregnant people with OUD. Effective coordination between carceral and community stakeholders is critical for MOUD implementation and continuity of treatment post-release. COVID-19 spurred innovation, increasing telehealth in carceral MOUD programs. Future research should explore MOUD program transition from early adoption to wide-scale implementation, considering external factors, sustainability, and evolving policies.

Introduction

Approximately 15% of incarcerated individuals in the United States (US) have an opioid use disorder (OUD), about 10 times the rate of the general population (Cates & Brown, 2023; National Academies of Sciences et al., 2019). This group is at dramatically elevated risk of fatal overdose during the weeks following their release from carceral facilities (Binswanger et al., 2007, 2013; Cates & Brown, 2023; Hartung et al., 2023; Ranapurwala et al., 2018). This is due in part to forced withdrawal in custody, which lowers tolerance when people return to use when returning to the community (Mitchell et al., 2009). Subgroups such as pregnant people with OUD in custody are at uniquely high risk of adverse outcomes during incarceration, with quadruple the risk of fatal postpartum overdose and contributing to maternal mortality (King et al., 2021). Jails, typically managed by local counties, are short-term facilities used for pretrial detention or sentences of less than one year. Prisons, operated by state or federal authorities, house individuals serving longer sentences for more serious crimes. In 2024, there were approximately 1.8 million individuals incarcerated, with the number of individuals in local jails just below 660,000 (Kang-Brown & Zhang, 2024).

There are currently three FDA-approved forms of medications for opioid use disorder (MOUD): methadone, buprenorphine, and naltrexone. All three forms of MOUD are proven to be safe and effective medication for OUD (National Academies of Sciences, 2019). Guidelines indicate that patients should be offered a choice of medications, as patients differ in their preferences and therapeutic needs (Box 1).

Box 1. Medications for Opioid Use Disorder.

The American Society for Addiction Medicine and the National Commission on Correctional Health Care support the provision of all three forms of FDA-approved MOUD to be available in jails and prisons; this availability if often not the case. States differ in their legal and policy mandates regarding MOUD provision in jail and prison settings.

Methadone:

The oldest approved MOUD, methadone is a full opioid agonist. Under federal regulations, methadone maintenance treatment can generally only be dispensed through regulated opioid treatment programs, or at a correctional facility that is registered with the Drug Enforcement Administration (DEA) as a hospital/clinic when treating a patient who has another medical diagnosis. Methadone can be administered to patients undergoing withdrawal and can be continued on a maintenance basis. Overdose risk is decreased by more than 50% for patients on methadone compared to abstinence-only treatment.

Buprenorphine (Suboxone, Subutex, Sublocade and Brixadi):

A partial agonist that is effective at managing the core symptoms of OUD and has a relatively low potential for abuse or overdose. Buprenorphine is formulated as a sublingual tablet, strip, transdermal patch, subcutaneous injection, and subdermal implant. Until 2023, buprenorphine could only be prescribed by clinicians with a special waiver but can now be prescribed by any clinician who possesses a controlled substance license and who has undergone a DEA-approved 8-hour training. Buprenorphine is a MOUD option for individuals experiencing withdrawal. Due to its affinity for the mu receptor, buprenorphine can displace lower affinity agonists from the receptor, initiating buprenorphine too early can risk precipitated withdrawal.

Naltrexone (Vivitrol):

A long-acting opioid antagonist that works by blocking the opioid receptors. Naltrexone is typically administered as a long-acting intramuscular injection. Naltrexone is not subject to any controlled substance regulations but can only be started for patients who are fully withdrawn from opioids (minimum 7–10 days). While naltrexone may be preferred by patients who are relatively stable in their recovery, many patients find it difficult to remain on naltrexone for extended periods of time.

Carceral-based MOUD programs are a critical evidence-based public health and medical intervention for incarcerated individuals with OUD, as MOUD is associated with improved recovery outcomes in custody and reduced risk of overdose, adverse health outcomes and re-incarceration upon release (Bahji et al., 2019; Evans et al., 2022; Klemperer et al., 2023; MacArthur et al., 2012; National Academies of Sciences, 2019; National Institute on Drug Abuse, 2018). Carceral-based MOUD programs in the US vary in their design and services. Some carceral facilities partner with community opioid treatment programs (OTPs) to administer MOUD onsite, while others have in-house buprenorphine providers. Facilities may also choose to become certified OTPs themselves or contract with external medical providers to offer treatment. Despite positive outcomes of carceral MOUD programs, MOUD is not available in most carceral settings in the US (Macmadu et al., 2020; Maruschak et al., 2023; Sufrin et al., 2023). Between 2022–2023, less than half (43.8%) of jails participating in a nationally representative study offered MOUD to incarcerated individuals and only 12.8% of jails provided MOUD to any individuals with OUD while in custody (Balawajder et al., 2024).

Recent years have seen increased efforts to scale-up carceral MOUD programs. From 2016 to 2021, buprenorphine use increased by over 220 times across US carceral settings, with the most marked increase occurring between 2020–2021(Thakrar et al., 2021). Despite reduced operational capacity early in the COVID-19 pandemic that impeded MOUD implementation in some carceral settings, other concurrent factors have accelerated expansion (Saunders et al., 2022).

