Where health systems are spending their AI dollars

Hospitals and health systems are largely using artificial intelligence for purposes like patient scheduling and disease prediction, several digital and data executives told Becker's.

That lines up with findings of an Oct. 18 Deloitte report on the state of AI, which found the top healthcare applications include customer service operations and computer-assisted diagnostics.

Winston-Salem, N.C.-based Novant Health, for example, is using AI to optimize scheduling and cut down on patient wait times. The health system also employs AI for its hospital-at-home program and other remote monitoring efforts.

"By combining our diagnostics, predictive health scoring, as well as vitals management, we are looking for being able to move as much to avenues that will provide the patient with a more comfortable and familiar environment," said Karl Hightower, chief data officer and senior vice president of data products and services of Novant Health.

According to the Deloitte report, the top AI uses for patient experience and marketing were customer service operations and image recognition/digital radiology (tie, 41 percent); computer-assisted diagnostics (40 percent); personalization and patient vitals monitoring (tie, 38 percent); and omnichannel experience management (35 percent). The consulting firm surveyed 260 global healthcare and life science organizations from across the globe.

"The leading industry-specific healthcare AI use cases focus on outcomes and monitoring in such potentially transformational areas as AI-assisted diagnoses including predictive diagnoses, patient engagement, insurance fraud detection and smarter hospitals," the report stated.

The data and digital health leaders interviewed by Becker's found these categories to be largely representative of their uses of AI.

"Mayo Clinic has AI projects that fall under the diagnostics, customer service, and patient vitals monitoring categories, but we also have many AI initiatives that would be considered operational, or administrative, uses," said Ajai Sehgal, chief data and analytics officer of the Rochester, Minn.-based health system.

Those include patient volume prediction, resource allocation, and billing. At Mayo's Center for Digital Health, most use cases involve natural language processing, taking spoken word or text and turning it into a structured format.

Mayo Clinic's digital front door app has a natural language processor to match a patient's primary concern to the health system's scheduling tree to expedite new appointments.

New York City-based NewYork-Presbyterian Hospital has deployed such tools as predictive machine learning, robotic process automation and AI imaging.

"Particularly, over the course of the last several years, and because of the COVID-19 pandemic, we've heavily leveraged AI within our customer service operations," said Peter Fleischut, MD, senior vice president and chief transformation officer of NewYork-Presbyterian.

Its NYP Connect App offers self-scheduling, EHR access and telehealth visits. Patient-facing calls are automated through scheduling and conversational AI.

Most of the AI efforts at Duarte, Calif.-based City of Hope fall under clinical decision support, a combination of computer-assisted diagnostics, vital sign monitoring and risk detection, said Chief Digital Officer Mark Hulse.

City of Hopes combines AI, genomics and other clinical data to predict the risk of sepsis for transplant patients, 30-day unplanned readmissions, surgical complications, and survival and early disease. AI also helps forecast financial metrics and patient readiness for early discharge.

At Sioux Falls, S.D.-based Sanford Health, the enterprise data analytics team has developed algorithms to try to predict which patients might be at risk of Type 2 diabetes or which women should have mammograms earlier than recommended.

"We are applying predictive models and risk stratification to our electronic medical record that consider and review all aspects of patients' histories to inform clinical decision-making," said Doug Nowak, vice president of enterprise data analytics for Sanford Health.

Jacksonville, Fla.-based Baptist Health works with AI to cut back on repetitive tasks for employees and optimize operating room scheduling.

"My prediction is that as data is shared more widely, coupled with significant increases in computational ability, and we turn towards utilization of quantum computing for big data analysis, AI will truly be unlocked with numerous more categories and hard return-on-investment use cases," said Aaron Miri, Baptist Health's senior vice president and chief digital and information officer.

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.


Featured Whitepapers

Featured Webinars