Emergency department–based efforts to offer medication treatment for opioid use disorder: What can we learn from current approaches?

Abstract

Background:

The opioid epidemic remains a public health crisis and most people with opioid use disorder (OUD) do not receive effective treatment. The emergency department (ED) can be a critical entry point for treatment. EDs are developing and implementing ED-based efforts to address OUD to improve access to OUD treatment. This study’s objective is to identify features of ED-based OUD treatment programs that relate to program implementation, effectiveness, and sustainability.

Methods:

We obtained data through literature review and semistructured interviews with ED physicians and leaders. The study analyzed these data to develop a framework of key components of ED-based efforts and highlight barriers and facilitators to implementation and program effectiveness.

Results:

We identify five key features of ED-based opioid treatment programs that vary across programs and may influence effectiveness and impact: patient identification methods; treatment approaches; program structure; relationship with community partners; and financing and sustainability. Successful implementation of ED-based OUD treatment includes having a champion, a reliable referral network, and systematic tracking and reporting of data for monitoring and feedback.

Conclusion:

Going forward, attention to these features may help to improve effectiveness. As researchers conduct studies of ED-based care models, they should assess the impact of variation in key features to improve program effectiveness.

1. Introduction

Opioid-related morbidity and overdose deaths continue to be a significant public health problem. The prevalence of opioid use disorder (OUD) has increased dramatically in the past two decades, with opioid-related deaths in the United States numbering more than 50,000 in 2019 (Ahmad, Rossen, & Sutton, 2020). Public and private initiatives have expanded prevention, treatment, and emergency care. Nevertheless, the opioid epidemic remains a national emergency.

Three medications for opioid use disorder (MOUD)—methadone, buprenorphine, and XR-naltrexone—are effective, evidence-based treatments. While not required for treatment, MOUD is associated with improved outcomes, including treatment retention and functioning, less drug use and health care utilization, and lower risk for overdose (Bart, 2012; Lo-Ciganic et al., 2016; Morgan, Schackman, Weinstein, Walley, & Linas, 2019; Schackman, Leff, Polsky, Moore, & Fiellin, 2012; Thomas et al., 2014; Volkow, Frieden, Hyde, & Cha, 2014). However, MOUD is used less often than recommended (Larochelle et al., 2018). Capacity to offer MOUD is limited in many parts of the United States and does not meet the need for care. In 2018, less than 20% of individuals with OUD received treatment (Substance Abuse and Mental Health Services Administration, 2019). Opioid medication treatment rates declined among privately insured populations between 2009 and 2016 (Shen, Barrette, & Dafny, 2020).

One strategy to improve access to MOUD is to reach patients during emergency department (ED) visits. For individuals with OUD, the ED is a critical entry point to potential treatment (Kaczorowski et al., 2020); and studies have recommended that all ED patients with OUD who are not already engaged in medication treatment be considered for MOUD (Strayer et al., 2020). Models of care are emerging that offer medication initiation in the ED and link patients to ongoing care. These programs may have significant potential to save lives, support individuals with OUD in entering treatment, and reduce the treatment gap. However, many EDs treat overdose or withdrawal symptoms alone, and offer referrals to community providers. Such practices can be insufficient to ensure successful, continued treatment and recovery.

As EDs implement ED-based efforts, health systems, providers, and policy-makers must understand how key features of these programs relate to program implementation, effectiveness, and sustainability. By identifying key program features and barriers and facilitators to operating ED-based efforts to offer MOUD, we may be able to identify health system, payer, state, and federal policy and regulatory strategies to improve access to MOUD. The purpose of this study is to examine the literature and conduct and analyze expert interviews to identify a framework of key components of ED-based efforts to offer MOUD and highlight what we know about programs’ reach and effectiveness.