Litigation against carceral facilities, including lawsuits from the US Department of Justice, has increased pressure to provide MOUD in carceral settings (Office of Public Affairs, 2022). Fourteen US states have also mandated the provision of MOUD in some carceral facilities (Homans et al., 2023). The substantial interest and high need to increase carceral MOUD programs provide an important opportunity to assess experiences to date.

New Contribution

A previous scoping review published by Grella and colleagues on MOUD implementation in US and international carceral facilities reviewed studies published through 2018 (Grella et al., 2020). The review described programmatic barriers to implementing MOUD that included capacity gaps in providing quality, guidance-concordant MOUD to incarcerated individuals resulting in poor experiences for patients and cascading impacts on continuity of care in the community upon release (Grella et al., 2020). The review identified a limited number of implementation facilitators of successful implementation of MOUD programs in carceral settings, primarily related to staff training and education. The study highlighted the need to identify additional facilitators of successful MOUD programs in carceral settings to characterize and document best practices (Grella et al., 2020).

The period captured in the Grella et al. review pre-dates the latest proliferation of carceral MOUD programs following a number of federal and state efforts to encourage MOUD adoption in jails and prisons. Further, the review focused on studies published before major changes in carceral and OUD policy spurred by the COVID-19 pandemic that increased MOUD provision. Thus, an updated review of barriers and facilitators of MOUD implementation during this recent period is warranted (Thakrar et al., 2021).

Since the Grella et al. review, expert bodies have issued best practice guidelines for MOUD programs in jails and prisons (National Commission on Correctional Health Care & National Sheriffs’ Association, 2018; New York State Department of Health, 2022; Solomon L et al., 2023; Substance Abuse and Mental Health Services Administration, 2019; Teixeira da Silva et al., 2023). These guidelines recommend comprehensive carceral-based MOUD programs that include use of validated screening and assessments, and offering a choice of MOUD medications, counseling, and reentry support. While they serve as a benchmark for evaluating programs, implementation varies across facilities due to differing ecosystems and practices (Bandara et al., 2021). This variation in implementation can help strengthen our understanding and evidence base of best practices for overcoming barriers and successfully implementing MOUD in these settings (Krawczyk et al., 2022). The aim of this narrative review is to provide an updated synthesis of implementation of MOUD programs in carceral facilities during a period of rapid growth in MOUD provision in jails and prisons and a corresponding increase in research on MOUD implementation.

Methods

Conceptual Framework

To guide our data abstraction and analysis process, we used the Exploration, Preparation, Implementation, Sustainment (EPIS) Framework. The EPIS framework is a conceptual framework that outlines the process of implementing evidence-based practices in public sector settings (Aarons et al., 2011). This framework has previously been applied to research assessing the implementation of substance use and mental health care in criminal justice settings (Kang et al., 2023; Pivovarova et al., 2022a; Zielinski et al., 2020). MOUD implementation in the context of this review refers to the process of integrating, executing, and sustaining MOUD programs in carceral settings.

There are four main constructs of the EPIS framework: the outer context, the inner context, bridging factors, and innovation factors (Aarons et al., 2011). When applying the EPIS framework to carceral MOUD programs, the inner context refers to decisions, support, attitudes, and systems inside carceral facilities that affect MOUD programs like education for staff to build knowledge on MOUD. The outer context includes external factors like stakeholders, policies, and the environment that impact carceral MOUD programs. Bridging factors refer to elements of cooperation and collaboration such as jails working with community providers for MOUD during and following incarceration; and innovation factors are novel approaches or strategies for implementing carceral MOUD programs or in connection with the community.

Study Inclusion and Identification

We followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) methodology as illustrated in Figure 1. Our inclusion criteria were studies 1) published between 2019–2023 (i.e. following the previous scoping review); 2) published in English; 3) focused on the United States; and 4) focused on provision, implementation, or assessment of implementation of MOUD in carceral facilities with justice-involved individuals and/or correctional professionals such as health care providers, custody officers, and jail or prison administrators.

Figure 1.

PRISMA.

Excluded – protocols, pure commentaries, hypothetical cases, international studies, non-English papers, modelling outcomes, MOUD in the community and not in carceral settings and MOUD programming impact evaluations

We included studies that used both qualitative and quantitative methodologies. We identified studies by searching CINAHL, PubMed, and Google Scholar using search terms outlined in Table 1. We screened and reviewed all the titles and abstracts to determine if an article met inclusion criteria.

Table 1.

Literature search methodology.

Keywords

“MAT”, “medication assisted treatment”, “MOUD”, “opiate/opioid substitution therapy”, “buprenorphine”, “methadone”, “naltrexone”, “prison”, “implementation”, “inmate”, “incarceration”, “penal”, “offender”, “detention”, “imprison”, “evaluation”, “implementation science”, “MOUD programming”, “jail”, “prison”, “incarcerated”, “correctional”, “carceral facilities”, “lessons learned” “medications for opioid use disorders”

Data Abstraction

We extracted full texts for all studies meeting inclusion criteria. For this analysis, we organized our findings on MOUD implementation during incarceration according to the four EPIS constructs. We used the EPIS constructs and its subcategories to organize the findings from our synthesis; we modified subcategories to align with carceral-based MOUD. Issues related to reentry and the interaction between carceral systems and the community treatment environment are the subject of a separate manuscript. The scope of this manuscript is on implementation of MOUD within carceral settings.

