Implementing the Collaborative Care Model as Part of a Countywide Initiative

The Washtenaw Health Initiative (WHI) was formed in 2011 and is a voluntary, countywide coalition of more than 60 health services organizations, including the major health systems in Michigan’s Washtenaw County. Participating organizations include the University of Michigan Health System, the St. Joseph Mercy Health System, the Veterans Affairs Ann Arbor Healthcare System, Washtenaw Community Mental Health, numerous nonprofit human services organizations, and several county government agencies, including the Washtenaw County Public Health Department (1). The WHI, a developing accountable health community (AHC), is staffed by the Center for Healthcare Research and Transformation, and its goal is improving access to high-quality, coordinated care for low-income, uninsured, and Medicaid populations in the county.

The WHI convened a working group on mental and substance use disorders and a primary care working group to determine unmet need and gaps in services among low-income county residents and to develop and prioritize initiatives to address these gaps. In 2012, group members coalesced around the issue of inadequate access to mental health services, particularly psychiatric services, as a pressing concern for low-income individuals in the county.

Although the supply of mental health services was considered to be limited for all Washtenaw residents, working group members noted that patients with higher incomes had some level of access to private-sector mental health services and that individuals with serious and persistent mental health conditions had access to public mental health services. However, low-income individuals with depression, anxiety, or other mild to moderate mental health conditions often did not have access to either public- or private-sector services. Instead, these individuals usually relied on their primary care providers (PCPs) for mental health care.

Formulating a Response to Lack of Access

Given the working groups’ concerns about the lack of access to mental health services, academic faculty members in the groups shared information about integrated behavioral health care models and suggested that such models could partly address ongoing access and quality issues. Integrated models have increasingly been emphasized by policy makers, professional societies, and large managed care organizations to address the limitations of separate primary care and mental health services (2–4). Integration of services could assist overburdened PCPs, reduce the impact of psychiatric physician shortages, and serve patients who prefer to receive mental health care from their PCPs.

The academic group members emphasized that one model of integrated behavioral health care—the collaborative care model (CoCM)—had the most extensive research support for efficacy. This model is implemented by a team consisting of a primary care clinician, an embedded care manager (usually a master’s-level social worker), and a consulting psychiatrist and includes the following key elements: measurement-based care, in which patients’ symptoms and treatment response are routinely assessed with standardized measures; care manager–led coordination, with regular assessment of treatment adherence, support of self-management, and regular communication with PCPs; and “treatment to target” for patients who fail to improve as expected, with a consulting psychiatrist conducting a panel review with the care manager and making treatment recommendations to be conveyed to the PCP (5).

More than 80 randomized controlled trials (RCTs) had shown that patients from a variety of populations (including low-income, geriatric, medically complex, and racial-ethnic minority populations) who are enrolled in CoCM programs have depression treatment outcomes superior to patients in usual care (6). CoCM programs also improve outcomes for patients with anxiety disorders (6).

Both the primary care and behavioral health working groups recommended to the WHI steering committee that CoCM be pilot tested in safety-net clinics. The steering committee subsequently endorsed the proposed initiative as a WHI-sponsored effort.

Finding Funding

Unfortunately, CoCM endorsement by the WHI did not automatically come with funding, because the WHI served as a forum for planning and community collaboration and was not a funding agency. However, several members of the WHI steering board were fiduciaries of the major health systems in the county and sympathetic to initiatives arising from this countywide collaboration.

With the assistance of WHI staff, academic partners in the behavioral health working group took the lead in writing internal proposals to the WHI finance committee and grant proposals to external foundations and other agencies. After reviewing the internal proposal, two of the large health care systems participating in the WHI, the University of Michigan Healthcare System and the St. Joseph Mercy Health System, provided substantial funds for implementation of the CoCM program in county safety-net clinics. Some, but not all, of these contributions benefited their own operations because two of the four safety-net clinics were in the University of Michigan system and one safety-net clinic was in the St. Joseph Mercy system. The fourth was an unaffiliated community clinic.

The WHI steering committee strongly suggested that an evaluation of the new CoCM program accompany implementation, and proposals to external granting agencies resulted in further funds for implementation and evaluation of the project. These funds came from the Blue Cross Blue Shield of Michigan Foundation, the Michigan Department of Community Health Innovation Grant, and Medicaid match grants.

Project Implementation and Additional Partnership Building

Securing funding was accomplished over nine months, and hiring of project staff and implementation of the CoCM program began soon afterward. The WHI working groups included several leaders in the primary care clinics in which the CoCM program would be initiated. Participation by these leaders increased local ownership of the project, and project staff continued to work closely with them to tailor the CoCM program to the needs and resources of each clinic.

Collaboratively gathered data on usual care in the piloting clinics prior to CoCM implementation led to an appreciation of the unique populations served by each clinic, the mental health resources that were already in place, unique work flows, and other issues key to local implementation. For example, Clinic A, an unaffiliated community clinic that gained federally qualified health center status during implementation, had an embedded part-time social worker and limited psychiatrist time (four hours per week) dedicated to the care of the seriously mentally ill population. Clinic B, a university-affiliated clinic staffed primarily by medical residents, served many medically complex patients recently discharged from general medical inpatient stays and had extremely limited social work services and no psychiatric presence in the clinic. Clinic C, staffed primarily by medical residents in a community hospital, served large numbers of persons with substance use disorders as a result of referral patterns from substance use treatment facilities and had only a limited social work presence focused on concrete needs. Clinic D served a high-needs, low-income population of patients mainly from racial-ethnic minority groups and had a weekly four-hour embedded clinic staffed by a psychiatrist and residents and employed two full-time social workers focused on concrete social needs.

Project staff interfaced with all existing resources in each clinic, ensuring complementary services and smooth work flows while emphasizing that maintaining key CoCM elements was important to maintaining effectiveness. Numerous challenges were encountered when implementing this countywide project in four clinics that spanned three health systems. These challenges included, but were not limited to, completing several human participant applications and business associate agreements; managing three different hiring systems and procedures; interfacing with five different electronic medical record systems, with two clinics changing record systems; and frequently educating new primary care staff because of staff turnover in the clinics.

Evaluation

Even though RCTs have indicated that CoCM is effective, the WHI board strongly recommended including an evaluation component to address whether CoCM could be successfully implemented and improve outcomes locally. The evaluation of local efforts proved important in securing buy-in from community and clinic leaders, many of whom thought that current depression treatment practices were adequate, with little need for improvement. A stepped-wedge design was used in the evaluation: depression treatment outcomes of patients receiving usual care in the four safety-net clinics were first assessed and then compared with the outcomes of patients who enrolled in the CoCM program once it was implemented (7).

The evaluation found that local CoCM participation was associated with greater improvements in behavioral health outcomes. Over the first three months, 48% of CoCM patients reported a clinically significant reduction in depressive symptoms (≥5 points on the Patient Health Questionnaire–9), compared with 33% of patients receiving usual care (p=.01). In addition, 40% of CoCM patients reported a clinically significant reduction in anxiety symptoms (≥5 points on the Generalized Anxiety Disorder–7), compared with 26% of patients receiving usual care (p=.02).

Lessons Learned and Next Steps

AHCs that are similar in structure to the WHI may be an important avenue for bringing communities together to jointly identify local mental health needs, prioritize needs to be addressed, implement evidence-based models for improving care, fund these efforts (at least temporarily), and advocate for coverage and sustainability of treatment models that benefit the community, such as CoCM.

The problem identified by WHI partners regarding lack of access to high-quality mental health care is not unique to a single county or a single region. Many communities experience shortages of mental health providers. A comprehensive assessment indicated that 77% of U.S. counties had a severe shortage of mental health prescribers where half or less of mental health prescriber need was met (8), and 96% of counties had some level of unmet mental health need.

CoCM is a proven model for improving the outcomes of primary care patients with mental health needs. It leverages psychiatric physician expertise by using team care. However, despite substantial research support, dissemination has been difficult, with delays in translating CoCM into clinical practice and difficulties obtaining reimbursement for many indirect team activities (9).

Key to the success of the current implementation was the buy-in of community and clinic leaders and the ability to tap into funding from five different partners, including health systems, foundations, and state funds. As policy makers have noted, hospital community benefit programs that collaborate with AHCs, similar in structure to the WHI, are likely to increase the impact of their benefit funds on community health (10).

Because of the partnerships built by the WHI, efforts to roll out and fund CoCM are continuing beyond the original efforts in four safety-net clinics. A roll-out of collaborative care is now under way in a large provider organization in Washtenaw and nearby counties, Integrated Health Associates, with an additional six to eight primary care clinics expected to implement the model in the next two years. The University of Michigan Health System is developing a plan to implement CoCM programs throughout its primary care clinics. Discussions regarding longer-term reimbursement strategies are also ongoing; participants include local Medicaid health plans and the Michigan Behavioral Health and Developmental Disabilities Administration. Ongoing learning meetings, widespread sharing of expertise across community organizations and health systems, and open lines of communication with government officials and health plans have been critical to these follow-on efforts. The “backbone” structure of the WHI, a developing AHC, continues to be essential in this expanded implementation.

Conclusions

AHCs are increasingly being formed in response to the Affordable Care Act and the Centers for Medicare and Medicaid Services Innovation Center grants. These highly partnered efforts offer new opportunities for academic, community, and government leaders to meet, prioritize, plan, implement, and disseminate evidence-based interventions, such as CoCM, that address important gaps in mental health services.