Development of a Maturity Model for Learning Health Systems: A Framework for Self-Assessment and Continuous Improvement

Abstract

The concept of a Learning Health System (LHS) was first envisioned as a means to enable value-based health care through the digital transformation of health systems. An LHS aligns people, processes, and technology to support rapid learning cycles of knowledge discovery and implementation. In an LHS, data management and analytics capabilities enable the discovery of scientific evidence from data that are routinely collected from care delivery practices. Concurrently, scientific evidence is integrated into the point of care for practice change and improved patient outcomes. Despite published case studies describing how LHS learning cycles have been enacted in practice, there is limited guidance on how health systems should plan for, measure progress toward, and continuously improve their learning cycle capabilities. To bridge this gap, this dissertation describes the co-development of a maturity model intended to support learning cycle capability development. Maturity models are used to describe and measure components of a system through a series of phases, stages, or levels. To develop the maturity model and component parts, I engaged in a design-based, action research effort with a clinical department specializing in Physical Medicine and Rehabilitation (PM&R) within an academic health system. To construct the component parts of the maturity model, I incorporated multiple data sources through successive design iterations. Data sources for model design included inputs from clinical and operational stakeholders within PM&R, feedback from an expert panel of LHS researchers from outside PM&R, and a comprehensive literature review. Three design iterations were used to refine the component parts of the maturity model in an integrated manner. Findings from data analyses in each design iteration were used to enhance the component parts progressively until final versions were reached. The resulting maturity model is comprised of two parts: a process reference model for learning cycles and a self-assessment instrument for learning cycle capability measurement. The process reference model describes activities, outputs, and stakeholder roles required for the execution of learning cycles by clinical teams within health systems. The process reference model utilizes generalizable terminology and provides the basis for capability assessments. It also serves as a standardized reference to streamline project management for capability development across teams. The self-assessment instrument enables measurements of team maturity as a function of its learning cycle capabilities. The co-developed maturity model is intended to serve as a roadmap for clinical departments and health systems. The maturity model enables objective assessments of the current state of learning cycle capability across individual teams within a health system. Such measures are intended to facilitate continuous improvements and shorten the lead time for capability development across the health system. As envisioned by the LHS principles, improvements to patient outcomes are sought as a direct consequence of the use of the maturity model in practice. The design-based methods used in this dissertation provide an additional contribution to the LHS literature by serving as an example for how to collaboratively create a measurement instrument to support LHS capability development in practice.