For each included study, we abstracted and categorized the study findings according to EPIS inner, innovation, and bridging constructs. There were three publications that presented findings from the same data set from a nested study; for this analysis these papers were categorized as separate papers, as they focused on different aspects of MOUD implementation. Table 2 outlines information on the studies included with more detailed information in Supplement 1.

Table 2:

Included Studies

	Study	Study Setting and Population	Sample size	Data collection period	Study Design	EPIS construct
1.	(Ferguson et al., 2019)	Four correctional systems (two unified systems) in Rhode Island and Connecticut, and two jails in Massachusetts	N=4 correctional system teams	Not included	Monthly data and narrative reports from each correctional system	Bridging Innovation
2.	(Brinkley-Rubinstein et al., 2019)	Rhode Island Department of Corrections in Cranston, Rhode Island Formerly incarcerated individuals and enrolled in the MOUD program in a unified jail and prison system	N=40 interviews	Not included	Qualitative interviews	Inner context
3.	(Lam et al., 2019)	Rhode Island Department of Corrections Incarcerated individuals who are to be released soon	N=80 individuals	July 2018	Pre-post survey in watching an educational video	Innovation
4.	(Hanna et al., 2020)	One male and one female facility and three counties of release in Michigan Stakeholders-providers and facility staff, policymakers, and current and former program participants	N=28 structured meetings N=6 focus groups N=10 individual interviews	Not included	Observational field notes; notes from structured meetings and discussions with stakeholders; focus groups; individual interviews; site visits	Inner context Innovation
5.	(Grella et al., 2020)	Criminal justice settings	N=53 papers N=28 international; N=23 from the US	2002–2018	Scoping review	Inner context Bridging
6.	(Sufrin et al., 2020)	State prisons and county jails participating in the Pregnancy in Prison Statistics Study	N=6 county jails N=22 state prisons	2016–2017	Survey; epidemiological data analysis	Inner context
7.	(Bandara et al., 2021)	U.S. jails and prisons that implement MOUD programming Correctional system leaders who manage MOUD programs (medical director, warden)	N=20 interviews with 35 individuals N=19 jails	August 2019 to January 2020	Qualitative interviews	Inner context Bridging Innovation Outer context
8.	(Martin et al., 2021)	A large county jail system in Maricopa County, Arizona	N=1 jail system	Not included	Sequential Intercept Model (SIM) to map the process of the criminal justice system	Inner context Bridging Innovation
9.	(Duncan et al., 2021)	Minneapolis, Minnesota Hennepin County Jail	N=1 jail	January to February 2020 and April to May 2020	Pre – post case study	Inner context Innovation
10.	(Donelan et al., 2021)	Greenfield, Massachusetts Franklin County Sheriff’s Office (FCSO)	N=1 jail	2020	Descriptive case example	Innovation
11.	(Barnes et al., 2021)	California’s San Mateo County (SMC) Clinical staff implementing MOUD services, correctional health administrators, and community partners	N=8 interviews	2019	Qualitative interviews	Inner context
12.	(King et al., 2021)	U.S. states with high rates of maternal opioid use Providers and administrators from opioid treatment providers (OTPs)	N= 16 opioid treatment providers N=16 interviews	June 2019 to February 2020	Qualitative interviews	Inner context Bridging Innovation
13.	(Pivovarova et al., 2022b)	Jails in Massachusetts Medical, correctional, administrative, MOUD jail staff	N=7 jails N=61 individuals	December 2019 to January 2020	Qualitative interviews and focus group discussions	Inner context Bridging Innovation
14.	(Scott et al., 2022)	U.S. counties with high opioid-related death rates Jail administrators, medical directors, health administrators, program directors	N=185 jails	December 2019 to February 2021	Interview questions via structured survey	Inner context Bridging Outer context
15.	(Krawczyk et al., 2022)	County jails from across New Jersey Custody facility	N=17 baseline surveys N= 11 interviews	October 2019 to September 2020	Baseline survey and qualitative interviews	Inner context Bridging Innovation Outer context
16.	(Matsumoto et al., 2022)	Jails in Massachusetts Clinical, correctional, senior jail administrators	N=7 jails N=61 individuals	December 2019 to January 2020	Qualitative interviews and focus group discussions	Inner context Bridging Innovation
17.	(Stopka et al., 2022)	Seven counties in Massachusetts Medical, supervisory, and administrative staff at community-based MOUD programs serving jail-referred 14 individuals	N=36 interviews N=18 community-based MOUD programs	March to October 2020	Qualitative interviews	Inner context Bridging Outer context
18.	(Ludwig et al., 2022)	U.S. Jail Systems Jurisdiction team lead	N=16 jurisdictions	August 2019 to April 2020	Checklist at baseline and follow-up	Inner context Bridging Innovation
19.	(Cheng et al., 2022)	New York City jails Recently released individuals with OUD on Extended-release buprenorphine (XRB)	N=16 interviews	October 2019 to May 2020	Qualitative interviews	Inner context
20.	(Victor et al., 2022)	Rural communities in the U.S. Criminal legal and treatment stakeholders – court systems, individuals with MOUD experience, public health organizations, medical professionals, advocacy org, and public safety professionals	N=7 interviews N=110 surveys	March 2020, to June 2020 – survey	Survey and semi structured interviews	Inner context
21.	(Evans et al., 2022)	Massachusetts jails Staff who implemented MOUD programming (administrative, security, behavioral health, and clinical staff)	N=7 jails N=61 individuals	December 2019 to January 2020	Qualitative interviews and focus groups	Inner context
22.	(Treitler et al., 2022)	New Jersey state prisons Individuals with OUD released from prisons	N=53 interviews	July 2020 to April 2021	Qualitative interviews	Inner context Outer context
23.	(Kaplowitz et al., 2022)	Rhode Island Department of Corrections Individuals receiving MOUD during incarceration	N=40 interviews	July to August 2018	Qualitative interviews	Inner context Outer context
24.	(Sufrin et al., 2022)	U.S. jails Medical and custody leaders	N=836 jails	August 19 -November 7, 2019	Survey	Inner context
25.	(Clark et al., 2023)	U.S jails and community-based treatment providers (CBPT) in 14 states	N=29 jails N=30 CBPT	Not provided	Baseline survey	Inner context Bridging
26.	(Harrington et al., 2023)	Jails in Massachusetts Jail staff who implemented MOUD programming	N=2 jails N=29 interviews	Oct 2020- Jan 2021	Qualitative interviews and focus groups	Bridging Innovation Outer context
27.	(Booty et al., 2023)	Kentucky Deportment of Corrections Social service clinicians	N=25 interviews	Not included	Qualitative interviews	Inner context
28.	(Pivovarova et al., 2023)	Courts in one state in Northeastern USA Drug court staff--judge, probation officer, prosecutor, defense attorney, clinician, and/or case manager	N= 7 courts N=21 interviews	March 2021 to June 2021	Qualitative interviews	Bridging Outer context
29.	(Staton et al., 2023)	A rural Appalachian County in Kentucky Individuals in jail or on probation with a history of opioid use	N=10 interviews	Not included	Qualitative interviews	Inner context
30.	(Kaplowitz et al., 2023)	Rhode Island Department of Corrections Incarcerated individuals receiving MOUD	N=40 interviews	June to August 2018	Qualitative interviews	Inner context Innovation
31.	(Whaley et al., 2023)	A jail in Camden, New Jersey	N=1 jail	Began in February 2019	MOUD adoption experience description	Inner context Innovation
32.	(Pourtaher et al., 2023)	Albany County Jail and Rehabilitative Services Center (ACCRSC) Incarcerated and released individuals who were enrolled in the MOUD program	N=375 individuals	January 19, 2019, and the end of December 2020	Site visits; ACCRSC jail management system and the facility’s electronic medical records system data; statewide health data	Inner context Bridging Innovation
33.	(Blue et al., 2023)	16 different jurisdictions across the U.S. Correctional and treatment staff	N=123 staff participants	August 2019 through April 2020	Surveys at baseline and one-year follow-up	Inner context
34.	(Thomas et al., 2023)	Counties in South Central Appalachia Sheriffs and jail administrators from detention centers distributed across	N=9 counties N=19 interviews	May 2022 to December 2022	Qualitative interviews	Inner context Outer context
35.	(Grella et al., 2023)	U.S. counties with high opioid overdose mortality Representatives from jails knowledgeable about MOUD that house pregnant people	N=174 jails	December 2019 and February 2021	Structured interviews	Inner context Outer context
36.	(Staton et al., 2023)	A rural Appalachian Community in Kentucky Criminal justice involved individuals with OUD	N= 9 individuals completed baseline and follow-up interviews	September 2018 - March 2022	Baseline interview with study participants; follow-up interview 3 months following baseline	Inner context

