Building Behavioral Health Homes: Clinician and staff perspectives on creating integrated care teams

Abstract

Adults with serious mental illness and substance use disorders have elevated risk of mortality and higher healthcare costs compared to the general population. As these disparities have been linked to poor management of co-occurring chronic conditions in primary care, the behavioral health setting may be a preferred setting for routine medical screening and treatment. This qualitative study describes early stages of integrating care teams in emerging medical homes based in mental health and addictions treatment settings. Clinicians and staff from 10 agencies engaged in the Behavioral Health Home Learning Collaborative participated in qualitative interviews exploring local definitions of “Behavioral Health Home” and initial barriers and facilitators to integration. Facilitators included clear staff roles, flexible scheduling, and interdisciplinary huddles and staff trainings. Challenges included workforce, limited use of electronic health records, and differing professional cultures. Participants advocated for new workflows and payment structures to accommodate scheduling demands and holistic case management.

Introduction

Frequently, behavioral health (BH) service locations are the main point of healthcare contact and continuity for patients with serious mental illness (SMI) and substance use disorders (SUD).¹ Integrating primary care (PC) services into BH settings has demonstrated improvements in health outcomes, quality of care, and health care utilization.^2–5 The patient-centered medical home (PCMH) model provides a framework for person-centered, coordinated, comprehensive care to effectively integrate additiction treatment and BH services with primary care.⁶ Section 2703 of the Affordable Care Act allowed states to design “health homes” for Medicaid beneficiaries with chronic conditions, and there is great national interest in developing a variant of the medical home model to serve SMI/SUD populations.⁷ PCHMs must provide services, such as comprehensive care management, care coordination, individual and family support, and use of health information technology to link services.^6,8

Best practices for integration are usually described for integrating BH clinicians into PC settings.⁹ Few studies have identified challenges and facilitators specifically orientated to integration of PC clinicians into BH settings. The Substance Abuse and Mental Health Services Administration’s (SAMHSA) Primary and Behavioral Health Care Integration program (PBHCI)¹⁰ has identified challenges including staff recruitment and retention, data collection, licensing, physical space, and patient enrollment.¹¹ Other research cites challenges associated with defining staff roles, obtaining medical equipment, implementing new workflows, and building relationships with PC agencies.^3,12

In May 2014, The Oregon Health Authority (OHA) launched the Behavioral Health Home Learning Collaborative (BHH LC) to support BH and addictions treatment agencies working to increase access to PC services for persons with SMI/SUD. The BHH LC was funded through the Adult Medicaid Quality Grant Program as part of a broader improvement project intended to increase the proportion of Medicaid enrollees in Oregon’s patient-centered medical homes, known locally as “Patient Centered Primary Care Homes” (PCPCH). The BHH LC was loosely structured around the Insitute for Healthcare Improvement’s Collaborative Model, in which hospitals or clinics embarked on similar improvement projects participate in a “short-term (6- to 15-month) learning system.”¹³ The BHH LC was originally planned as a 1-year project, but the scope of change and breadth of challenges across agencies and implementation models required extensive support, and OHA extended the collaborative for two additional years, with a no-cost extension from the funder.

During the second year of the BHH LC, the project’s lead investigators from ORPRN and OHA authors conducted qualitative interviews with 10 participating agencies to elucidate local definitions of a “Behavioral Health Home” and begin to identify best practices for integration. The analytic approach in the present study, was based on Peek’s 2008 operational, clinical, and financial framework for successful integration.¹⁴ Although other studies have identified challenges and facilitators for building care teams, they have not used this framework in the BH setting.

Methods

BHH LC sites

OHA contracted with the ORPRN to provide practice coaching, formative evaluation, and curriculum development for all three years of the BHH LC implementation. The principle intervention of the BHH LC was intensive, individualized practice coaching to help participating sites design and implement their chosen integration model. Other activities included seven in-person learning sessions, specialized cross-training, and webinars. ORPRN’s Practice Enhancement Research Coordinators, or practice coaches, used a range of organizational development, project management, quality improvement, and practice improvement methods to help participating organizations conduct improvement activities.

From its inception, the BHH LC focused on expanding access to the PCPCH model to high-need, high-cost populations with significant BH issues and SUDs. At end of year 2, most sites had begun to provide integrated primary care and behavioral health through one of three general models: co-location of PC and BH services without shared administrative oversight, in-house PC and BH services with shared administrative oversight, or facilitated referral to PC. Approximately 2,500 clients received integrated care across the ten participating BHH LC sites. Characteristics of participating BHH LC sites can be found in Table 1.

