Recovery Coach Program Implementation Across an Integrated Health System

In 2014, MGH launched a substance use disorder initiative with the mission of making treatment readily available, standardized, and of high value across its system to ensure that patients could receive addiction treatment seamlessly, regardless of setting. This initiative required integrating addiction services into the emergency department, hospital inpatient units, and primary care practices and starting a new low-barrier outpatient addiction treatment program called the Bridge Clinic, which was designed to bridge any remaining gaps in the system by offering walk-in, immediate-access services. PRCs were made available in all of these settings (6–9).

Early data from the MGH initiative were examined to assess the impact of PRCs as well as integrated care models that included PRCs. A qualitative study of PRCs demonstrated the benefit patients found in working with them. Patients appreciated how accessible PRCs were, their shared lived experience, their help with system navigation, and their support with regard to behavior change (4). In one assessment, patients newly connected to PRCs experienced a 44% decrease in hospitalizations, a 9% decrease in emergency department visits, and a 66% increase in outpatient encounters in the 6 months after initiating recovery coaching compared with the 6 months prior to the first coaching contact. Among patients treated with buprenorphine, being engaged with a PRC was associated with greater odds of opioid abstinence and ongoing buprenorphine treatment (6). A retrospective cohort analysis of patients with a substance use disorder who received primary care at a site with PRCs and addiction nurses compared those patients with patients who were treated at practices without PRCs and addiction nurses; the authors found a mean difference of 7.3 fewer inpatient hospital bed days per 100 patients and a lower number of emergency department visits for the patients with access to PRCs and addiction nurses. Although this evaluation was not limited to the effect of PRCs, these professionals were a component of the intervention. Buprenorphine and naltrexone prescribing rates were also higher at sites with PRCs and addiction nurses (7). These studies provided a promising demonstration of the potential of PRCs to contribute to higher-quality and lower-cost care for patients with substance use disorders.

Deployment of PRCs throughout the MGB system began in 2017 as part of a larger strategy to optimize addiction care, which also involved increasing primary care providers’ treatment competency through addiction medicine case conferences, buprenorphine training, electronic specialist consultations, peer mentorship, and adaptations to the electronic health record system to integrate universal substance use screening and clinical decision support. This system expansion adopted best practices from the MGH program, including the dual supervision model. A systemwide recovery coach collaborative was launched to create a space for PRCs and site supervisors to discuss best practices and build cohesion across sites. PRC documentation was also standardized with an electronic health record template that allowed for streamlined notes and data extraction and that was linked to a dashboard to track patient-PRC contacts across the system.

Support for this expansion was bolstered by the transformation of the Massachusetts Medicaid system in 2018, when the full adoption of accountable care organizations (ACOs) further catalyzed efforts to address substance use (10). In an ACO environment, health care systems begin to assume increased financial risk and to be accountable for the cost and quality of patient care. Because of the high health care cost of untreated addiction and research showing that substance use disorder treatment reduces health care costs, addressing addiction emerged as a key priority for MGB’s ACO transformation. On the basis of MGH’s preliminary data, the expanded role of PRCs was incorporated as a cornerstone of the MGB substance use disorder treatment strategy.

Both the MGH and the MGB addiction recovery programs integrated PRCs as essential nonclinical care team members. PRCs focused on engaging and building relationships with patients, providing hope and support as patients worked toward their individual goals. PRCs were embedded in primary care, behavioral health, emergency department, inpatient hospital, Bridge Clinic, perinatal clinic, and community and street medicine settings.

Applying their own lived experience and perspectives, PRCs engaged with patients and helped them to surmount barriers to accessing clinical care and other services. The PRC model includes four domains: advocacy, mentoring and education, recovery and wellness support, and ethical responsibility. To maintain fidelity to the model, the PRC-patient relationship was a nonclinical one of peers. PRCs leveraged their lived experience and motivational interviewing to inspire change and provide mentorship. PRCs met with patients in clinical or community settings and also communicated via telephone. In addition, PRCs used text messaging to connect with patients who consented to such contact. PRCs enhanced the limited support that the traditional care team could offer to patients by providing valuable insight and education to their clinical colleagues. By sharing their lived experience with clinicians, PRCs provided a recovery-oriented perspective and a tangible reminder that most people recover from substance use disorders.

PRCs were supervised by both a PRC supervisor and a site supervisor, each of whom could be a nurse, physician, or behavioral health clinician. The PRC supervisor provided training and education opportunities, supported PRC self-care, and ensured fidelity to the PRC model.

In addition to building a supervision structure, preparing sites before embedding a PRC was critical to implementation. Site administrators were oriented to the role and function of PRCs, asked to find space where a PRC could work and meet with patients, and tasked with creating referral pathways. PRCs participated in team meetings and received referrals from providers. Policies and procedures were developed and refined for each site. The management team also introduced the concepts of confidentiality and boundaries as they applied to the PRC role and made teams aware of how clinical and PRC roles differed. For example, although coaches document their work in the electronic medical record, their documentation is minimal to protect patient confidentiality and trust.

PRC applicant requirements included having lived experience of addiction; having demonstrated the skills needed to complete 2 years of sustained remission; and having attended the Recovery Coach Academy, a state-run, 5-day training course. If an applicant had not yet attended this training course but was otherwise an ideal candidate, the applicant was permitted to attend after they were hired. PRCs were also required to work toward becoming a state-certified addiction recovery coach, which entailed attending specific training sessions, logging a set number of supervision hours, and passing a standardized examination.

Between July 1, 2017, and October 31, 2021, MGB had 23 full-time PRCs who received 6,785 referrals and enrolled 5,662 participants, an enrollment rate of 83.4%. Some PRCs worked in multiple settings. For example, a PRC might work in the Bridge Clinic and a primary care practice or in an inpatient hospital setting and the emergency department. Of the 23 PRC positions, the equivalent of five were in the Bridge Clinic, eight were in primary care practices, two were in inpatient hospital settings, two were in emergency departments, four were in behavioral health settings, one was in a perinatal clinic, and one was in street medicine. The most common referral sources were social workers, emergency departments, word of mouth, and primary care providers. Among patients referred, 25.5% (N=1,730) were enrolled in the MGB Medicaid ACO, 14.1% (N=957) were enrolled in a commercial ACO, 9.2% (N=624) were enrolled in a Medicare or Medicare Advantage ACO, and the remainder did not have insurance. Patients were predominantly White (78.6%, N=5,333) and non-Hispanic (82.3%, N=5,584); 8.4% (N=570) were Black and 12.9% (N=875) identified as Hispanic. The race of referred patients varied considerably by site. For example, the race of referred patients who identified as White ranged from 50% to 89% across sites, those who identified as Black ranged from 3% to 24%, and those who identified as “other” race ranged from 2% to 16%.

The substance most frequently used by patients referred to a PRC was alcohol, followed by heroin or fentanyl, cocaine, prescription opioids, benzodiazepines or sedatives, cannabis, and tobacco. The primary addictive substance used by referred patients varied somewhat over time, although alcohol and heroin or fentanyl remained the two most common types of substances used by those referred to PRCs. Type of substance used was similar among patients who were successfully enrolled with a PRC compared with those who were referred but not enrolled.

In calendar year 2022 across MGB, there were 11,671 primary care patients with alcohol use disorder, 5,171 with opioid use disorder, and 6,929 with other drug use disorders. Among patients with alcohol use disorder, 55.4% (N=6,466) were engaged with a therapist, an office-based addiction treatment nurse, or a PRC, and these patients completed 5,467 contacts, either in person or virtually, with a PRC. For patients with opioid use disorder, 62.6% (N=3,237) were engaged in care and completed 2,784 PRC contacts. Among patients with other drug use disorders, 55.7% (N=3,859) were engaged in care and completed 2,527 PRC contacts. In calendar year 2023, an additional 14,032 PRC contacts with patients occurred across the system.

In response to pilot data demonstrating the effectiveness and potential value of PRCs through patients’ reduced acute care use, we were able to successfully implement a recovery coach program across a large, integrated health care system. This program supported patients with addiction, provided the clinical team with the perspective of individuals with lived experience, and facilitated the expansion of addiction treatment integration within general medical settings. To our knowledge, ours is among the first health care systems to have PRCs provide both acute and longitudinal support to patients across a range of care settings, including emergency departments, inpatient hospital settings, primary care practices, a perinatal clinic, and outpatient behavioral health clinics.

The development and implementation of our recovery coach program was facilitated by key resources and components. First, having senior leadership support was crucial. This support ensured adequate funding for the program and provided a clear message to local sites that the model was supported. Leadership engagement was also helpful in addressing challenges that arose during the implementation process. Second, engagement of local stakeholders was critical and was best fostered when sites identified areas of need and PRCs stepped in as a source of help. Third, a strong central system of supervision helped support PRCs and reassure local site administrators that PRCs would have the infrastructure needed to succeed. Fourth, early data from MGH on positive patient outcomes and reduced acute care use made the case that PRCs contributed to value. Last, forming a recovery coach collaborative to share best practices helped maintain site engagement.

We encountered several challenges with this implementation. First, because of the way addiction is criminalized, many of our PRC applicants had experienced interactions with the criminal legal system, which were identified during the hiring process. We worked closely with human resources to develop flexible guidelines that allowed us to hire people with a recent history of incarceration. As part of the interview process, we involved a dedicated recruiter who understood the unique role of PRCs and could be a resource to candidates. Second, because of the chronic course of addiction, some PRCs experienced a recurrence of substance use that affected their ability to work. Such recurrences highlighted the importance of promoting staff wellness, including making PRCs aware of available resources to support their own remission or to get immediate help in the event of substance use recurrence. We collaborated with occupational health services, human resources, and our employee assistance program to compile guidelines for taking a full or intermittent leave of absence to access treatment. Last, the offsite, on-demand nature of PRC work posed administrative and managerial challenges. We created a system that ensured accountability for timekeeping but also allowed flexibility. For example, instead of signing in and out at the end of a shift, PRCs could submit an electronic form to notify their manager when they started or ended work. PRCs were encouraged not to work at times other than their regularly scheduled hours to ensure work-life balance, but if they did provide support to a patient after hours, they could submit those hours worked on the standard electronic form to ensure they were paid.

Despite these barriers, implementation of this large recovery coach program across a range of clinical settings was feasible and successful. Early evaluation of the program and of PRCs as a component of primary care integration, as well as ongoing review of a PRC dashboard summarizing patient contacts by site and by coach, helped maintain leadership support. PRCs have supported thousands of recoverees across our system since the launch of the program and have served as the “human glue” of this work, offering critical support to patients at a cost lower than that incurred by many other interventions. Ongoing evaluation of how PRCs facilitate improved clinical outcomes and reduced cost will be important for further dissemination of this model and for its long-term sustainability. Amid the ongoing crisis of substance use, integrating PRCs into general medical settings may be an important component of multifaceted, value-based interventions that address addiction.

The authors acknowledge the leadership and collaboration of the leaders across Mass General Brigham substance use disorder programs and the tireless work of the peer recovery coaches.