Identifying and Predicting Patients with Preventable High Utilization
Full Title: Identifying and Predicting Patients with Preventable High Utilization
PI: Rainu Kaushal, M.D.
PI Institution: Weill Cornell Medical College
OneFlorida Site PI: Betsy Shenkman, Ph.D.
Co-PI Institution: University of Florida
Study Staff: Katherine Blackburn
Funding Agency: PCORnet
Abstract: Five percent of patients, or “high utilizers,” account for 50% of health care utilization.
Understanding the needs of these individuals and designing appropriate interventions is
fundamental to improving the health of these patients as well as the efficiency, effectiveness,
and quality of the U.S. health care system. As health systems increasingly become responsible
for entire populations, ineffective management of these patients adds unsustainable burden to
the healthcare delivery system, jeopardizing the quality of health and healthcare with resultant
poor clinical outcomes for individual patients. High utilization often indicates that patients
with complex or specific needs are failing to receive the appropriate care in the appropriate
setting at the appropriate time. Many algorithms have been designed to predict risk of future
utilization, with mixed practical utility for healthcare delivery systems. Further, there are gaps
in the evidence for patients with high utilization, which contribute to variations in practice
patterns and clinical uncertainty.
This project will 1) solicit feedback from patients, clinicians, and health system leaders about
which group of patients with preventable high utilization should be prioritized and which
evaluative criteria should be emphasized in the resulting algorithm; 2) develop data sets and
computable phenotypes to identify and characterize patients with prioritized preventable high
utilization; 3) develop and compare the effectiveness of different methodological approaches
and data sources to predict a patient’s annual risk of falling into one of the prioritized groups
of patients with preventable high utilization; and 4) disseminate these results back to health
system stakeholders. OneFlorida is partnering with the Chicago Area Patient-Centered Outcomes
Research Network (CAPriCORN) and the New York City Clinical Data Research Network
(NYC-CDRN), the lead site, on this project.
OneFlorida Partner Sites
University of Florida
Page Last Updated: 11/02/2017