Times are displayed in (UTC-04:00) Eastern Time (US & Canada)Change
Early prediction of patient admission likelihood in emergency departments
Efficiently transitioning patients from the emergency department (ED) to inpatient units is a critical challenge in the US, often contributing to ED overcrowding. Early prediction of patients' admission likelihood and the demand for inpatient units enables hospital managers to apply proper capacity management strategies before real changes in demand happen. Our developed prediction model incorporates clinical, demographic, and visit-related information, as well as unstructured triage notes. This model serves as a decision support tool to provide the inpatient unit managers with near real-time predictions of their unit's demand when the patients are still in the early stages of their caregiving process in the ED. The study proves that predicting the likelihood of admitting patients to inpatient units will help identify the system's changing status and make proactive decisions before any surge in demand happens. Such a responsive decision-making approach can improve healthcare systems' operational efficiency and capacities.
Author(s):
Farzad Zeinali Kevin Taaffe | Professor | Clemson University Chris Gaafary | Prisma Health Ronald Pirrallo | Dr | Prisma Health William Jackson | Prisma Health Michael Ramsay | Dr | Prisma Health Jessica Hobbs | Prisma Health
Early prediction of patient admission likelihood in emergency departments
Category
Abstract Submission
Description
Primary Track: Data Analytics and Information Systems