The use of artificial intelligence in hospitals is ramping up so fast — and with such little transparency — that it is impossible to track how any given product is impacting the cost or quality of care. Whether AI is monitored at all is entirely up to individual health systems.
“Everybody is doing this differently, and many health systems are not doing it,” said Michael Pencina, chief data scientist at Duke Health. The lack of systematic monitoring of AI in medicine creates safety and financial risks, he said, because poorly performing products can easily escape scrutiny.
Pencina is among those proposing to fill the void with a new approach — a national registry of AI tools that would list where products are being used and provide background information about their development and performance. It is not an entirely novel idea.
This article is exclusive to STAT+ subscribers
Unlock this article — and get additional analysis of the technologies disrupting health care — by subscribing to STAT+.
Already have an account? Log in