2. Methods

Data come from two sources: a review of the literature, including both academic peer reviewed literature and nonacademic literature (e.g., government reports and white papers); and semistructured interviews with physicians and researchers from across the United States expert in ED-based OUD treatment. The university IRB determined that this study did not meet the definition of human subjects research. We searched PubMed and Google Scholar to identify articles related to implementation and effectiveness of ED-based OUD treatment and to identify potential experts to interview. To focus on the most recent information, the literature search included English-language research studies published between 2015 and 2020. Searches used the following keywords: opioid use disorder, buprenorphine, Suboxone, medication-assisted treatment, medications for opioid use disorder, and emergency department. We examined reference lists in identified sources to “snowball” additional sources, including those published prior to 2015. We identified papers that analyzed efforts to change practice, described implementation, and described barriers and facilitators to ED-based OUD treatment. Our search identified 56 articles describing or testing ED-based efforts to address OUD and all but three were peer reviewed manuscripts. We excluded fourteen papers because they focused exclusively on harm reduction efforts, which were beyond the scope of this paper. Thirty-two papers specifically addressed MOUD in EDs; thirteen of these examined effectiveness. Ten manuscripts focused on peer supports in EDs; six of these reported on effectiveness. We describe findings from the literature in the section on evidence of effectiveness of ED-based interventions to address opioid misuse and the section on metrics to determine effectiveness. This study used both the literature and interviews to identify barriers and facilitators to offering these programs.

The study team identified potential subject matter experts first by scanning recent conference proceedings and highly cited articles in the literature. Through initial expert interviews, we identified additional informants. Using an interview guide, the research team conducted semistructured interviews between November 2019 and January 2020 with seven emergency medicine and addiction medicine physicians and researchers who have developed and/or are key leaders within ED-based OUD programs. Experts were based in academic and community health centers, EDs, and addiction medicine bridge clinics, and all were part of programs that offered MOUD. Via telephone conversations with the authors, the research teams asked respondents to provide an overview of how their ED-based OUD treatment program worked; to identify key features of the program; and to identify barriers and facilitators to implementation, operation, and sustainability. The team analyzed the interviews for major themes and insights using an inductive approach. We identified aspects of program design and key features by applying a framework analysis approach to the literature and interviews (Ritchie & Spencer, 1994).

The research team used the findings from the interviews to identify key features of ED-based programs to offer MOUD treatment and to provide context to information gathered in the literature.

3. Results

We found evidence in the literature of efforts to initiate OUD medications in the ED and/or connect patients from the ED with outpatient care operating in more than 80 EDs. Interviewees estimated that there are many more such programs across the nation and expect continued development and implementation. Efforts fall into one or more of three overall approaches: those that initiate and/or maintain patients on MOUD; those that offer peer support and assist with connections to treatment; and those that consist of harm-reduction efforts (i.e., providing naloxone as a take-home emergency medication). In this context, peers are individuals with lived substance use disorder recovery experience (McGuire et al., 2020). While harm-reduction efforts such as distributing naloxone to those who use opioids are important, feasible, and recommended practices (Adams, 2018; Hawk, Vaca, & D’Onofrio, 2015; Samuels, Baird, Yang, & Mello, 2019; Strayer et al., 2020), they are beyond the scope of this study. Our analyses focus on the first two approaches. Within these overall approaches, programs vary in several ways.

Key features are aspects of programs that vary across programs and may influence who receives services, program outcomes, and program sustainability. We identified five key features: 1) methods for patient identification, 2) treatment approaches, 3) program structure, 4) program relationship with community partners, and 5) program financing and sustainability. Table 1 lists these features and we describe them next.

Table 1.

Key features of ED-based interventions for OUD.

