Separating the dimensions of network performance for coordinated care networks

Background

Coordinated care networks (CCNs), especially those that integrate health and human services, have recently emerged as a new paradigm to connect people with the services they need. As public, private, and nonprofit funders have continued to finance these networks, the need for evaluation has grown. CCNs primarily self-evaluate in ways that prioritize specific funder or management objectives, but a common evaluation framework can enable comparison of CCN performance and facilitate identification of evidence-based best practices. 

Objective

 1) Provide a set of metrics by which to commonly evaluate CCN performance. 2) Determine whether distinct patterns of performance exist. 3) Develop recommendations for managers of CCNs based on their performance profile. 

Methods

Drawing on data from a community referral technology, we aggregate data from 11 AmericaServes networks across 2015-2023, representing 28,697 military-connected clients and 71,991 requests. We then cluster the network aggregations based on 1) the time in days it takes them to match a client to a provider, 2) the percent of requests rejected by providers, and 3) the percent of requests resolved by providers. 

Results

Cluster analysis reveals four distinct profiles of network performance. We term these profiles: 1) well-rounded, 2) low-efficiency, 3) low-accuracy, and 4) low-resolution. These four profiles capture distinct performance tendencies in networks, recognizing both strengths and challenges. 

Conclusion

CCNs exhibit distinct performance profiles consistent across geographies, at least for the military-connected population. A common evaluation framework makes it possible to identify these profiles and suggests pathways to more targeted management interventions. Future research should seek out such interventions and dig further into the performance of CCNs, focusing on how performance varies across services.

Presentation slides