Moving from a registry to a learning health system: A case study of a Dutch prostate cancer registry

Learning health systems (LHSs) are systems that seamlessly embed continuous quality improvement based on real-world data. To establish LHSs, several infrastructures need to be in place. Registries already have part(s) of this infrastructure and could therefore be leveraged to establish LHSs. This study aims to identify key factors facilitating the transition of registries into LHS to support continuous learning from real-world data.

Eleven interviews with 12 stakeholders, including medical specialists and nonmedical stakeholders, were conducted in the context of a prostate cancer registry. Findings were coded deductively based on seven previously identified facilitators for learning: complexity, relative advantage, compatibility, credibility, social impact, actionability, and resource match. These facilitators cover technical, social, and organizational aspects. An inductive phase followed to pinpoint factors for continuous learning and LHSs. Subsequently, two focus groups were conducted to ensure accurate interpretation of findings, and five expert panels to provide additional context.

Complexity within healthcare systems emerged as a significant challenge, attributed to multiple stakeholders and the rapidly changing healthcare landscape. The advantage of LHSs is the timely availability of population-based data for real-time care adjustments. Compatibility of the system with stakeholders' needs was considered pivotal requiring a relatively flexible infrastructure. Credibility of data and results was supported by creating transparent processes in which stakeholders could review data from their own patient population. Social influences, including interpersonal trust and engaged leadership, fostered collaboration within LHSs. Actionability of the findings and resource match were vital for knowledge translation and sustainability.

Our findings provide practical recommendations to support registries in transitioning towards LHSs by leveraging and expanding their infrastructure for continuous learning. We identified technical, interpersonal, and organizational factors that facilitate continuous and rapid learning using real-world data, create transparent and collaborative infrastructures, and help to navigate the complexity of the healthcare system.