Key feature	Description	Examples
Patient identification	How are patients identified in the ED as possibly needing OUD treatment?	Screening all ED patients, rely on individual clinician identification, use of real-time algorithms to flag potential cases, patient self-identification in response to community outreach or signs and pins posted throughout ED
Treatment approaches	How are OUD treatment services delivered?	Offer medication vs only referral to community-based treatment; Medication typically used (buprenorphine vs others); initiate in ED or provide buprenorphine take-home doses; dosing protocols
Program structure	How is the ED-based program structured (e.g. hours, staffing, coordination)?	Capacity, 24-hour services vs. shorter hours; model to support coordination (e.g., recovery coaches embedded or available, involvement of hospital-based social workers or clinicians); how programs coordinate care beyond ED; health system include an outpatient opioid treatment program or hospital-based bridge clinic
Relationship with community treatment programs	How does the program integrate with other community programs?	Availability of outpatient treatment in community, strength of connection with outpatient partner; warm handoff or referral to community partner; follow-up strategy with the patient and/or community partner; actively working to partner with EMT organizations, law enforcement, or social services for referrals to identify patients eligible for OUD treatment engagement
Program financing, sustainability, maturity	How mature is the program, how is it financed, is it sustainable?	Individual hospital funded, state program, federal funds; operating within an ACO; Plans for sustainability past grant period if applicable, time implemented

3.1. Key Features of ED-based interventions to address OUD

3.1.1. Patient identification

ED-based OUD programs have varying approaches to identifying patients who might be appropriate for OUD treatment; this feature is one of the most important, as it influences program reach. Patient identification is often a prerequisite to notifying peers to reach out to a patient (McGuire et al., 2020). Identification methods range from universal screening, to identifying individuals with OUD even when they present for other reasons, to relying entirely on patient self-identification, for example in response to a hospital informational campaign (Edwards et al., 2020). One multi-site program described in an interview that it uses highly visible signs posted in the ED to educate patients and families about treatment options so that they can feel comfortable about self-identifying. One expert reported that their program, “takes patients with few rules and restrictions/requirements”, making it a low-barrier option available to anyone. Another expert described a text alert system that uses medical record data to identify patients who may be candidates for buprenorphine and automatically alerts clinicians and peers. They reported that the text alert system was helpful because it “took out many of the steps required for a clinician to understand whether a patient was a good candidate”, which minimized disruption to patient flow.

3.1.2. Treatment approaches

The majority of programs that we identified focused on initiating buprenorphine. However, some programs link patients to outpatient treatment without offering medication. Multiple interviewees said that it is important that all three OUD medications (buprenorphine, methadone, XR-naltrexone) be available in the ED. One interviewee shared that “offering only buprenorphine is not enough, methadone should be offered as well.” Regardless of approach, programs sought to engage patients in ongoing care, “We need to use EDs as a way to enroll people into treatment and get them to the care, that’s the end game. Therefore, one dose of buprenorphine provided in the ED is not sufficient and programs must focus on making sure patients also receive a prescription for their next dose.” To connect patients with ongoing care, EDs usually facilitate “warm-handoffs” to specialized treatment providers, including hospital-based bridge clinics and community-based outpatient providers.

Specific protocols for initial MOUD induction vary across programs. While some organizations developed guidelines to initiate MOUD in the ED, no national standard exists for starting OUD treatment in U.S. EDs (American College of Emergency Physicians Clinical Policies Subcommittee on Opioids, 2020). Clinical experts in our interviews reported struggling with existing induction protocols. Some argue that recommended doses of buprenorphine for induction are too restrictive and low, and that patients are more likely to agree to treatment if they receive medication earlier in their withdrawal experience.

3.1.3. Program structure

Program structure includes where the program is based (ED vs. psychiatry or addiction medicine department), the staff engaged in OUD treatment (e.g., physicians, nurses, social workers, peers), availability of these staff (e.g., peers based in the ED or available through other means, days/hours available), and the model for coordinating care beyond the ED (e.g., hospital-based bridge clinic, warm-hand-off or referral to community-based treatment). Several interviewees said programs are more successful when initiated by the ED rather than a behavioral health department because ED providers are more likely to participate and change their behavior. One interviewee explained ED providers need to see, “champions who are from the ED, and people who are role modeling so that providers are seeing it every day, from nursing, pharmacy, physicians, social workers, counselors”.