Results

In total, we included 36 studies in the review. Following the EPIS framework, we organized the findings from the 36 studies according to the inner (system and individual level), bridging, and innovation constructs.

1. Inner Context

1.1. Organizational characteristics, resource allocation, and staffing

MOUD delivery in jails and prisons is resource intensive, requiring dedicated infrastructure and staffing.

MOUD delivery is frequently described as needing significant additional finances, staff, space, and time, which can strain already limited resources in carceral facilities (Bandara et al., 2021; Grella et al., 2020; King et al., 2021; Krawczyk et al., 2022; Pivovarova et al., 2022b; Scott et al., 2022). A study describing barriers and facilitators of early implementers of MOUD in US jail and prison systems described the lengthy process of buprenorphine dosing conducted in small groups with intensive security measures (Bandara et al., 2021). Medication dosing varied between facilities including diversion mitigation measures such as mouth checks, which require dedicated teams and time, ranging from several hours to a full day to complete (Bandara et al., 2021). Furthermore, MOUD program implementation, including eligibility screening, medication dispensation and provision of counseling, can disrupt existing carceral operations or clinical workflows. Security concerns regarding moving individuals with varied security classifications to receive MOUD can take additional resources including extra staff (Bandara et al., 2021). Partnerships with community providers can offer jails additional resources including trained jail staff to alleviate some burden or gaps in capacity (Clark et al., 2023).