Table 1.

BHH LC Site Characteristics and Interview Participants

ID	Agency Type	Integration Model*	Estimated Patients Receiving Integrated Services N	Onsite Psychiatrist N (total FTE)	Onsite Mental Health Nurse Practitioners N (total FTE)	Onsite Primary Care Providers Type: N (FTE)	Weekly Clinic Hours N days (N hours/day)	Focus Group Participants N	Exit Interview Participants N
A	Federally Qualified Health Center	Co-location	1,200	2 (1.25)	4 (4.0)	MD/DO: 1 (.75) NP: 1 (1.0) PA: 1 (1.0)	5(10)	8a,b	5a,b
B	Community Mental Health Center	Co-location	100	2 (.9)	4 (3.0)	MD/DO: 2 (.1 -.2)	1(8)	NA	1b
C	Mental Health & Substance Use Disorder	Co-location	640	2 (0.5)	0 (0)	PA: 1 (1.0)	4(7)	4a, b	2b
D	Mental Health & Substance Use Disorder	Co-location	400	6 (5.5)	2 (2.0)	MD/DO: 1 (.5)	5(9)	1b	2b
E	Mental Health & Substance Use Disorder	Co-location	80	1 (0.1)	1 (.8)	MD/DO: 1 (.2)	1(6)	9b	5b
F	Mental Health & Substance Use Disorder	Co-location	300	1 (1.0)	0(0)	Various: 17 (8.2)	6(9)	7a, b	2a,b
G	Mental Health & Substance Use Disorder	In-house	375-500	0(0)	0(0)**	PA: 1 (1.0)	4(9)	8a,b	3a
H	Mental Health & Substance Use Disorder	In-house	475	0(0)	1 (1)	NP: 2 (1.0)	5(9)	4	2a
I	Community Mental Health Center	Referral	30	2 (1.75)	2 (1.75)	N/A	5(8)	3b	NA
J†	Community Mental Health Center	Referral	20	2 (1.5)	2 (1.5)	MD/DO: 1 (.2)	NA	5b	2b

^*These agencies utilized three integration models recognized by SAMHSA: in-house, co-located partnership, and facilitated referral.²²

^**Psychiatric medication management provided by physician assistant

^†Focused on an integrated buprenorphine program at partner FQHC

^aAttended by a prescribing primary care clinician (e.g., Nurse Practitioner, Physician Assistant, MD/DO)

^bAttended by a licensed behavioral health provider (e.g., LCSW, CADC, QMHP)

Data collection and analysis

In Year 1 of the BHH LC (July 1, 2014 through March 31, 2015), ORPRN provided curriculum design and program planning, on-site practice coaching, and contributed to project administration. In Year 2 (April 1, 2015-January 31, 2016), a formative evaluation component was added to the BHH LC program design as part its expansion.

Mid-year focus groups

During the second year of BHH LC, each site was asked to participate in a structured focus group interview to identify agency definitions of a behavioral health home, including interpretations of the core elements of a medical home and the utility of the CMS Adult Quality Measures to further the agency’s service delivery model. Nine out of ten sites participated in BHH LC focus groups. Group participation ranged from one to nine team members (see Table 1). Groups were led by the project’s lead investigators at ORPRN (2^nd author) and OHA. Practice coaches were not present during the focus groups.

End of year leadership interviews

At the end of the BHH LC’s second year, before the third year of the collaborative had been approved, “exit interviews” were conducted with leadership at each site. The exit interview guide included questions about the agency’s perceived benefits of participation in the learning collaborative, strengths and weaknesses of specific program components, and the agency’s experience using a practice facilitator. Between one and five team members, representing integration leadership, at nine out of ten sites participated in exit interviews with (see Table 1). Interviews were led by the project’s lead investigators at ORPRN (2nd author) and OHA. Practice coaches were not present during the focus groups.