Peer support services are a key part of many ED-based efforts to address OUD. Peer services include education about treatment options such as medication treatment, emotional support, assertive outreach, and assistance overcoming barriers to treatment such as transportation or insurance; these efforts are the foundation of many ED-based programs. The peers providing services in the ED may also be called peer navigators, peer counselors, or peer recovery coaches and may be employed by the hospital or a community-based organization. Peer support services can be implemented both as part of a comprehensive program of ED-based OUD treatment that includes medication initiation and as a standalone intervention to help connect patients to outpatient treatment.

All programs described in the literature reported peer recovery coaches/navigators, a key point that interviewees confirmed. One interviewee reported that peers were “often more effective” than clinicians in getting patients to take buprenorphine. This expert described an anti-buprenorphine sentiment in their patient population that providers struggled to overcome. Peers accomplished this. A particular challenge is that some rural hospitals do not have the volume to support having ED-based peers and instead may rely on peers from a central location who travel to hospitals within an hour drive or use peers available via tele-health.

3.1.4. Relationship with community providers

Several interviewees reported that having a reliable referral option is critical to making ED physicians willing to initiate MOUD. One interviewee explained the importance of a reliable referral options for ED providers, “We have a bridge clinic that takes the initial follow up appointment so that the ED has a single point of referral that never changes. Every referral is a good referral, there is no criticism, and everything goes smoothly for the ED provider”. This interviewee reported that this is the most important feature of successful ED-based programs.

Some ED-based programs actively try to serve as a resource to the community, proactively reaching out to providers who interact with individuals with OUD, and “by being a 24/7 opportunity to initiate treatment for those centers (e.g., ex-detox programs, residential treatment, primary care providers) with no access to prescribers, or who do not feel comfortable treating those patients”.

3.1.5. Program financing and sustainability

ED-based OUD treatment programs are generally financed in several ways, including standard billing and reimbursement and grant funding. Clinical experts reported that financing an ED-based program is not difficult when patients have health insurance that covers treatment. One interviewee said, “if buprenorphine is covered without prior authorization or delay, the program can be self-sustaining.” The challenging part, however, is funding peer support services in the absence of grant funds, because “the least costly staff in the ED are the hardest to pay for”. When peer support services cross different institutions and services lines, billing is a challenge. Clinical experts reported it would be helpful for all insurers to pay for peer services.

Many existing ED-based efforts started with grant funding. The Drug Addiction Treatment Act of 2000 (DATA 2000), which first enabled office-based prescribing of buprenorphine for OUD, requires that most providers complete training before prescribing buprenorphine, and be waivered to prescribe buprenorphine through the Drug Enforcement Administration (DEA). Some grants provided funding for clinicians to obtain the DATA 2000 waiver to prescribe buprenorphine and pay for peer support services. Experts said while ED physicians are able to provide buprenorphine for a limited amount of time without a waiver, waiver training was nonetheless important because many ED physicians did not feel comfortable initiating treatment without it.

Experts disagreed about funding streams necessary to start ED-based OUD treatment. One said that it is only possible to start with grant funding, while another argued that grant funding is not a prerequisite because insurance can be billed and sustainability can be improved using nurse practitioners. Experts agreed that if grant funding was used, showing the program’s value to the community encourages the hospital to maintain the program. One interviewee explained that tracking and sharing data with hospital administrators and with the community has generated good will for the program and facilitated obtaining necessary funding from the hospital.

3.2. Evidence of effectiveness of ED-based interventions to address opioid misuse

Existing evidence of the effectiveness of ED-based MOUD initiation is based on one randomized trial, pilot studies, and several small observational studies (see appendix). In the randomized trial, individuals receiving ED-based initiation of buprenorphine with referral to primary care had higher rates of clinician-reported 30-day treatment engagement and larger declines in drug use than those receiving referral only or brief intervention and referral. Further analyses following 88% of the randomized sample show at 60 days, individuals randomized to the buprenorphine group were more likely to be engaged in treatment compared to the referral only or brief intervention and referral groups (D’Onofrio et al., 2017). The study identified no differences in treatment engagement at 6 and 12 months or in drug use and HIV risk behavior among the three groups (D’Onofrio et al., 2017). The D’Onofrio trial informed a pilot implementation effort in three South Carolina hospitals that included implementing universal screening, extensive training, and hiring peer supports. Following implementation, approximately 45% of eligible patients elected to received buprenorphine and of these, 46% were retained in treatment after 30 days (Bogan et al., 2020).

