Using Group Model Building to Develop Insight into a Local Behavioral Health Care System

Purpose:A persistent challenge in supporting people with mental illness and addiction is the breadth of services, advocacy, and public support needed for recovery. Differences in philosophies, funding streams, institutional priorities, fields, and needs often make it difficult to develop a common language, let alone coordinate activities. The resulting fragmentation of services results in systemic gaps that can have tragic consequences, as vulnerable persons seek help and supports and may instead experience stigma, loss of hope, and even injury and death.

The Brown School Social System Design Lab (SSDL) at Washington University worked with the Behavioral Health Network based in St. Louis to identify systemic gaps that may be leading to increasing numbers of individuals seeking supports and services in emergency departments (EDs). This project describes the key system insights and lessons learned from this process.

Methods:Approximately 20 advocates, clergy, providers, law enforcement, hospital administrators, housing support providers, and consumers (as peer recovery specialists) participated in five 3-hour group model building (GMB) workshops. GMB is a form of community-based participatory research based on principles of system dynamics (SD). SD uses causal maps and computer simulation models to understand relationships between elements in systems and how they change.

Results: Ultimately, several causal maps were generated that led to key system insights. We will present a revised map that incorporates three of the most important insights. First, frequent ED visits are both a cause and an effect of unmet need, with mutually-reinforcing frustration weakening relationships between patients and providers: “The more you go, the less help you get.” Second, interventions must reach beyond the ED into the community. Community supports such as peer recovery specialists, friends and family, and clergy can play an important role at multiple leverage points to address unmet need. Third, stigma and the lack of a recovery orientation in service providers likely exacerbates gaps in care by disempowering consumers. Thus, it may not be new services that are needed, but rather a change in what it means to be a consumer or provider in the behavioral health care system. Exemplary quotes will be shown to make clear how the session discussions lead to the causal maps. 

There are also important “lessons learned” from this project for those interested in using methods like GMB to address complex problems. Key among these is careful attention to who is invited and present and who is not, developing a shared language early on, and balancing training goals with the development of recommendations.

Implications: Causal mapping with stakeholders through methods like GMB leads to system insights by capitalizing on the rich knowledge of those who are daily experiencing the system of interest. For policy-makers and researchers, formal simulation models may follow causal maps to test policies for unintended consequences. For social work researchers, who understand the dynamicism and complexity inherent in the lives of people living with severe mental illness and addiction, causal maps can assist in improved theory-building and suggest future directions for research.