Brokering System Change: A Logic Model of an Intermediary-Purveyor Organization for Behavioral Health Care

Studies examining IPOs have suggested that they support EBI implementation (5), facilitate knowledge translation activities (6), and build organizational capacity to sustain EBIs (7). Franks and Bory (2, 3) identify service systems’ experiences with innovation initiatives, system needs, capacity and leadership, and stakeholder collaboration as IPO facilitators. They name seven IPO functional domains: consultation; best practice model development; purveyor of evidence-based practices; quality assurance-improvement; outcome evaluation; training, public awareness, and education; and policy and systems development. Examining child mental health IPO strategies, Proctor et al. (4) found that IPOs rely most heavily on education and training, followed by planning (e.g., addressing system barriers) and quality management (e.g., collecting and summarizing performance data). Although these preliminary studies begin to define IPO activities and contextual facilitators, further conceptualization of an IPO’s relationship to a system and its stakeholders is needed, particularly with regard to partnership-driven processes at the core of IPO work.

To advance understanding of an IPO’s role in system improvement, we summarize IPO inputs, objectives, and impacts in a logic model. (Details of this model are available in an online supplement to this Open Forum.) It synthesizes insights from the dissemination and implementation (D/I) literature (e.g., knowledge-brokering and facilitation) and incorporates practical lessons from more than 3 years of developing a university-based IPO serving a large public mental health system at the University of California, Los Angeles (UCLA). The Los Angeles County Department of Mental Health (DMH)+UCLA Public Mental Health Partnership (PMHP), established in 2018, focuses evidence-based training and technical assistance on two DMH programs that serve individuals with serious mental illness who experience or are at risk for homelessness: the Homeless Outreach and Mobile Engagement (HOME) program and full-service partnerships (FSPs). As a team-based intervention modeled on assertive community treatment, FSPs are field-capable multidisciplinary teams that provide intensive services to individuals with serious mental illness and other vulnerabilities (e.g., justice involvement). HOME’s street-based teams conduct outreach to difficult-to-engage unsheltered individuals needing behavioral health treatment. PMHP began with a vision of improving EBI adoption but evolved to incorporate additional systems-oriented objectives, such as leadership and quality assurance capacity development.

As depicted in our logic model (see the online supplement), IPO inputs include three main partnership groups: academic and technical experts, system partners, and community stakeholders. Collaboration among academics, practitioners, policy makers, and intermediaries has been shown to be critical for use of evidence in practice settings (8). Our model suggests, for example, that academic experts may share knowledge about D/I methods; practitioners offer information about service system needs and capacity; and community stakeholders, such as knowledgeable advocates, may highlight marginalized perspectives whose centering could increase readiness for change. This notion is congruent with cultural exchange theory, which posits that innovation results from bidirectional exchanges of knowledge, attitudes, and behavior across cultural groups (e.g., practice and academic) through debate, negotiation, and compromise (9). Further, participatory research approaches emphasize system stakeholder perspectives in formulating implementation goals and strategies, thereby supporting “the deconstruction of power and democratization of knowledge” (10). These frameworks underlie IPO inputs, supporting the idea that without a clear understanding of community resources and perspectives that shape uptake of innovations, sustainable system improvement will remain difficult to achieve.

Consistent with IPO studies (4–6), PMHP activities include consultation and technical assistance, such as sharing EBI knowledge and addressing system change barriers; best practice model development based on D/I and evaluation methods; training and education, including developing and distributing educational materials and convening learning communities; quality monitoring strategies, such as helping providers collect and use performance data; and planning and restructuring activities, such as prioritizing implementation approaches. These activities advance IPO objectives, including supporting knowledge and skill building at the levels of communities, organizations, and individual providers; negotiating evidence-informed strategies to fit service system culture and context; specifying problems and potential solutions, including a willingness to raise tough questions; and building long-term capacity to sustain policy, practice, and system changes.

The IPO mission, here stated as “realizing a vision for excellence in care for vulnerable populations that incorporates typically marginalized perspectives and builds capacity to foster health equity” is served by the objectives identified. To achieve excellence, an IPO centers previously marginalized perspectives, for example, by including historically underrepresented stakeholders’ perspectives that are often sidelined in the service of maintaining the status quo. Our model suggests that the mission is achieved when partnering dynamics catalyze IPO objectives, in turn, affecting intrasystem capacity outcomes (e.g., provider knowledge and skills) (11) as well as intersystem capacity outcomes (e.g., number and intensity of service sector collaborations) (12). The model indicates that such capacity improvements enhance the health of individual service recipients and communities through improved quality of care.

Participatory processes were exemplified in PMHP’s engagement with stakeholders to establish a plan for quality improvement in FSP, a program that incorporates elements of EBIs but with unique team structures, service arrays, and processes. System leaders did not view any single EBI as responsive to problems they prioritized. They also identified the need for greater programmatic consistency (e.g., appropriate clients enrolled and multidisciplinary staffing) and conceptual cohesion built on practices used by a range of EBIs (e.g., recovery orientation and assertive engagement). In a decentralized program such as FSP (13), promulgating or mandating any single EBI could undermine system strengths, expose mismatches between resources and expectations, and disadvantage some clients. PMHP engaged in extended communication and bidirectional learning with frontline staff and program leaders to integrate stakeholder knowledge with evidence (e.g., literature reviews), ultimately moving toward a codified FSP best practice model, largely derived from EBIs but with tailored staffing and service offerings. For example, some agencies adopted staffing models that utilized staff specialized in substance use or housing on FSP teams, whereas others employed general case managers depending on agency resources and service philosophy; yet there was a consistent expectation to address client substance use and housing needs. As in participatory research, goals were refined through collaboration, with every aspect of system change (e.g., priority setting, implementation planning, and resource development) guided by diverse system and community experts.

The objectives and activities in our model reflect an understanding that IPOs must respond dynamically to system context and the interplay between intervention, organization (e.g., billing requirements), provider (e.g., knowledge and skills), and client characteristics by negotiating evidence-informed strategies to fit service system culture (14). For instance, DMH initially proposed implementing the individual placement and support (IPS) model of supported employment within the FSP program. Questions emerged about the fit of a high-fidelity IPS approach to the inner and outer contexts of the program; for example, some IPS services were incompatible with funding streams. Moreover, dismantling long-standing services (e.g., volunteering and peer-run enterprises) to achieve IPS fidelity conflicted with organizational values. Thus, in collaboration with service system stakeholders, PMHP leaders helped delineate and disseminate an IPS-informed approach consisting of a range of employment activities rather than focusing solely on competitive employment. Without such responsiveness to context, interventions delivered with high fidelity may have limited clinical relevance, acceptability, and sustainability (11).

Social and relational processes—including exchange of values, knowledge, and skills among IPO partners—drive successful selection and pursuit of activities to improve behavioral health systems. A community-partnered approach will confront the tension between EBI adaptation and fidelity, such as in a highly decentralized program like an FSP in which sites offer unique strengths, diverse practice patterns, and a range of skills among staff. IPOs will need to utilize social networks and leverage political will and policy levers to develop a shared understanding of salient system features, stakeholder-relevant best practices, promising EBIs, and feasible implementation strategies and to identify individuals who are critical to the advancement of IPO activities. Our model suggests that closing the research-practice gap may require a more intentional approach to stakeholder involvement in defining service issues worthy of attention as well as potential solutions.

The present model can guide empirical study of IPO effectiveness. First, investigations should examine contributions of partnership structures and processes to IPO outcomes, including uptake of EBIs. Structural equation models can test causal pathways between partnership variables and system capacity outcomes, as well as whether these relationships are mediated by IPO interventions. Second, evidence regarding the efficacy of IPO strategies is needed, along with tests of interactions between IPO strategies and inner and outer contextual factors (e.g., between training protocols and intervention characteristics). Such investigations can guide decisions about which IPO strategies are best used under what organizational, policy, and intervention-related conditions. Third, to delineate common metrics of IPO impact, D/I frameworks of impact assessment (15) may illuminate relevant measurement strategies (e.g., number of providers delivering EBIs or number of EBIs resulting in new clinical guidelines or health system policies).

Finally, our model helps clarify the role of IPOs. IPO activities differ from those that academic researchers might pursue in research-practice partnerships, in evaluating community programs, or in supporting implementation efforts toward EBI adoption. Nonetheless, IPOs engage in evidence-driven and knowledge-generating activities common to each of these efforts. For example, PMHP evaluates its impact on system readiness for change, and IPOs may conduct program evaluations that contribute to program development goals. Our model can guide case studies, process evaluations, and outcome assessments that richly describe dynamic engagements between researchers and service systems and the impact of IPO activities on system performance.

The authors thank Drs. Jonathan Sherin, Patricia Lester, Marta Alquijay, Lisa B. Dixon, Lawrence A. Palinkas, Sam Tsemberis, Sapana Patel, Paul Margolies, Helle Thorning, Alejandro Silva, and Kara Taguchi; Ms. Gita Cugley, Ms. Emi Bojan, Ms. Teresa Halliday, Ms. Anna Bruce, Ms. La Tina Jackson, and Ms. Rosalinda Cárdenas; and other DMH partners, the Public Mental Health Partnership team, the UCLA Prevention Center of Excellence, and colleagues in the Jane and Terry Semel Institute’s Center for Health Services and Society.