1.2. Monitoring and Reporting

Few formal carceral-based MOUD program evaluations have been conducted and limited monitoring systems are in place to adequately track outcomes.

Evaluation and quality improvement for MOUD programs in carceral settings remains limited. Insufficient formal evaluations have been conducted to evaluate MOUD programs at or across facilities (Bandara et al., 2021; Clark et al., 2023). Few facilities have systems to track individual and/or program-level MOUD outcomes (Grella et al., 2020; Scott et al., 2022). This creates challenges in being able to make timely, evidence-informed decisions at a facility and policy level (Barnes et al., 2021; Grella et al., 2020; Scott et al., 2022).

A comprehensive system that tracks the administration of MOUD, medication storage, and transfers of individuals between facilities would be beneficial for monitoring patients along the correctional treatment cascade (Scott et al., 2022).Information shared across stakeholders, such as between clinical community providers and carceral facilities, facilitates improved continuity of care between the community and carceral facility (Martin et al., 2021). Concurrently, state-level reporting requirements on MOUD implementation information and data are described as duplicative and burdensome for some jail staff (Pivovarova et al., 2022b). Sharing data across and within carceral facilities can be challenging with the fragmented nature of the corrections system (Ludwig et al., 2022).

Indicators (when they were collected) included number of people receiving treatment while incarcerated, the number of individuals entering and retained in community treatment, the number of individuals who experience rearrest, and/or the number of individuals who experience an overdose in the community following release (Clark et al., 2023).

1.3. Organizational characteristics: structures and processes for diversion

Concerns regarding MOUD diversion were common among carceral and medical staff in jails and prisons.

Diversion is the practice of patients either selling or sharing their medication with people outside the program or self-administering it in a non-prescribed way. Diversion was reported frequently across judicial, carceral, and clinical stakeholders (Bandara et al., 2021; Barnes et al., 2021; Cheng et al., 2022; Victor et al., 2022). Due to its slow dissolving formulation (which can allow medication to be cheeked), concern arose more commonly in the context of buprenorphine versus methadone. One study described using buprenorphine strips instead of tablets to diminish the risk of diversion due to their ability to dissolve more rapidly (Pourtaher et al., 2023). Carceral facilities implement various surveillance measures to monitor and reduce potential diversion of buprenorphine (Evans et al., 2022). Surveillance approaches include elaborate medication dosing procedures, the administration of regular and/or random urine drug tests and monitoring the incarcerated individual’s commissary funds to identify unusual activity (Evans et al., 2022). A study conducted in seven Massachusetts jails described how the jails implemented specific ratios of carceral staff to MOUD patients (3 staff:15–20 patients) to allow for thorough mouth checks, observations during dosing, and camera surveillance activity during MOUD provision (Evans et al., 2022). Efforts to minimize diversion risk were described as resource intensive, requiring significant human, financial, and logistical resources (Krawczyk et al., 2022; Scott et al., 2022).

Procedures intended to mitigate the risk of diversion included the use of protocols related to dosing and diversion prevention, patient education, and implementing the same dosing for all patients on MOUD (Bandara et al., 2021; Evans et al., 2022). Diversion was reported less frequently in jails that established MOUD administration protocols using measures such as mouth checks (Bandara et al., 2021; Evans et al., 2022; Thomas et al., 2023). Around half of the jails in a 29-jail sample instituted protocols to minimize diversion for methadone and/or buprenorphine (Clark et al., 2023). Protocols, such as those for MOUD dosing, that facilitated a level of flexibility to allow staff discretion to meet individual patient needs were described by corrections and clinical staff as useful and facilitated a more secure atmosphere (Barnes et al., 2021; Evans et al., 2022). In one study, sheriffs and jail administrators were not concerned with diversion given the use of detailed policies governing the environment and process for medication administration (Thomas et al., 2023).

Intentionally informing incarcerated patients that jail staff were aware of diversion techniques and educating patients on the significance of taking MOUD as prescribed were approaches implemented to reduce diversion risk in Massachusetts jails (Evans et al., 2022). Carceral staff in these jails reported their perceptions that patients may be unaware of how diversion could impact their withdrawal symptoms and MOUD experience (Evans et al., 2022). Other strategies used to reduce diversion risk included concealing which patients were in the MOUD program to limit identifying them as targets for coercion by other facility residents (Evans et al., 2022). Some facilities placed all individuals on the same dose or set maximum dosage limits for buprenorphine despite varying patient needs, due to concerns that higher doses posed higher diversion risk (Bandara et al., 2021).

Carceral facilities vary in their responses to MOUD diversion.

