Children aren’t just little adults, goes the saying in pediatrics. But with little financial incentive for companies to develop kid-specific treatments, pediatricians are often left to use adult medicines and medical devices off-label for their patients — or none at all.
As medical devices powered by artificial intelligence and machine learning have exploded in the last decade, “we were concerned that this same pattern would emerge,” said Ryan Brewster, a pediatrics resident at Boston Children’s Hospital.
In a new analysis of AI and ML-enabled devices authorized by the Food and Drug Administration, Brewster and colleagues at Boston Children’s and Harvard Medical School found just that. Digging into FDA documents — a process that took half a year — they found that a minority of AI and ML devices, 149 of 876, were clearly labeled for use in kids under 18. Among those pediatric devices, there were gaps in the data used to validate the algorithms, calling into question their efficacy in younger patients.
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