Regulatory Approaches to Digital Medical Devices in the U.K., EU, and U.S.

New digital technologies are revolutionizing all sectors, including healthcare, with the potential application of artificial intelligence (AI) and machine learning (ML) leading to advancements in diagnostic tools, personalized treatments, sensor-based remote monitoring, and medical data analysis. 

However, regulating these emerging technologies is complex, as authorities find themselves navigating intricate definitions, potential harm to sources from bias or data leakage, and tasked with defining the distinction between medical and non-medical tools. Regulators are aiming to balance safety with innovation through initiatives such as the FDA’s pre-determined change control plan and the U.K.’s regulatory sandbox for AI-based medical devices.

Potential Application of New Technologies

AI is poised to transform healthcare in several areas, spanning from diagnosis and treatment to patient monitoring and management. In diagnostic imaging, AI and ML algorithms can analyze medical images such as X-rays, MRI scans, CT scans, and ultrasounds to identify specific features or structures, aiding in the diagnosis of various conditions. For remote patient monitoring, AI-powered devices can continuously gather data on vital signs, activity levels, and other metrics, enabling early detection and timely intervention.

Personalized medicine is also benefiting from AI/ML algorithms that analyze patient data, including genetic information, medical history, and lifestyle factors, to create tailored treatment plans. Clinical decision support systems (CDSS) are increasingly utilizing AI/ML to examine patient data, assisting in diagnosis, treatment planning, and decision-making. Wearable health devices equipped with AI/ML can monitor health parameters such as heart rate, blood pressure, sleep patterns, and physical activity. In robotic surgery, AI-powered robots can assist surgeons in minimally invasive procedures by enhancing precision, dexterity, and control. 

Predictive analytics for healthcare management involves AI and ML models analyzing large volumes of data, including electronic health records, insurance claim data, and operational metrics, to identify patterns, trends, and risk factors, ultimately improving healthcare management. 

As healthcare systems become more crowded, AI-powered devices can assist professionals with remote patient monitoring by automatically interpreting data on vital signs and other metrics, enabling early detection of negative health trends that could lead to early intervention and therefore help reduce the incidence of more serious health outcomes. 

In underfunded areas of medical research, AI and ML could be transformative, helping to detect co-morbidities and environmental or genetic factors that increase disease risk for certain individuals. Beyond early detection, these technologies could potentially predict disease risk years in advance, enabling preventive treatment or protective lifestyle changes.

Regulatory Approaches in Key Markets

Despite all potential AI/ML applications in the healthcare sector, implementing the technology can pose challenges to manufacturers. This starts with distinguishing digital health tools from digital medical devices. Digital health encompasses a spectrum of technologies, ranging from non-medical devices designed to monitor well-being through to medical devices tailored for specific medical purposes. The many mindfulness apps, sleep tracking software and many other apps flooding the internet, for example, are usually not medical devices. However, a change in a single word within a marketing claim can sometimes be the distinguishing factor between a digital health product and a regulated medical device. 

The International Medical Device Regulators Forum defines Software as a Medical Device (SaMD) as “software intended to be used for one or more medical purposes that perform these purposes without being part of a hardware medical device.” Regulatory authorities for medical devices have been developing guidance to help clearly identify SaMD differentiating them from that wide and confusing range of other tools such as well-being and lifestyle software products, IVD software, and companion diagnostics.¹ However, even with available guidance identifying whether a product is a medical device or digital health product can still be challenging for manufacturers.

Europe and the EU AI Act

In Europe, the upcoming EU AI Act will also encompass medical devices—possibly introducing additional complexities and enhancing scrutiny for AI-based devices—that now typically fall under Class IIa classification. 