Once diversion is identified, facilities varied in their response (Bandara et al., 2021; Evans et al., 2022; Krawczyk et al., 2022; Pourtaher et al., 2023; Whaley et al., 2023). Several studies described the use of small teams to identify the reason diversion occurred; once transpired, teams worked to address the contributing factors which could include coercion from other individuals or receiving more medication than needed (Evans et al., 2022; Krawczyk et al., 2022; Whaley et al., 2023). Patient education or counselling was provided in combination with the small teams-based approach (Pourtaher et al., 2023; Whaley et al., 2023). Clinical jail staff, administrators, and custody officers who participated in one study emphasized the importance of including patients suspected of diversion in the investigation to allow them an opportunity to provide input and be part of solution discussions (Evans et al., 2022). In the case of continued diversion, dosing adjustments, or altering medication administration practices, (e.g., by providing medication individually rather than in a group setting) were implemented (Evans et al., 2022; Whaley et al., 2023). Some studies also described punitive responses, which included dose reductions or termination of patients from the MOUD program (Bandara et al., 2021; Evans et al., 2022). In one study, in the case of ongoing diversion, an individual would be removed from the MOUD program until two weeks before their release at which point, they were re-initiated on MOUD and would receive re-entry support (Pourtaher et al., 2023).

Implementation of MOUD was perceived to reduce diversion of contraband and associated incidents, including violence, inside jails and prisons.

Despite initial apprehension, jail staff in some facilities shared how their concerns regarding diversion lessened once they began systematically implementing MOUD (Evans et al., 2022; Whaley et al., 2023).With the initiation of MOUD programming, carceral staff reported fewer observed efforts to smuggle buprenorphine in from outside the facility and less violence and “predatory” conduct associated with victimization of people receiving MOUD (Evans et al., 2022). The reduced smuggling of contraband was perceived to increase staff approval of the MOUD program (Evans et al., 2022).

1.4. Organizational characteristics: implementation processes for pregnant people

MOUD provision for pregnant and postpartum people remains suboptimal, lacking protocols, capacity, and policies that are pregnancy-specific to guide administration in carceral settings.

Despite the established standard of care and the known risks of withdrawal in pregnancy for pregnant people with OUD, access to MOUD in jails and prisons is limited in pregnancy (Grella et al., 2020, 2023; King et al., 2021; Sufrin et al., 2020, 2022). Results from a nationally representative survey found that only 32% (267/836) of jails in the analysis initiated and continued MOUD for pregnant people (Sufrin et al., 2022). Notably, nearly a third of jails (28%; 237/836) only continued MOUD if the pregnant person was already on MOUD in the community before arriving to jail, but would not initiate treatment (Sufrin et al., 2022). Some differences in access to MOUD was reported based on the jail’s health care provider arrangements; in jails with privately contracted health care, there was a 1.49 greater odds (95% CI, 1.03–2.14; P = .03) of MOUD continuation in pregnancy compared to jails without privately contracted care, such as direct service delivery or contracts with individual medical providers or local health care clinics (Sufrin et al., 2022).

Multiple studies reported that high proportions of jails and prisons discontinue MOUD for pregnant people following birth (Sufrin et al., 2020, 2022). From a nationally representative survey of jails, 76% (384/504) of the jails providing MOUD in pregnancy discontinued treatment postpartum (Sufrin et al., 2022). This finding of postpartum MOUD discontinuation was echoed in another study that included 22 state prison systems, in which three-quarters did not allow treatment after pregnancy (Sufrin et al., 2020). Carceral staff lack information on the benefits of pregnancy MOUD and held negative views of MOUD in pregnancy, highlighting the need for interventions to address misinformation and discrimination toward pregnant people (King et al., 2021; Scott et al., 2022).

1.5. Individual characteristics: attitudes towards MOUD

Stigma among custody and medical staff remains a challenge, especially among staff with no previous MOUD experience.

Stigma from custody, legal professionals, law enforcement medical staff towards MOUD and people with OUD was a common theme across studies. Perhaps because of generational differences, research suggested that younger custody staff tended to be more accepting of MOUD compared to older correctional staff (Pivovarova et al., 2022b). Some studies indicated that acceptance of MOUD within these groups improved over time (Bandara et al., 2021; Whaley et al., 2023).

The existence and experience of stigma in peer recovery groups and from custody officers, as well as self-stigma, influence patient preferences and choice of MOUD. For instance, stigmatizing attitudes from carceral staff led some patients to choose monthly injections of extended-release naltrexone, as it reduced daily interactions with correctional staff (Cheng et al., 2022; Kaplowitz et al., 2022). In some cases, custody staff shared stigmatizing views on MOUD, which could influence the patient’s decision to begin or continue MOUD (Booty et al., 2023).

Researchers have identified a variety of education and training to address stigma (Blue et al., 2023; Grella et al., 2020; Lam et al., 2019; Stopka et al., 2022; Victor et al., 2022). One study found that staff training was associated with increased familiarity with and referrals for MOUD and produced significant improvements in viewing MOUD as helpful for patients (Grella et al., 2020). Another study reported more favorable attitudes toward MOUD and increased knowledge among custody and medical staff following targeted educational MOUD activities (Blue et al., 2023). One study explored how ongoing staff education integrated into routine meetings built buy-in for MOUD programming (Whaley et al., 2023). Another study reported using a video intervention with incarcerated individuals that portrayed stories of fictional incarcerated people using MOUD; this intervention increased self-reported MOUD knowledge and improved positive attitudes towards MOUD among incarcerated individuals (Lam et al., 2019).

