New OpenEvidence study shows it outperforms general-purpose LLMs in health-tech research
A recent study funded by OpenEvidence compared its platform to general-purpose large language models (LLMs) in health technology research and concluded that OpenEvidence delivered more accurate and reliable results.
The findings directly challenge an earlier high-profile study released weeks before, which had suggested LLMs were on par with specialized tools.
In the same newsletter, Dexcom announced FDA clearance for its continuous glucose monitor to be used in children, expanding its pediatric market.
An additional survey of American Medical Association members highlighted growing interest in wearable devices among clinicians.
This writeup was produced by pharmadog from original reporting by STAT.
Original headline: “STAT+: OpenEvidence backs study that finds it beats LLMs”
read at STAT ↗
comments(0)
5-min edit window · permanent after that