ECRI Expands Reporting Network for AI Errors in Patient Care to Address Growing Healthcare Challenges

ECRI Enhances Problem Reporting Network to Address AI Errors in Healthcare

In response to the increasing integration of artificial intelligence (AI) in the medical field, ECRI, a renowned nonprofit dedicated to patient safety, recently announced the expansion of its Problem Reporting Network. This initiative aims to specifically capture and investigate errors, malfunctions, and near misses associated with AI tools utilized in patient care. As healthcare systems increasingly rely on advanced technology, the call for vigilance in reporting AI-related incidents has never been more crucial.

ECRI is urging healthcare providers, systems, and clinicians across the nation to report any situation where an AI-enabled tool may have caused an error or posed a risk to patient care. By actively triaging and investigating these reports, ECRI will offer valuable insights back to each submitter. This expansion complements the already impressive efforts of the ECRI and ISMP Patient Safety Organization (PSO), which has analyzed over 8 million safety event reports from healthcare professionals.

The Promise and Perils of AI in Healthcare

AI is currently revolutionizing healthcare by enhancing clinical outcomes and increasing operational efficiency. However, this rapid deployment of AI tools and devices has outstripped the existing safeguards intended to prevent, identify, and respond to potential failures. AI's roles in healthcare range from assisting with diagnostic imaging to providing clinical decision support, and even functioning as personal chatbots that interact with patients. Nonetheless, there is currently no centralized mechanism to monitor the accuracy and reliability of these AI systems, leading to concerns over how often they produce incorrect or misleading outputs that may affect patient care.

Scott Lucas, PhD, Vice President of Devices, Therapeutics, and Technology at ECRI, highlighted the need for thorough scrutiny before adopting AI. He stated, “While we recognize the immense potential of AI, we still lack a comprehensive understanding of its impact on healthcare. Without a solid foundation of data and analysis, the industry cannot adequately refine the design and integration of these tools.” This sentiment underscores the importance of gathering evidence to navigate the evolving risks associated with the use of AI in healthcare settings.

Survey Highlights Concerns Over AI Outputs

A recent survey conducted by ECRI involving 124 healthcare leaders—primarily focusing on quality, safety, risk, or compliance—revealed a mix of confidence and uncertainty regarding AI outputs. Approximately one-third of respondents reported encountering AI-generated outputs that they believed to be incorrect or misleading within the past year. Notably, only 9% acknowledged that an AI error had directly reached a patient or influenced a clinical decision. The survey revealed that ambient scribes were the most frequently encountered AI tools, followed by embedded clinical decision support systems and clinical large language model (LLM) assistants or chatbots.

ECRI has conducted evaluations on various AI-enabled devices and tools related to medical settings, including applications in imaging, fall prevention, anesthesia, and wearable technology for chronic diseases. With the aim of ensuring patient safety, their evaluations help identify efficacy and reliability within AI applications.

How ECRI's Problem Reporting Network Operates

ECRI's Problem Reporting Network has been a cornerstone for the healthcare community since its inception in 1972. This initiative provides a confidential platform through which issues related to medical devices and technology can be reported. Each submission undergoes triage and investigation by ECRI’s team of clinical and engineering experts. Importantly, ECRI follows up with submitters to communicate the findings and insights derived from the reports.

When risks associated with a device or technology are identified, ECRI promptly informs all relevant stakeholders, including manufacturers and regulatory agencies, via hazard reports that detail the issues and recommend actions to mitigate risk. Annually, the Problem Reporting Network contributes to vital research reports that highlight the most pressing health technology hazards and patient safety concerns.

With this recent expansion, healthcare providers now have a dedicated pathway to submit reports concerning suspected AI-related errors or unsafe AI behaviors encountered in practice. Clinicians and healthcare professionals who suspect AI tools have contributed to an error—whether it reached a patient or was identified beforehand—are highly encouraged to report their experiences to ECRI. In addition, ECRI emphasizes the necessity of submitting safety reports through all relevant state and federal reporting channels.

For further information regarding the safe use of AI in healthcare, interested parties can visit ECRI's AI Resource Hub.

About ECRI

ECRI is an independent nonprofit organization dedicated to enhancing the safety, quality, and cost-effectiveness of healthcare across various settings. With a rich history spanning over six decades, ECRI is recognized for its integrity and commitment to evidence-based insights. Designated as an Evidence-based Practice Center by the U.S. Agency for Healthcare Research and Quality, ECRI's credibility ensures it is at the forefront of advancing healthcare safety standards. In addition to its extensive safety initiatives, ECRI has made progressive strides by acquiring organizations such as the Institute for Safe Medication Practices (ISMP) in 2020 and The Just Culture Company in 2024, solidifying its position as a premier resource in healthcare quality and safety.