1.6. Leadership

Committed leadership was key to drive culture change and increase MOUD buy-in among custody staff (Grella et al., 2020; Hanna et al., 2020; Pivovarova et al., 2022b; Thomas et al., 2023). Staff often took their cues from leadership, such as wardens and medical directors (Bandara et al., 2021). Leadership also supported MOUD implementation by developing workflows, defining staff roles, and creating the flexibility needed for MOUD programming (Pivovarova et al., 2022b; Thomas et al., 2023).

2. Outer context

2.1. Funding availability and Contracting

Adequate funding to support MOUD implementation and sustainment is a significant concern for carceral facilities (Kaplowitz et al., 2022; Krawczyk et al., 2022; Scott et al., 2022). Funding affects the number of carceral staff available to operationalize a MOUD program and directly impacts decisions around services offered and contract negotiations with external medical providers who provide medical services inside jails (Thomas et al., 2023). Insufficient funding and staffing resources further affect facilities’ capacity to proactively support discharge planning for individuals with OUD prior to their release, adversely affecting treatment continuity in the community (Stopka et al., 2022). The availability and implementation of certain MOUD treatment, like extended-release buprenorphine, is limited by funding constraints due to high medication costs (Treitler et al., 2022).

2.2. Service Environment and Policies

Regulatory burdens on MOUD licensing impact feasibility and decision-making in carceral facilities

MOUD programs in jails and prisons are constrained by complex federal and state regulations, which are often described as burdensome (Bandara et al., 2021; Krawczyk et al., 2022; Pivovarova et al., 2023). The most salient are federal regulations around methadone, which regulate that only certified opioid treatment providers can dispense methadone. This requires facilities to either establish an onsite OTP, transport patients offsite to an OTP, or partner with an OTP in the community under a guest dosing arrangement to bring doses onsite. Buprenorphine is less regulated, particularly after the elimination of the X-waiver in 2023, but is still regulated under federal controlled substance requirements. One study reported most jails that offered methadone did so by transporting patients offsite for dosing or used guest-dosing for onsite administration; jails that provided buprenorphine had primarily on-site administration through a contracted provider (Grella et al., 2023) Few jails were licensed to administer methadone as an OTP or received a buprenorphine waiver (previously known as the X-waiver) for provision in-house (Grella et al., 2023).

All current policy options for methadone require significant coordination between clinical and carceral entities (Bandara et al., 2021; Krawczyk et al., 2022). Some facilities formed agreements with community providers to dispense methadone directly in the correctional facility or provide 1–2 weeks’ worth of doses for the correctional medical staff to dispense under a guest dosing agreement (Bandara et al., 2021). One site in Massachusetts described providing methadone on-site at the jail with take-home bottles instead of transporting people to local methadone clinics; this was described as a positive innovation for optimization of administration as well as for the experience of the patients not being “paraded in belly chains and shackles” in a community setting (Harrington et al., 2023). Facilities without medical contracts are restricted in how they can administer MOUD; bringing patients to community clinics or hospitals for treatment is necessary for provision of medication (Thomas et al., 2023).

3. Bridging Factors

3.1. Community collaboration

Coordination between carceral facilities and community providers is a key determinant of MOUD program success in jails and prisons.

Coordination and communication between carceral facilities and community providers play a critical role in the successful administration of MOUD among individuals in carceral facilities (Ferguson et al., 2019; Grella et al., 2020; King et al., 2021; Matsumoto et al., 2022). Carceral facilities and community provider relationships can take various forms. Carceral facilities can partner with a community OTP to administer MOUD on site, transport individuals to community-based OTPs for routine dosing, or contract with community entities to supply MOUD doses with administration done onsite by corrections staff. Ongoing communication provided opportunities to clarify roles among carceral and community staff, understand different organizational work cultures, and share challenges in a non-confrontational manner (Hanna et al., 2020). Effective communication models included regular phone calls, meetings, and having a contact person at the carceral facility to facilitate coordination (Hanna et al., 2020; King et al., 2021; Pivovarova et al., 2022b; Stopka et al., 2022). A key focus of communication was to establish linkages to critical social and clinical supports during the re-entry period (Hoffman et al., 2023; Pourtaher et al., 2023). In Maricopa County, Arizona, community service providers are credentialed through the county jail system and have access to rosters and information on patients to facilitate confirmation of MOUD prescription and dosages (Martin et al., 2021). This coordination between community and correctional entities allowed for the proactive identification of individuals for MOUD continuity upon entry into jail or prison (Martin et al., 2021).

By contrast, poor communication between correctional and community providers hindered planning and treatment efforts. This affected MOUD initiation, release and continuity planning, and referral decisions made by court stakeholders (Brinkley-Rubinstein et al., 2019; Pivovarova et al., 2023; Stopka et al., 2022).

4. Innovation

4.1. Innovation and Adaptation

COVID-19 created an environment for innovation and tele-health utilization that was extended to MOUD programming in jails and prisons.

