A new risk on top of ECRI’s annual list of health technology hazards: AI

For almost two decades, the global healthcare safety organization ECRI has published its list of Top 10 Hazards of Health Technology. This year’s edition is headlined by a rapidly changing technology with enormous promise for healthcare but no shortage of downside risks: artificial intelligence.

WHY IT’S IMPORTANT
“While AI has the potential to improve efficiency and outcomes, it poses significant risks to patients if not properly assessed and managed,” ECRI said in the announcement of the 2025 report.

“AI has expanded from its early applications in medical imaging to impact virtually every area of ​​healthcare, including diagnosis, documentation and appointment scheduling. Even AI applications involving support systems, which are not regulated as medical devices, can have a profound impact on patient care,” ECRI said.

Of particular concern, of course, are AI hallucinations or other misleading feedback from improperly calibrated algorithms. The output of machine learning models — especially if those models are trained on biased data — can vary widely across patient populations, researchers note, posing risks to health care equity and patient safety for underrepresented or underserved communities.

ECRI’s top 10 health technology hazards for 2025:

  1. Risks with AI-enabled health technologies

  2. Unmet technology support needs for home care patients

  3. Vulnerable technology suppliers and cyber threats

  4. Substandard or fraudulent medical devices and supplies

  5. Fire hazard due to additional oxygen

  6. Dangerously low default alarm limits on anesthesia machines

  7. Improper handling of temporary deductions from medication orders

  8. Poorly managed IV lines

  9. Harmful medical adhesive products

  10. Incomplete investigation into incidents involving infusion systems

ECRI defines a health technology hazard as any “flaw in a device or system, design feature or method of use that could, under certain circumstances, endanger patients or users.”

The group takes what it calls a “total systems approach to safety,” with the goal of helping healthcare professionals, administrators, device manufacturers, policy makers, researchers and patients themselves reduce incidents of preventable harm during the delivery of care.

The teams focus on various aspects – human factors, technology, device safety, medication safety, infection control – that influence and arise from technology implementation in healthcare, and work to recommend system-wide safety solutions.

ECRI notes that the topics and technologies listed each year “are not necessarily the most reported problems or those associated with the most serious consequences”, although these factors are a core consideration.

Instead, the annual report reflects his assessment of “which risks should receive attention now to help healthcare providers, device manufacturers and others prioritize their patient safety efforts.”

The full Top 10 Health Technology Hazards report, for ECRI members, provides detailed steps that healthcare systems, vendors and other IT leaders are taking to reduce risks to patient safety. Access to an executive briefing here.

THE BIG TREND
The ECRI 2024 list had some similar themes, focusing on the risks associated with remote monitoring of patients and other medical devices at home (#1) and inadequate management of AI (#5). In recent years, we’ve focused on the dangers of IV pumps and software management gaps, and of course, cyber threats like ransomware.

ON THE RECORD
“The promise of artificial intelligence’s capabilities should not distract us from its risks or its potential to harm patients and healthcare providers,” said Dr. Marcus Schabacker, president and CEO of ECRI, said in a statement. “Balancing AI innovation with privacy and security will be one of the most difficult and defining endeavors of modern medicine.

“AI is only as good as the data given to it and the guardrails that govern its use,” he added. “Healthcare stakeholders at all levels should think critically about the integration of AI, as they would with any new technology.”

Mike Miliard is editor-in-chief of Healthcare IT News
Email the writer: mike.miliard@himssmedia.com

Healthcare IT News is a HIMSS publication.