Description
Predictive analytics can provide crucial information for decision making about patients with multiple chronic conditions (MCC). However, it also raises an important question about the appropriate representation of MCCs. We present a novel representation of MCCs and their treatment pathways that utilizes known relationships among various entities in EHR data to capture disease severity in a natural way. Results show that the proposed representation significantly outperformed traditional methods in predicting mortality and major cardiovascular events.
Learning Objective: After participating in this session, the learner should be better able to:
1) Understand multiple chronic conditions (MCC) and its burden on patients.
2) Learn the known treatment pathways of MCC in electronic health records (EHR) and other interactions
3) Understand survival models with age as the time scale
4) Learn a new representation of MCC and treatment pathways for predictive analytics
Authors:
Che Ngufor (Presenter)
Mayo Clinic
Pedro Caraballo, Mayo Clinic
Thomas O’Byrne, Mayo Clinic
David Chen, Mayo Clinic
Nilay D. Shah, Mayo Clinic
Michael Steinbach, University of Minnesota
Gyorgy Simon, University of Minnesota
Presentation Materials: