Skip to main content

LEARNING HEALTH SYSTEMS (LHS)

BIOBANKING SERVICES

WHAT IS A LEARNING HEALTH SYSTEM?

A Learning Health System (LHS) is a system in which care generates evidence, and evidence immediately reshapes care.

A Learning Health System is one in which science, informatics, incentives, and culture are aligned for continuous improvement and innovation — with best practices seamlessly embedded in the care process, patients and families as active participants in all elements, and new knowledge captured as an integral by-product of the care experience.

Adapted from the Institute of Medicine, Best Care at Lower Cost (2012), developed through the Roundtable on Value & Science-Driven Health Care. The National Academy of Medicine’s current wording adds equity, and broadens the frame from care to health delivery.

The idea was introduced by the Institute of Medicine (now the National Academy of Medicine) nearly two decades ago as a unifying commitment to speed the generation and use of evidence. The definition is deliberately broad. It does not prescribe a technology stack, an organizational chart, or a research method — it describes a capability: the ability of a health system to study itself and act on what it finds, continuously, at scale.

In practice, an LHS is a cycle rather than a project.

The Learning Cycle. Practice - Care delivered to real patients. Data - EHR, claims, patient-reported, community, Knowledge - Analysis, trials, models, evaluation, Change - Order sets, workflows, policy, decision support

The Four Foundational Elements

The NAM framing rests on four pillars. A system can be strong in one pillar and still fail to learn; the elements have to work together.

ElementWhat it means in practice
ScienceRigorous methods applied to real-world care — pragmatic trials, quasi-experimental designs, implementation and improvement science.
InformaticsData that can actually be found, trusted, linked, and returned to the bedside. Common data models, computable phenotypes, decision support.
IncentivesAlignment so that learning is rewarded rather than tolerated — in payment, in promotion criteria, in operational priorities.
CultureLeadership, trust, transparency about variation, and genuine partnership with patients, families, clinicians, and communities.

What an LHS is not

  • Not simply a data warehouse. Data infrastructure is necessary but not sufficient; without governance, method, and a route back into practice, a warehouse produces reports, not learning.
  • Not simply quality improvement. QI is a core engine of an LHS, but an LHS adds research-grade rigor, generalizable knowledge, and dissemination beyond the local site.
  • Not simply research on clinical data. Secondary analysis that never returns to change care completes only half the loop.
  • Not a one-time project. The defining feature is that the cycle repeats, and each turn is faster and cheaper than the last.