The National Institute of Diabetes and Digestive and Kidney Diseases awarded Philadelphia-based Temple University $2.5 million to develop EHR-based tools that help better predict which patients with diabetes are more likely to need additional hospital visits.
The research team will build from the Diabetes Early Readmission Indicator, a model previously generated by Temple, according to the Dec. 10 news release. Using machine learning technology, the researchers aim to improve the models to predict 30-day readmission risk among patients with diabetes, based on EHR data.
The new models will use data from the National Patient-Centered Clinical Research Network's clinical data research network for the project. Once the new models are finished, the researchers will translate them into a readmission risk prediction tool that automatically captures EHR data and identifies risk factors for individual patients with diabetes.
The tool will be tested in patients with diabetes at Temple University Hospital and built in its Epic EHR system.