At HiMSS19, Michael Sanders, FAAFP, MD, chief medical and informatics officer of Flagler Hospital, led a session on the hospital’s progress in tackling clinical variation with artificial intelligence (AI).

Applying Ayasdi's Artificial Intelligence (AI)-powered Clinical Variation Management (CVM) software application to pneumonia, Flagler Hospital saved an average of $1,350 per case, reduced the average length of stay by two days and reduced readmissions by seven times – eliminating nearly $850,000 in unnecessary costs.

Flagler Hospital, drawing data from its electronic medical record (EMR) and using the Fast Healthcare Interoperability Resources standard, tapped into unsupervised learning capabilities with CVM to understand the optimal events, sequence and timing of care. These results were presented to physician teams via an intuitive interface that allowed them to understand exactly why each step (and the timing of the step) was recommended. Final care paths were deployed as new order sets to the EMR.

The intelligent application has also been applied to sepsis and will be used with 18 other conditions in the coming months.

"Our ability to rapidly construct clinical pathways based on our own data and measure adherence by our staff to those standards provides us with the opportunity to deliver better care at a lower cost to our patients. Given the complexity of the challenge, using AI to solve this problem is the right approach, and Ayasdi's CVM application makes it accessible to our practitioners, a key part of the value proposition given Flagler does not have dedicated data science resources,” said Sanders.