Description
Abstract: Social risk factors are a key driver of overall health status and are responsible for unnecessary and/or preventable use of costly services. We developed a decision support tool using machine learning algorithms based on a combination of local health information exchange, and small-area aggregate socioeconomic and public health indicators to identify adult patients who may need a referral to various social and behavioral services. The intervention was rolled out in 9 clinics and studied in a stepped-wedge trial. The intervention was associated with an 7% percentage point reduction in the expected number of emergency department visits after controlling for other factors and temporal trends (p=0.005). The intervention group experienced 1,237 fewer emergency visits over the 6 month post-period for an estimated cost of care reduction of $1,718,193.
Describe the new knowledge and additional skills the participant will gain after attending your presentation.:
Participants will understand how implementing risk stratification to identify patients with the highest need for social services can result in substantial cost savings.
Authors:
Joshua Vest (Presenter)
Indiana University
Paul Halverson, Indiana University
Lisa Harris, Indiana University
Dawn Haut, Eskenazi Health
Nir Menachemi, Indiana University