Columbia University Learning Health System Launches Pilot Projects
The Columbia University Learning Health System (LHS) initiative is supporting its first two pilot projects. The first will use machine learning to more quickly identify sepsis, and the second will create a pediatric equity and quality dashboard.
“This initiative combines the expertise of world-class physicians and researchers from NewYork-Presbyterian [NYP] and Columbia to harness technology to improve the quality of care for all,” explained Laureen Hill, M.D., M.B.A., group senior vice president and chief operating officer at NYP/Columbia University Irving Medical Center (CUIMC), in a statement. “We are delighted to work together to develop innovative tools to advance care and health equity for our patients.”
The LHS initiative builds on years of expertise generated in clinical care, informatics, health systems research, policy, industrial engineering and operations, digital health, behavioral science and, importantly, translational science excellence at CUIMC to solve problems for clinical practices and providers by feeding researchers real-time data from those same sources.
“At ColumbiaDoctors, we accelerate the path to bring innovation to the point of care,” says Timothy J. Crimmins, M.D., chief medical information officer for CUIMC, in a statement. “Connecting our world-class physician-scientists to world-class patient care through a dynamic learning health system drives the transformation needed to provide the best experience and outcomes for our patients, staff, and providers.”
The chosen projects will leverage the breadth of expertise and experience within the unique environment at NewYork-Presbyterian/CUIMC and the full weight of the institutions’ conscience and commitment to close disparities in health care.
Using Machine Learning Algorithms to Quickly Identify Sepsis
Sepsis, an illness caused by the body’s dysfunctional response to infection, is a leading cause of death of hospitalized patients. Once the condition sets in, every hour that sepsis treatment is delayed is associated with increased mortality.
At NewYork-Presbyterian and hospitals around the world, sepsis is identified using a set of criteria that relies on abnormal vital signs and lab tests. However, machine learning algorithms drawing data from the electronic health record may more quickly detect sepsis.
The research team will use a methodology that employs learning with observational data, a recently described approach that hasn’t been used for sepsis. The hope is that their work will lead to improved care for patients who contract sepsis at NYP and, potentially, a larger federal study.
Identifying and Addressing Inequity in Pediatric Emergency Care
Equity dashboards dynamically track real-time quality metrics stratified by sociodemographic characteristics to identify inequities and have the potential to inform and ultimately enhance rapid-cycle quality improvement initiatives focused on reducing healthcare inequities.
A CUIMC team of pediatricians will study and pilot the development of a Pediatric Equity and Quality Dashboard & Implementation Roadmap in the Morgan Stanley Children’s Hospital Emergency Department. The NYP Dalio Center for Health Justice will host and maintain the dashboard.
The pilot Dashboard will display inequities in a single, patient-centered metric – emergency department length of stay – with documented disparities in the literature. The team expects the dashboard and roadmap will reduce inequities in pediatric ED length of stay, identify successful strategies that could be tested at other sites, and create a platform to recognize and assess meaningful pediatric equity measures and local infrastructure to support pediatric population health initiatives.
Columbia’s Learning Health System Initiative
Columbia University’s LHS initiative is a collaboration between ColumbiaDoctors, NYP, CUIMC, Columbia Engineering, New York State Psychiatric Institute, and the Irving Institute for Clinical and Translational Research, as well as other collaborators across the Columbia campuses.
Last fall, these stakeholders gathered for the first LHS symposium. The pilot awards were announced at this symposium and ideas or teams generated from that session were encouraged to apply. The pilot award research teams are being paired with Columbia Engineering experts, who will advise the teams on dynamic optimization, developing networks, risks analysis, and computational support.
The LHS initiative will eventually form a network of faculty working groups to help address implementation barriers in the learning health system cycle. The overarching goal is to capitalize on Columbia University’s world-renowned scholars to integrate informatics, implementation science, and other methodologies at the point of care to improve quality and accelerate the translation of scientific discovery into practice.
The LHS pilots are the first in a set of Strategic Priorities Pilot Awards created by the Irving Institute. Home to CU’s Clinical and Translational Science Award (CTSA) Program hub, the Irving Institute is one of about 60 medical research institutions across the nation that work together to speed the translation of research discovery into improved patient care.