Developing and Validating EHR-integrated Readmission Risk Prediction Models for Hospitalized Patients with Diabetes (eDerri) )

Study PI(s): Dan Rubin and Anuradha Paranjape, Temple University/Temple University Health System

Study Design: Observational Study

Project Summary: Hospital readmission is an undesirable, costly outcome for both patients and hospitals that is often preventable. Patients with diabetes are at higher risk of readmission within 30 days of hospital discharge than patients without diabetes. This project aims to develop new versions of the DERRI model enhanced by EHR (electronic health records) data (eDERRI) that predict unplanned 30-day all-cause hospital readmission risk among patients with diabetes.

Using data from all participating PaTH sites, the aims are:

• To build an automated eDERRI tool integrated with an EHR in EPIC at Temple
• To prospectively validate the eDERRI tool at two sites with different EHR platforms
  • Temple in EPIC
  • Penn State in Cerner
• To build an automated, EHR-integrated eDERRI tool that automatically captures EHR data to predict readmission risk

PaTH Partners:

• Temple University / Temple University Health System
• University of Pittsburgh / UMPC (Mary Kortikowski, Site PI)
• Penn State University / Milton S. Hershey Medical Center (Wenke Hwang, Site PI)
• Geisinger Health System (Les Kirschner, Site PI)
• Johns Hopkins University (Nestoras Mathioudakis, Site PI)

Sponsor: NIH