Analysis

All interviews were conducted face-to-face, audio-recorded, and professionally transcribed. A unique numerical identifier was randomly assigned to replace participant names on transcripts. Transcripts were entered into the qualitative data management and analysis program Atlas.ti 7.5.10. Initial review of the transcripts by the authors revealed that interviews contained rich content describing barriers and facilitators of team-based implementation of the BHH, although this was not a specific topic on the interview guide. Informed by these formative discussions, the lead author coded all focus groups and Year 2 exit interviews for challenges and facilitators to operational, clinical, and financial components of integration. All coding was done electronically using Atlas.ti Version 7. Although data was originally to be presented according to integration model, there was not variation associated with the specific clinic model adopted. Qualitative data collection and analysis was overseen by the [BLINDED] Institutional Review Board. An IRB-approved information sheet was presented to each participant prior to the interviews.

Results

Table 2 summarizes analytic codes and findings related to clinical, operational, and financial aspects of integrating PC services in BH and addictions treatment settings.

Table 2.

Barriers and Facilitators to Integration for Clinical, Operational, and Financial Components of BHHs

Component	Sub-Code	Challenges	Facilitators
Clinical	Staff Background	Staff did not have education or experience in both physical and behavioral health Staff were frequently recent graduates that were ill-prepared to work with complex SMI/SUD populations	Have flexibility with provider credentials Staff must be motivated and ready to work with SMI/SUD populations Request for educational programs to include physical and behavioral health components Staff to “bridge both sides” Staff motivated to work with SMI/SUD populations
	Staff Training	Staff did not understand integration or view it as part of their job Shortage of time for training	Training through the context of care, (i.e., “osmosis”) Staff retreats for team building Touring each other’s’ spaces (i.e., “Open houses”) Cross-training and shadowing
Operational	Access and Scheduling	Difficult to get a functional panel size Not enough patients needed/wanted integrated services Fast-paced primary care scheduling did not fit the needs of SMI/SUD populations	Flexible scheduling (e.g., walk-ins, late appointments) Back-to-back primary care and behavioral health appointment scheduling Staff assigned to identify patient barriers to care
	Staff Recruitment and Roles	Lack of behavioral health and primary care clinicians locally and nationally High burnout and turnover Understaffed agencies were difficult to operate Staff were not prepared to work with SMI/SUD populations Unclear role definitions Staff did not execute roles for unknown reasons	Personal growth activities to encourage staff to stay Cross-training for primary care and behavioral health clinicians to increase empathy Requested unfilled roles: IT/data management, nurse care managers, peer support, community health workers, psychiatrist, and integration “champions”
	Communication	Different communication styles between primary care and behavioral health clinicians Shared meetings were expensive with the highest paid staff Primary care and behavioral health electronic health records were not integrated Electronic health record workarounds required staff time Concerns about access to records and privacy of sensitive information (e.g., SUD)	Shared primary care and behavioral health clinician meetings to discuss integration projects and shared cases Primary care and behavioral health clinician huddles to discuss yesterday’s problems and today’s patients Co-location made impromptu communication easier Staff training to ensure that staff understood how and what to communicate Second layer or protection for sensitive information Integrated electronic health records or identified staff member with access to all records
	Metrics	Technology limitations on the ability to record, track, and manage data No integrated electronic health records; pulling data from multiple sources was cumbersome	Data was utilized to improve care Metrics meaningful for SMI/SUD populations User-friendly technology More resources and staff time for data management More data from payers
Financial		Behavioral health home reimbursement rates were equivalent to primary care service rates despite extra services and staff time needed for SMI/SUD individuals Inequity between primary care and behavioral health clinician funding affected the ability to retain staff Concerns about sustainability affected staff motivation to work on integrated care	Financial support increased motivation for integration work A global payment system that accounts for smaller, high-need panels Incentivized metrics meaningful to SMI/SUD populations More data from payers, including emergency department and hospitalization utilization

Clinical components

Staff background

Participants frequently described how PC clinicians were not well-trained or experienced in provision of primary care to patients with SMI or SUD, including how to manage psychiatric medications. Concurrently, BH clinicians were described as not receiving basic training in screening for physical health conditions. Participants expressed concern that new clinicians on both sides were often recent graduates, ill-equipped to work with high-need SMI/SUD populations. One BH administrator explained, “[BH clinicians] went to school to be a therapist…it’s really kind of disenchanting…when you find out that what you are going to be doing is [coordinating] housing, benefits, skills training, insurance” (Site E).

