A project to predict emergency department treatment times for the Erdos Institute Data Science Bootcamp, Spring 2025
Using the emergency department (ED) and hospital data from Mimic-IV https://mimic.mit.edu/docs/, we will build an interpretable model to predict total treatment time for a patient based on patient characteristics gathered before and including triage. This will allow ED administrators and researchers to better schedule patients with regards to limited resources like rooms, tests, and staffing.
Our objectives are to:
- Predict the total length of time a patient will spend in the ED
- Identify key patient characteristics that predict treatment time
- Understand how treatment time predictions interact with admission times
Stakeholders: hospital administrators, healthcare practitioners, insurance companies
- Zhaolong Han, University of California - San Diego
- Haley Kottler, University of Wisconsin - Madison
- Dr. Emilie Wiesner, Ithaca College
- Haoyu Zhang, University of California - San Diego
- Yiming Zhang, University of California - San Diego