Overview
- ECRI formally expanded its Problem Reporting Network on Aug. 25 to include a dedicated pathway for errors, malfunctions, and near misses tied to clinical AI tools and AI-enabled devices.
- The organization is asking clinicians, health systems, and risk managers to report any incident where an AI tool may have contributed to a mistake or introduced risk in care delivery.
- Each submission will be reviewed by ECRI’s clinical and engineering experts who follow up with the reporter and may issue hazard advisories to manufacturers, providers, and regulators if a systemic safety risk is found.
- A recent ECRI survey of 124 hospital safety and quality leaders found 31% encountered an AI output they believed was incorrect or misleading in the past year and 9% said an AI error reached a patient, with 35% unsure whether errors had occurred.
- ECRI’s move aims to fill a national post-market surveillance gap for clinical AI by building a centralized evidence base that can reveal design or integration failures, change vendor behavior, and help hospitals spot problems before patients are harmed.