Facilitators for recruiting well-prepared providers included flexible hiring. Nurse practitioners and physician assistants with a BH background were often hired in place of physicians (MDs/DOs) and described as effective. Some agencies recommended incorporating a comprehensive care approach into local graduate training programs. One agency’s PC clinician (Site G) gave a presentation on detoxification to physician assistants at a local university. Another agency (Site I) suggested having a staff member that could “bridge both sides” to educate new staff members. Half of the agencies emphasized that staff need to be motivated to work with SMI/SUD populations: “Unless this provider is really passionate for this type of population…that right provider who…has that passion for this population, you are…bound to have the turnover that we’ve had” (Site E).

Staff training

Four agencies (Sites A, C, G, I) reported that staff members struggled to recognize integration as part their own work: “It’s a sort of a slow education for our…200-plus employees to get them to understand what we’re trying to do here. Because I think mostly they just see it as, ‘Oh good, that’s a resource I refer that client to” (Site A). Another challenge was finding time for trainings.

Staff trainings were beneficial to integration efforts and varied across agencies. Four agencies (Sites A, C, F, G) described PC and BH clinicians learning about each other’s disciplines in the context of providing care or “by osmosis” (Site A). Huddles, shared meetings, and quality improvement projects were all helpful. Agencies also participated in more structured trainings, including a half-day staff retreat for team building (Site F). One agency (Site F) had 6-8 sessions with medical professionals presenting information to BH clinician on chronic health issues. A few agencies (Sites A, F) toured each other’s clinical spaces. One agency explained how these “open houses” helped their communication:

It’s been very beneficial for [the PC staff] to go to the [SUD agency] and see just how chaotic and crazy it is. You have someone coming in, someone storming out…if you called and you heard all this in the background, you’d be like, ‘What’s going on there?’…I think that makes a big difference because it lowers the communication barriers between different types of professionals (Site A).

Operational components

Access and scheduling

Four agencies struggled to obtain the correct panel size to fill both PC and BH clinicians’ schedules. One agency (Site D) had too many patients needing PC services, while another (Site E) reported not having enough. Two other agencies (Sites I, C) did not have enough patients receiving integrated care, citing that patients did not need/want services or the agency leadership failed to promote integrated services to patients. Second, half the agencies described that the “fast-paced” and “problem-oriented” scheduling of typical PC did not work for SMI/SUD populations. Agencies described SMI/SUD populations needing longer appointment times, walk-in accommodations, and easy rescheduling of missed appointments.

Providing easy access and scheduling required flexibility and effective communication between PC and BH clinicians. One agency explained, “There are times where we need to make exceptions…if you have someone who is agoraphobic and you’ve been working for six months to get them to see their doctor, if they show up…five minutes after the cut off time you don’t turn them away” (Site B). One agency (Site H) worked to accommodate walk-ins and assigned staff to call patients who “no-showed” to identify barriers to care. Three agencies (Sites A, G, H) emphasized having back-to-back BH and PC appointments. This was especially beneficial for patients in crisis or active drug use.

Staff recruitment and roles

All agencies expressed challenges recruiting and retaining staff. Several agencies cited a lack of PC and BH clinicians as a local and national problem. Some agencies could not fill critical roles, including clinicians who could prescribe antipsychotic medications. High turnover was common; one agency (Site E) reported that 73% of their staff stayed less than two years. Agencies described staff as ill-equipped and un-motivated to work with SMI/SUD populations (See Clinical Components). One agency in an underserved rural area (Site I) disclosed that new graduates could work a few years at their agency to have their student loans forgiven. High burnout rates made the agencies understaffed and difficult to operate. Furthermore, once positions were filled, staff struggled to understand their roles in integration: “Our [BH clinicians]…want to know ‘what’s my role in this integrated care concept?’…It wasn’t defined…it didn’t have roles and responsibilities attached to it” (Site G). A few agencies reported that staff members did not execute their responsibilities, “pawning it off” to others or not completing it entirely, which affected the agency’s ability to sustain its integration efforts.

For retention, one agency (Site E) emphasized professional growth opportunities to encourage staff to stay. Another (Site A) had PC and BH clinicians learn about each other’s daily routines to increase empathy and decrease burnout. All agencies expressed needing more staff overall, and half reported needing information technology staff to oversee information exchange and manage quality metrics. Three agencies (Sites A, C, J) had unfilled nurse care managers to track population health positions. Agencies recognized the importance of having an integration “champion” within agency leadership. One staff member described, “[The] challenge is we have a lot of support for integrated care…but we don’t have a lot of champions. And we need more champions to shift the culture across the agency” (Site B). Other roles that often went unfilled due to local shortages included psychiatrists (Sites A, I), community health workers (Site A), peer specialists (Site E), and individuals to perform triage and ER/hospitalization follow-up calls (Site E). Agencies stated that staff at all levels needed to be committed to integration.