The eight observational studies are small, ranging from 18 (Dunkley et al., 2019) to 950 (Monico et al., 2020) patients. Four studies examined buprenorphine initiation rates. These studies provide support for the feasibility of ED-based buprenorphine induction, but the variation shows that more information is needed to understand features that influence effectiveness, standardize measures, and identify the best way to implement such a program.

We identified five studies of ED-based peer support services for individuals with OUD. Existing research shows that peer support services are feasible to implement in the ED and acceptable to patients (Waye et al., 2019), associated with shorter time to treatment engagement (Samuels et al., 2018), and patients are more likely to be discharged with a referral to treatment when they meet with a peer (Samuels, Baird, Yang, & Mello, 2019).

Forthcoming evidence

In addition to published literature, several efforts are in process to rigorously study ED-based approaches for OUD. For example, the EMBED trial is testing a clinical decision support tool to help providers offer MOUD in 20 EDs across five U.S. health systems (Melnick et al., 2019). A pragmatic multi-site cluster randomized trial is underway in Indiana (Watson et al., 2020). The ANCHOR trial of peer support for individuals presenting to EDs is underway in Rhode Island (Goedel et al., 2019). The NIDA Clinical trials network is studying an implementation facilitation model to test whether it improves ED-based MOUD in four urban EDs (D’Onofrio et al., 2019). Massachusetts Health Policy Commission is funding an evaluation of nine hospitals’ efforts to expand ED-based treatment (The Massachusetts Health Policy Commission, 2018). An effort in Pennsylvania to improve opioid care in hospitals broadly is also underway (Kilaru et al., 2020).

3.3. Metrics to examine effectiveness of ED-based efforts for OUD

Efforts to examine the effectiveness of ED-based OUD programs employ a variety of patient-level process and outcome measures and system-level measures of health care system structure. Table 2 lists measures that recent studies have used. Process measures are most common, primarily engagement in outpatient treatment at various time points following initiation of MOUD. Although validated quality measures have been developed related to OUD identification and treatment (National Quality Forum, 2019), the studies that we identified generally do not use these measures, but instead use de novo customized measures. Outcome measures are more difficult to track because often follow-up information is limited to a subset of patients engaged in treatment, and the studies have no controls.

Table 2.

Metrics used to examine effectiveness of ED-based OUD treatment.

