Abstract: In an attempt to improve emergency department (ED) fall screening tools, we created a short fall screening questionnaire based on prior literature and evaluated the effect of this tool’s integration into an existing machine learning algorithm in predicting return visits to the emergency department for a fall. Incorporating the screening tool data to a machine learning approach slightly increased the performance of predictive models compared to a machine learning approach alone.

Describe the new knowledge and additional skills the participant will gain after attending your presentation.: Attendees should be able to understand the impact of introducing targeted questionnaire-style data on machine learning algorithms in predicting emergency department return visits for a fall.

Authors:

Alexander Clegg (Presenter)
University of Wisconsin School of Medicine and Public Health

Collin Engstrom, University of Wisconsin School of Medicine and Public Health
Brian Patterson, University of Wisconsin School of Medicine and Public Health