Socio-technical infrastructure for a learning health system

This commentary is in many ways a follow-on to, and elaboration of, the commentary published in the July issue of this journal.¹ The previous commentary introduced three characteristics that contribute to the uniqueness of learning health systems (LHSs) as an approach to health improvement. The three characteristics introduced there were: “(1) a multi-stakeholder learning community that is focused on the (targeted) problem and collaboratively executes the entire cycle; (2) embracing, at the outset, the uncertainty of how to improve against the problem by undertaking a rigorous discovery process before any implementation takes place; and (3) supporting multiple co-occurring cycles with a socio-technical infrastructure to create a learning system.”

This commentary focuses on the very important third characteristic, infrastructure. It examines the role of infrastructure in the overall architecture of an LHS and describes LHS infrastructure in terms of 10 interconnected socio-technical services accompanied by a brief description of each. Like the previous commentary, this one seeks to bring an increased level of focus to discussions of LHSs and move an emerging field, what is coming to be called “Learning Health System Science”,² toward a sharper conception of its core principles.

Critically, LHS infrastructure must extend beyond digital technology in order to support improvement of individual and population health. The infrastructure must be socio-technical in the sense that it incorporates the roles that a wide range of people must play at different levels of social organization: as individuals, as teams, as members of organizations, and as citizens of civil society.⁵ Technology, alone, only establishes a potential for health improvement through an LHS.

Viewing its infrastructure in terms of socio-technical services could be beneficial in several ways beyond working toward a consensus view of LHS structure and function. Most notably, such a modular approach could lead to sharing of interoperable infrastructure components and the possibility that sharing of such components might promote the more rapid adoption of LHS methods. Moreover, compatibility of LHS architectures could enable smaller scale LHSs to compose into a single system that functions at larger scale. Logical next steps to mature LHS infrastructure would include building consensus around the constituent services and developing specifications for each one.

The authors have no conflicts of interest to declare.