Telehealth options for correctional health care in general, and for MOUD specifically, were employed before and ramped up during the COVID-19 pandemic. Telehealth approaches ranged from education-focused activities and psychosocial therapy to utilization of virtual platforms to initiate MOUD (Duncan et al., 2021; Harrington et al., 2023; Krawczyk et al., 2022). The relaxed federal telehealth regulations during COVID-19 facilitated virtual buprenorphine initiation through computer and audio/visual communication between incarcerated individuals and community medical providers (Duncan et al., 2021). Utilization of phone-based services and virtual education also increased utilized during COVID-19. A county jail in New York utilized phones and pre-recorded videos to conduct naloxone training (Pourtaher et al., 2023).

Discussion

We reviewed 36 studies that included information on the implementation of MOUD programs in US jails and prisons published between 2019 and 2023. This period coincided with an era of unprecedented expansion of carceral MOUD programs, a growing body of published research on this topic, rising overdose rates, and the onset of the COVID-19 pandemic, which fundamentally altered health care in jails and prisons.

Our review identifies several instances of unique problem-solving. These included the novel use of telehealth, new strategies to reduce stigma and educate correction and medical staff, and administrative workarounds to improve access to post-release services (Donelan et al., 2021; Harrington et al., 2023; Krawczyk et al., 2022). However, the resources and innovation needed to successfully implement MOUD programs also indicate that starting and sustaining programs requires facilities to go “above and beyond” normal operational procedures and likely explains why comprehensive MOUD programs in jails and prisons remain relatively rare compared to the number of carceral facilities and the number of people who use opioids and have criminal legal system involvement. Additional research to examine ways carceral systems can implement MOUD programs in less resource intensive ways would be beneficial, as would research into other policy measures such as the development of national standards, metrics for evaluating program outcomes, and oversight for MOUD programs in jails and prisons around screening, administration, and re-entry.

The implementation experiences from across the US highlight the diverse approaches employed without a one-size -fits-all model, emphasizing the value added from experience sharing across correctional settings and systems. Several technical assistance platforms provide opportunities for peer-to-peer learning, including the Tele-Echo model - a virtual cross-learning platform for custody professionals, policy academies, and learning labs. As this arena of evidence grows with the maturity of carceral MOUD programming, the systematic documentation of best practices and approaches to address systemic, internal, and external challenges will be critical (Health Management Associates, 2023; Kawasaki et al., 2022). Given the recent adoption of MOUD carceral programs, much of the literature focused on relatively new programs or on programs that were early to adopt. Diversion was a key factor affecting MOUD availability and implementation in carceral settings, particularly during rollout and ongoing administration, due to the resources needed for a proactive response. Sustainability of MOUD programming in carceral settings is not well explored. Implementing MOUD programs requires dedicated investment in resources, staffing, and culture change to ensure sustainment long-term. Funding needs vary based on carcel MOUD program specifics while also occurring in the context of often resourced constrained settings with implications for planning and operationalization of programs across states (Ryan et al., 2023).

Policies focused on and affecting MOUD programs in carceral settings continue to evolve with an increasing momentum towards adoption, such as the repeal of the X-waiver which removed the obstacle of the federally required waiver to prescribe buprenorphine or the expansion of state laws mandating MOUD implementation. Policies such as these can be pertinent in addressing key challenges highlighted in included studies such as ongoing challenges accessing MOUD during incarceration. This also includes the new regulations concerning methadone which expanded the types of healthcare providers who are eligibility to prescribe and dispense methadone, allowing for initiation of methadone via telehealth and for provision of take-home doses, alleviating burdens around daily clinic attendance (Asher Funk et al., 2024).

Directions for future research

Implementation research will likely need to pivot to describe the implementation of MOUD programs as it progresses from the early adoptive phase into a scaled-up, more widely implemented service in carceral settings. This includes exploring the influence of external factors with a changing policy landscape such as the Medicaid 1115 waivers of the Inmate Exclusion Policy that were started in 2023. As of January 2025, these waivers had been granted in 19 states and 7 states and D.C. had pending applications (Roth et al., 2025). The 1115 waivers give states more flexibility to provide Medicaid covered clinical and substance use provisions before an incarcerated individual’s release and funding availability, which has implications for sustainability of MOUD programs (Hinton et al., 2024). It will also be important to examine how the new methadone regulations impact methadone implementation and access in carceral settings. Several areas of further exploration remain, including capturing lived experience of incarcerated and recently released individuals from more mature MOUD programs and evaluations assessing efficacy and effectiveness of MOUD programs in their various forms. Assessing changes in organizational climate with initiation and continuation of MOUD programs in carceral settings also remains an area for evaluation.

Conclusion

The literature on MOUD implementation in US jails provides insights into the experiences of vanguard facilities, as care remains far from the OUD standard approach to managing OUD in carceral settings. Given the adaptation and contextual considerations of MOUD programs, implementation research will need to evolve to capture evidence-based practices informed by all stakeholders to be utilized for decision making for adaptation and the future of MOUD programs. The findings of this narrative review call attention to subsequent directions for research, clinical practice, and policy evaluation. Developing systematic approaches to adapt MOUD service delivery in correctional settings is essential to the overall goal of addressing the opioid epidemic and preventing overdose deaths.

Data availability statement

No new data were created or analyzed in this study. Data sharing is not applicable to this article.