Measure	Data Source	Peer Reviewed Study
ED-based medication initiation studies
Patient experience measures
Patient experience	Patient self-report	(Snow et al., 2019)
Process measures
Medication initiation	Hospital EMR	(Bogan et al., 2020)
	Hospital EMR	(Holland et al., 2020)
	Hospital EMR	Kaucher, 2019
	Hospital EMR	(Kelly et al., 2020)
	Hospital EMR	(Monico et al., 2020)
Referral to outpatient treatment	Hospital EMR	(Holland et al., 2020)
Initial follow-up with outpatient treatment	System EMR	(Dunkley et al., 2019)
	Outpatient tx pgm	(Bogan et al., 2020)
	Outpatient tx pgm	(Edwards et al., 2020)
	Hospital EMR	(Monico et al., 2020)
Engagement in formal outpatient addiction treatment at various timepoints (e.g., 30-days, 2-, 6-, 12- month follow-ups)	Clinician report	D’Onofrio 2015
	Clinician report	D’Onofrio 2017
	System EMR	Dunkley et al., 2019)
	Outpatient tx pgm	MA HPC, 2018
	Outpatient tx pgm	Bogan 2020
	Outpatient tx pgm	Edwards 2020
	Outpatient tx pgm	Kelly 2020
	Medical center data	Kaucher et. al., 2019
	Clinician or patient self-report	Srivastava, 2019
HIV risk assessment score	Patient self-report	D’Onofrio 2017
Wait time for outpatient treatment	Primary data collection	(Kawasaki et al., 2019)
Outcome measures
Number of days of illicit opioid use in the past 7 days	Patient self-report	D’Onofrio 2017
Opioid-negative urine test results (2 and 6-months)	Primary data collection	D’Onofrio 2017
Mortality	Hospital EMR	MA HPC, 2018
Lethal and non-lethal overdose	Hospital EMR	MA HPC, 2018
Presence or absence of adverse medication effects	Hospital EMR	(Edwards et al., 2020)
Structural measures
Number of newly waivered providers	Primary data collection	Kawasaki 2019
Provider buprenorphine induction adoption rate	Hospital EMR	Holland 2020
Provider knowledge and ability to treat OUD	Survey data collection	Kawasaki 2019
ED 30-day revisit	Hospital EMR	MA HPC, 2018
	Hospital EMR	Berg 2007
Patient received naloxone from ED	Hospital EMR	Bogan, 2020
	Hospital EMR	Holland 2020
Peer Support Studies
Process measures
Accept referral to treatment	Program data	Waye 2019
	Program data	Welch 2019
Engaged in MOUD (initial and 6 month)	Private insurance and Medicaid claims	(Watson et al., 2020) Welch 2019
Engagement with a formal SUD treatment program within 30 days of the initial ED visit	State SUD administrative data & PDMP data	(Goedel et al., 2019)
Initiation of OUD medication	State PDMP data	(Samuels et al., 2018)
Outcome measures
Recurrent ED visit(s) for a suspected opioid overdose over 18 months	Statewide EMR & DPH surveillance data	Goedel 2019
ED visit for overdose	Statewide EMR	Samuels 2018
1-year mortality	State Vital Records Database and National Death Index	Samuels 2018

EMR: Electronic medical record; PDMP: prescription drug monitoring program; DPH: Department of Public Health

3.4. Facilitators and barriers to implementation of ED-based OUD treatment

Interviewees identified several aspects of ED culture and program design as facilitators for implementing ED-based OUD treatment, including having an ED champion, a reliable referral network, and systematic tracking and reporting of data. One clinician/administrator explained, “ED champions who model the work so that other providers see it every day will help motivate other ED providers to offer OUD treatment”. Another expert said that having the leadership and staff come from within the ED rather than from another department is important. They explained, when staff are assigned to the ED but structurally reporting elsewhere, it can hinder growth and success. In the absence of an ED champion, a clinician reported their program was successful when designed to make the work of ED providers easier, by helping to move patients out of the ED to a bridge clinic, with the support of an institutional champion at a higher level than the ED.

Experts agreed that a bridge clinic or reliable outpatient program, available to accept patients initiated in the ED, is critical. Several experts reported ED physicians will not offer treatment without having a place to refer, and they need to feel confident that they will not have problems with the referral or handoff. One interviewee reported difficulty establishing referral relationships for ongoing community treatment because most community programs in their area were abstinence-based. Without a sure place to refer, ED providers would not try to start patients on MOUD.

Studies suggest that systematically tracking treatment process and outcome data and feeding this information back to providers is crucial (Bogan et al., 2020). Interviewees supported this idea. For example, one expert tracks and reports the number of patients who receive a buprenorphine prescription among those initiated in the ED, to inform efforts to increase this rate. They said sharing feedback and success stories about people who successfully engage with outpatient treatment can generate excitement and motivation for this work among ED providers.