PC and BH clinician communication

Agencies described a range of communication challenges. One BH agency (Site A) struggled to communicate with its SUD treatment partner about patients on methadone treatment who appeared overmedicated and consequently could not participate in therapy. This agency also cited communication style differences between PC and BH clinicians, such as PC clinicians being “curter.” Another agency (Site D) explained that interdisciplinary staff meetings were very costly as the highest paid staff—PC physicians and psychiatrists—preferred to be present together.

Other challenges arose around electronic communication via electronic health records (EHRs). While EHRs were frequently shared across agencies, no agency had an integrated EHR, and some BH records were still on paper (Sites D, G). Agencies created workarounds, including manual entry, shared care plans saved securely outside the EHR, and identified staff members to access multiple EHR systems. However, agencies criticized workarounds as requiring extra time and creating privacy problems.

Strong communication “diminished the silos” (Site H) between PC and BH clinicians, but this required planning. One agency (Site G) built more communication time into schedules and determined what and how information needed to be exchanged. PC and BH clinician communication was valued and allowed staff to determine which patients needed integrated services, develop individualized treatment plans, and coordinate care. All but two agencies discussed the benefits of shared PC and BH clinician meetings. Meetings were used to discuss integration progress, inform staff of integration projects, discuss shared cases, and cross train. Three agencies (Sites C, D, G) employed morning PC and BH clinician huddles to review problems from the previous day and discuss patients receiving integrated services that day. Co-located agencies used impromptu forms of communication (e.g., phone calls, office visits). Specifically, three agencies (Site C, D, G) identified impromptu communication helpful about medications, including refills and concerns about side effects and non-compliance.

For electronic communication, a few agencies had secure messaging but emphasized it was clumsy and not equivalent to having an integrated EHR. Agencies expressed that ideally they would have either an integrated EHR systems or identified staff members with access to all records. Agencies recommended that all staff receive training on their electronic communication protocols, so everyone knows how to send and receive information. Additionally, agencies suggested a second layer of protection for sensitive information, such as substance use.

Metrics: tracking patient data

All agencies were limited in their ability to record, track, and manage patient data. Since none had an integrated EHR, pulling data from two EHRs was cumbersome and required a lot of staff communication and time. One agency (Site A) described using six different tracking systems for the PC agency, BH agency, SUD agency, and payers. Some agencies struggled to pull reports from their EHRs and resorted to separate outside tracking systems (e.g., spreadsheets).

Some agencies had the ability to track data in their EHR, and a few were able to use data to improve care: “There’s a lot of people who just don’t come in for blood draws or they don’t show up…One of the first things [we] did was run a huge thing about these people who have this gap in care. We’re going to do outreach” (Site F). Regardless, more user-friendly technology was requested that incorporated BH and PC data. Half the agencies wanted more staff time and financial resources for data management. One PC agency explained that meaningful quality metrics need to be selected for the SMI/SUD populations:

The carrot drives the system…there are whole systems that are being created to capture that data and a great deal of staff time because we know that there’s going to be reimbursements…But if there’s no financial incentive, there’s no reason for anybody to go after it (Site A).

Lastly, a couple of agencies (Sites D, J) requested more data from payers, such as high-utilizers of the emergency department.

Financial

Most participants expressed aversion to the current funding models for reimbursement of PC services provided in BH settings. They described how SMI/SUD populations require expansive services, and consequently BH agencies had fewer patients per day compared to PC practices. However, their reimbursement rates were the same as usual care. One agency (Site E) attributed problems with BH clinician recruitment and retention to the inequity in salary between a PC and BH clinician, and their respective support staff. A few agencies reported discontent with fee for service because it did not encourage preventative services and care coordination. Ongoing concerns about payment models affected confidence in the BHH model: “The rumor was that…the program wasn’t making enough money…that was what the reasoning was behind not allowing the providers to be a hundred percent dedicated to the program…to be able to provide the wraparound level of care that these higher-level need patients need (Site I).” Participants advocated for a global payment system that accounts for SMI/SUD populations needing more services and staff time. They also requested financial support for integration efforts and metrics. One agency described financial components as motivating for integration work: “Funding was really positive because that pushed the clinic. Even though it wasn’t a ton of money, it was enough money to say, okay, let’s give this a shot” (Site I).

