Description
This retrospective study aims to predict the disposition decisions (discharge home, admit to hospital floor, or admit to intensive care unit) for patients visiting an emergency department. We use available patient-specific electronic health record data at different acute care delivery stages to develop and validate progressive predictive models with increasing accuracy. We also demonstrate the contribution of CCMapper, a natural language processing-based chief complaint mapping algorithm, in improving the performance of prediction models.
Learning Objective: After participating in this session, the learner should be better able to understand the potential role of natural language processing and machine learning in early prediction of disposition decision and to discuss how designing such tools, capable to be integrated into the electronic health record systems, can contribute to reducing overcrowding and boarding time in an emergency department.
Authors:
Mohammad Samie Tootooni (Presenter)
Mayo Clinic
Kalyan Pasupathy, Mayo Clinic
Heather Heaton, Mayo Clinic
Casey Clements, Mayo Clinic
Mustafa Sir, Mayo Clinic
Presentation Materials: