Towards a Learning Health System to Reduce Emergency Department Visits at a Population Level

Elliot Brannon
Tianshi Wang
Jeremy Lapedis
Paul Valenstein
Michael Klinkman
Ellen Bunting
Alice Stanulis
Karandeep Singh
Peer-Reviewed Article
December 2018

This resource describes an effort to apply a learning health systems approach to reducing ED utilization.

  • A learning health system focuses on improving people’s health at scale through continuous cycles of knowledge discovery and implementation of best practices.
  • This study developed a predictive model to harness data from three Michigan health systems and identify residents at highest risk for future ED utilization in order to more effectively target a care coordination intervention.
  • The intervention consisted of steps including entry of data into a care management electronic health record (EHR), a needs assessment, and care plan development.
  • The model was able to accurately predict the number of ED visits during the subsequent six months to within a quarter to a half of a visit.
  • Transforming care coordination from a reactive to a proactive activity has potential to reduce ED utilization and therefore reduce costs.
Posted to The Playbook on
Level of Evidence
What does this mean?