Discussion

Kathol et al.¹⁵ have identified seven components necessary to achieve sustainable, value-added integration in the PCMH: 1) combined medical and behavioral payment pool, 2) targeting complex patients, 3) use of onsite behavioral ‘teams’, 4) matching clinical expertise to escalating treatment needs, 5) defined, measurable, and systematic outcome, 6) use of evidence-based treatments, and 7) use of cross-disciplinary care managers in assisting the most complicated and vulnerable. Although BHH LC sites used a variety of integration models, there was a remarkable consensus on the challenges and facilitators for effective implementation, and their stories document barriers in five out of these seven components. The remaining two components – systematic outcomes and evidence-based treatments—were outside the scope of the focus group and exit interview discussions.

As identified in past work describing the Colorado-based Advancing Care Together¹⁶ and national Integration Workforce studies,¹⁷ integration of BH and PC services in real-world settings challenges the organizational culture, workflow, communication, and financial stability of transforming practices. Moreover, SAMSHA’s PBHCI program and similar studies identified staffing problems, data collection and management, consumer recruitment, and adapting workflows as challenges to integration PC services in BH settings.^3,10,12 BHH LC participants described all of these challenges. Additionally, BHH LC participants specifically identified staff backgrounds, panel size, scheduling, and the lack of integrated EHRs as challenges, which were not emphasized in previous trials. Upgrading EHR functionality to effectively support integrated care requires financial support as well as cooperation among clinicians, vendors, and regulators among others.¹⁸ In the previous literature, facilitators of integration were identified less frequently than barriers. However, Gerolamo et al. 2014 found that successful implementation of PC services into BH settings required defined staff roles, time for staff training and communication, consumer input, and a process to track integration efforts and consumer outcomes.¹² BHH LC participants endorsed all these facilitators except obtaining consumer input, which was only reported by one participant (Site G).

Initial BHH integration findings fit within an implementation science framework, and may provide an innovation model for future BHH implementation efforts. The National Implementation Research Network (NIRN) identified four stages of change: exploration, installation, initial implementation, and full implementation.¹⁸ BHH LC sites were all in the exploration stage, or in the initial phases of installation and implementation. Challenges, resources, and policies were still being identified and developed to facilitate integration. Findings from the BHH LC qualitative interviews directly align with each key NIRN implementation driver, including competency, organization, and leadership.¹⁹ Specifically, competency drivers reflected in the BHH LC interviews included challenges related to staff selection, such as recruiting and retaining staff motivated to serve populations with SMI and SUD, and the importance of ongoing staff and administrative training and coaching to assist integration efforts. Across organizations, work with the practice coaches was described as helpful for formulating integration goals, and for fostering communication across disciplines. Organization drivers noted as systems-levels barriers to BHH integration included service access, workflows, scheduling challenges, concerns about financial stability and payment models, communication issues, and a lack of integrated electronic health records. Finally, BHH interviews revealed the importance of strong executive leadership to sustain integration efforts and ensure successful implementation across the organization.

This study has several limitations. Data were initially collected to evaluate the impact of the BHH LC on advancing each site’s selected integration model. Agencies were not asked specifically about clinical, operational, and financial components, and specific challenges and facilitators for each site were not pursued extensively. Although a diverse set of agencies were represented, results may not be generalized to all BH settings. The BHH LC focused on the process of building an integrated care team from the perspective of clinical and administrative staff rather than clinical or patient-centered outcomes of patients with SMI or SUD. Future research should investigate how to prevent staff burnout, identify BHH components valued by patients with SMI/SUD, and explore of alternative payment models.

Implications for Behavioral Health and Public Health.

The need for culture change in both workforce development and clinic workflows to promote integrated care was described across Behavioral Health Home Learning Collaborative sites and was consistent with prior work.^16,17,20 BHH LC participants emphasized the importance of defined roles, shared team meetings, and interdisciplinary staff trainings. Building an effective integrated team is challenging and requires communication across disciplines. For the most complex patient cases in BH settings, collaborative care includes real-time discussion of patient symptoms and the development of a joint care plan.²¹ As described by BHH LC participants, building in adequate resources for adults with SMI or SUD requires time, financial resources, and shifts in professional culture that require substantial commitment from both clinical leadership and payers.