A Tale of Two AI Implementations in Healthcare
A Tale of Two AI Implementations in Healthcare

Abstract: 

Implementing artificial intelligence into healthcare is a very difficult task. In this presentation, we will discuss how we implemented two predictive models based on machine learning at Fresenius Medical Care North America, a company that provides dialysis to approximately 200,000 patients with kidney failure. The two models are: 1) predicting which patients are at risk for all-cause hospitalization in the next 7 days and 2) which patients are at risk for peritonitis (i.e. infection) in the next 30 days. One of these models had high performance on area-under-the-curve (AUC), the other did not. One had actionable reasons, the other did not. One was rolled out abruptly, the other gradually. One integrated end-user feedback into the final product, the other did not. One had thousands of users, the other had less than a hundred. One had actionable metrics tracked throughout development, the other did not. One was productionized through a different department, the other was not. One was expanded, the other was terminated. Come to the talk to listen to challenges and lessons learned about implementing AI in healthcare.

Bio: 

Andrew Long is a Data Scientist at Fresenius Medical Care North America (FMCNA). Andrew holds a PhD in biomedical engineering from Johns Hopkins University and a Master’s degree in mechanical engineering from Northwestern University. Andrew joined FMCNA last year after participating in the Insight Health Data Fellows Program. At FMCNA, he is responsible for building predictive models using machine learning to improve the quality of life of every patient who receives dialysis from FMCNA. He is currently creating a model to predict which patients are at the highest risk of imminent hospitalization.