With so much money flowing into healthcare AI and funding many unproven solutions that will ultimately fail, a correction will be good for medicine, a Health Affairs commentary argues.
AI spending in healthcare outpaces other sectors by a factor of 2.2, reaching $1.4 billion in 2025, triple the year before, the authors noted in the Feb. 4 article, “The Coming Clinical Correction: Why Health Care Needs Its AI Bubble To Burst.” But when many of these startups eventually go belly up, health systems will be stuck with the contracts, using algorithms based on old data.
“Tools that provide measurable, reproducible benefits will survive: ambient scribes that verifiably reduce documentation burden by hours, not minutes; diagnostic aids with peer-reviewed accuracy improvements; and workflow systems with demonstrated ROI,” wrote Maxim Topaz, PhD, RN, Jing Dong, PhD, and Zhihong Zhang, PhD, RN, of New York City-based Columbia University and Laura-Maria Peltonen, PhD, RN, of University of Eastern Finland in Kuopio. “Like telehealth finding its home in behavioral health and chronic care management, AI will settle where outcomes justify adoption.”
Health systems can prepare for the potential bursting of the AI bubble by incorporating vendor stress testing, setting up internal AI governance, and implementing operational contingency plans, the authors said.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.