Through our interviews we also identified several barriers to initiating OUD medication treatment in the ED, including lack of training, regulatory requirements, difficulty obtaining detailed treatment history, ED culture, stigma, and lack of resources. Clinical experts reported ED physicians feeling unprepared to treat OUD and that this was addressed by offering waiver trainings. At the same time, clinical experts reported DATA 2000 regulatory requirements make prescribing difficult because of the need to spend time and money on training, and providers may not feel that they have the necessary expertise if they do not complete waiver training. One clinician reported that their hospital was able to overcome this reluctance through extensive education about the “three day rule” (Kaucher et al., 2019), which allows ED physicians to provide patients buprenorphine for three days without obtaining the waiver. One clinician explained that another barrier is the exclusion of methadone information from prescription drug monitoring program data. They said that patients may say they are not currently on methadone, but providers do not know if it has been days or weeks since the patient had methadone. One interviewee reported, “The tribalism of medicine and the idea that the ED should not be involved beyond the initiation step or in long-term care” is a barrier to serving patients. This clinician would like to see ED programs extend beyond initiating treatment and connecting patients to care. Finally, hospitals with fewer patients with OUD may have a difficult time sustaining staff to offer peer services for OUD. One interviewee argued that peer interventions can be broadened to address all substance use disorders as a way to streamline work for ED clinicians and allow peers to help more patients.

We identified three studies in the peer reviewed literature and one doctoral dissertation examining barriers and facilitators to ED-based OUD treatment. A mixed methods study in four EDs across the United States identified several barriers, including lack of training and experience using buprenorphine, lack of established protocols, waiver requirements and a lack of understanding regarding regulations, provider concerns about whether the patient would be able to engaging in ongoing treatment, a lack of support and resources from leadership, and provider perception that treating OUD was outside the scope of the ED (Hawk et al., 2020). A survey of 100 ED clinicians in one academic medical center found that providers think they should be able to offer buprenorphine, but do not feel they have adequate training in using buprenorphine (Im et al., 2020). Through interviews with physicians, the authors found that most providers reported offering referrals to substance use treatment and did not provide buprenorphine in the ED. Facilitators that these studies identified include training, protocol development, and receiving feedback on patient experience and outcomes (Im et al., 2020)(Hawk et al., 2020). Both studies identified ED culture and physicians’ perspectives on the role of EDs in treating OUD as barriers. Historically, EDs have limited their role to referring individuals with substance use disorder for treatment and have not provided treatment in the ED; in many EDs this remains true (Marshall, Derse, Iserson, Kluesner, & Vearrier, 2019; Teferi et al., 2020). As a result, providers may not see treating OUD as part of their role.

4. Discussion

Emergency department–based efforts to address opioid use are expanding across the country, but programs are early in their evolution and vary across many dimensions. In a recent study, only 21% of ED providers reported readiness to initiate buprenorphine for patients in the ED with OUD (Hawk et al., 2020). Multiple studies identify a need for education, training, protocols, and reliable referral sites to improve provider readiness (Hawk et al., 2020; Im et al., 2020). Evidence of program effectiveness is limited by the small number of studies, but more work is in process. Although the D’Onofrio trial identified short-term effectiveness at 30 days and 2 months, but did not find evidence of effectiveness at 6 or 12 months following initiation, the intervention concluded after 10 weeks and referred patients to nonstudy community treatment settings or, when requested, supported them through discontinuing medication and connecting with outpatient care. Thus, the study did not test effectiveness of ED initiation followed by long-term community-based treatment; more research is needed to investigate this approach.

The key features of ED OUD programs that this paper identified may influence success and outcomes of ED-based programs. For example, the approach to patient identification affects access to care and may also influence evaluation results because programs will vary in the type of patient they identify and this may affect outcomes. Availability of different medications and opportunities to engage with peers may influence patients’ willingness to engage in treatment.