The recently published final text for the AI Act specifies one of the core objectives for healthcare AI regulation: “in the health sector where the stakes for life and health are particularly high, increasingly sophisticated diagnostics systems and systems supporting human decisions should be reliable and accurate. The extent of the adverse impact caused by the AI system on the fundamental rights protected by the Charter is of particular relevance when classifying an AI system as high-risk.”²

Due to wider definitions of high-risk AI uses, certain digital health products may also be subjected to CE marking regulation for the first time under the AI Act regulatory assessment of high-risk AI systems, where they are not considered medical devices. The legislation specifically addresses certain digital health technologies as being high risk, for example AI systems used for emergency healthcare patient triaging and systems used in evaluating eligibility for certain healthcare services. These digital health products will now require notified body assessment and CE marking as an AI system, which will be based on different criteria to existing medical device requirements.

The FDA’s Benefit-Risk Framework

The U.S. Food and Drug Administration maintains a publicly available list of AI enabled tools with marketing clearance and applies a “benefit-risk” framework. This framework confirms that devices must conform to some basic principles such as demonstrating sensitivity and specificity for diagnostic devices, validating intended purpose and stakeholder requirements against specifications, and development that ensures repeatability, reliability, and performance. The FDA has also introduced pilot processes allowing pre-authorized software changes such as re-training to be deployed without additional regulatory assessment,³ addressing the unique nature of AI models and providing proportionate regulation. This is a significant milestone in regulatory innovation, as traditional assessment methods still used in EU medical device assessments can struggle to foster similar effective and proportionate regulation.

UK: Regulatory Sandbox and Innovation-Boosting Programs

The U.K.’s pilot regulatory sandbox, announced last fall, provides a safe space for healthcare AI tool developers to trial products under regulatory oversight before implementation.⁴ This initiative aims to foster innovation while updating regulatory safety standards safety. The MHRA also plans to develop a guidance-based system to allow for more frequent updates to keep abreast of technological advancements in AI systems, collaborating with international regulators to promote high standards in AI/ML practice. Alongside the FDA and Health Canada, the MHRA has outlined 10 guiding principles for developing Good Machine Learning Practices (GMLP) that are safe, effective, and promote high-quality AI/ML-enhanced medical devices.⁵

To turn the U.K. into an international hub for responsible medical device software innovation, the Software and AI as a Medical Device Change Programme has been updated, to achieve safety assurances, defining clear guidance and processes for manufacturers and liaising with key partners like NICE, NHS England, and the IMDRF (International Medical Device Regulators Forum). Addressing potential bias and inequalities, the MHRA also recognizes that SaMD and AIaMD must perform across all populations within the device’s intended use and serve the needs of diverse communities.

Given the novelty and complexity of regulating AI-based medical devices, businesses should determine early in development whether AI/ML functionalities are integral or supplementary to product functionality, as this impacts the level of regulatory scrutiny performed by medtech regulators.

Additionally, manufacturers should consider partnering with professionals who possess in-depth knowledge of regional regulations and emerging trends in digital health to expedite market entry and ensure compliance with evolving standards and requirements. By embracing collaboration and expertise, innovators can navigate regulatory challenges and pave the way for transformative digital health solutions.

References

¹ Gov.co.uk, Software and AI as a Medical Device Change Programme – Roadmap, June 2023

² EUR Lex, Proposal for a Regulation laying down harmonised rules on artificial intelligence, (27)

³ bit.ly/4eZL7rp

⁴ Gov.co.uk, MHRA to launch the AI-Airlock, a new regulatory sandbox for AI developers, 1st November 2023

⁵ Gov.co.uk, Good Machine Learning Practice for Medical Device Development: Guiding Principles

With 14 years of experience in the medical device industry, Timothy Bubb has built a wide breadth and depth of knowledge across multiple disciplines spanning regulatory, engineering, clinical, design and development, and quality assurance. Bubb is passionate about fostering medical device innovation, combining commercial insight, pragmatism, and solution-oriented approaches to help clients improve patient care and outcomes worldwide. He has held various leadership roles within the industry, focusing on achieving and maintaining market access in global territories. One of Bubb’s key areas of focus is on software and hardware, including AI/ML, with experience bringing complex lifesaving and life-enhancing products to market. Additionally, he is active in knowledge sharing and training programs within the industry and academia, addressing challenging topics in medical device regulation, safety, and compliance for novices and experts alike.