For several key features of ED OUD programs, limited consensus exists about the best approach. For example, we identified a range of strategies to identify patients, from passive information posted, to only approaching those who present with an opioid-related event, to universal screening. In addition, little consensus exists about when to intervene and how to treat. EDs that offer MOUD may administer the medication in the ED, send patients to a bridge clinic, or send the patient home with the medication or a prescription and instructions for taking it when withdrawal progresses. Systematic information, including patients’ and providers’ perspectives, on how these different approaches work and for whom would help to improve access to and quality of care. Finally, researchers should carry out additional analyses of the cost-effectiveness and sustainability of programs. Many programs report that they have anecdotal information about their effectiveness, but without more rigorous data on effectiveness and cost, providers may find it difficult to support an argument to maintain efforts, especially when grant funding is no longer available.

This analysis is limited to a set of interviews with clinical experts and existing literature. We spoke with eight clinicians who are actively working to develop and improve ED-based OUD programs that offer or connect patients with medication treatment. While their views may differ from other providers in the field and in particular do not represent programs that do not offer medication treatment, their experiences and insight may be helpful for others developing similar programs. It is also important to learn from individuals in the community who use substances. Future work should include their perspectives. Existing effectiveness literature is limited, does not consider programmatic factors that may influence effectiveness, and may be subject to publication bias as many programs are not described in the literature. Studies tend to be small and limited to single institutions. With the exception of the D’Onofrio trial, studies do not employ control groups and most are only able to follow patients who engage in treatment. Literature is also limited by the lack of standardized metrics for measuring effectiveness of ED-based OUD programs. Most programs are measuring engagement in outpatient treatment following the ED visit, but are not consistently using the same measure. A multi-stakeholder group proposed a quality framework that adopts structural, process, and outcome measures, including the proportion of ED OUD patients initiated on medication treatment and linked to care (Samuels, D’Onofrio, et al., 2019). As ED-based OUD programs continue to develop and researchers conduct more studies, a consensus on the best measures would help other researchers and treatment provides compare studies.

5. Policy implications

Our review identified several barriers to increasing ED-based OUD treatment, as well as some steps that have or could be taken to increase access. State and federal governments should consider how MOUD regulation affects treatment availability and consider modifications that further promote access while ensuring patient safety. Policy-makers can require EDs to provide MOUD—Massachusetts took this approach in 2018 (An Act for prevention and access to appropriate care and treatment of addiction. MA-HR. 4742. 190th General Court., 2017). Some steps have been taken to increase the available workforce that can prescribe: physicians assistants and advance practice nurses are now allowed to complete DATA 2000 waiver training to prescribe buprenorphine. Reduced restrictions for offering methadone would make it easier for EDs to support patients in starting any of the three medications approved for OUD. Policy changes made due to the COVID-19 pandemic have included changes in telemedicine requirements for providers, HIPAA allowances for additional forms of communication between providers and patients, changes in 42 CFR Part 2 to allow more data sharing, and flexibility in medication regulations (e.g., take home methadone) (CARES Act, 2020). These changes may provide an opportunity to expand access to MOUD. In addition, stigma is an important problem to address. Interviewees and the literature identified educational efforts to reduce stigma around MOUD among front-line providers and patients as an ongoing need.

Financing mechanisms to support reliable linkages between acute care hospitals and community providers are critical to successful ED-based OUD treatment. Payers could use reimbursement strategies that facilitate these linkages through incentive or bundled payment initiatives. This method would help providers currently offering ED-based MOUD and support additional providers in adding ED-based MOUD. Payers could also improve coverage and reimbursement for peers. Many of the ED-based OUD treatment programs use peer support providers, but funding for these roles is often through temporary grants. When these funds are no longer available, how hospitals will pay for peer support services is unclear.

6. Conclusion

Access to OUD treatment needs to be expanded. Emergency departments can play an important role by helping to initiate MOUD in a low-barrier setting. Many programs are in development and early implementation. As programs evolve, attention to key features that we identified here may improve effectiveness and help to explain variation in outcomes.

Highlights.

• Emergency departments can play an important role in improving access to opioid treatment

• ED-based efforts to address opioid use vary in key ways including how to identify and treat patients

• Strong referral networks and systematic data reporting are needed to improve programs

• In future work it will be important consider program key features to understand effectiveness