Medical Device Regulations

🚨 BREAKING: Over 10,000 #digital tools in use across the #NHS could be putting #patients at #risk!🚨 Just published, our national study reveals an uncomfortable truth: 3 out of every 4 #digitalhealth #technologies used in NHS trusts don’t meet minimum #clinical #safety or #legal requirements. Yes, you read that right: tens of thousands of apps, telehealth tools, devices and platforms lack assurance against critical risks that could harm patients. Youssof Oskrochi Elliott Roy-Highley Dr Keith Grimes UCL Global Business School for Health have carried out this eye opening research. Despite mandatory standards (DCB0129 & DCB0160) required by law in the NHS since 2012, the scale of noncompliance is staggering. The absence of proper #safety assessment and documentation isn’t hypothetical, it’s happening NOW!!! Legacy systems and new technologies alike are left unchecked, and the impact? Unseen, unreported, and unquantifiable risks to every patient and every #clinician relying on digital care. With digital #transformation at the heart of the NHS Long Term Plan, we must demand robust #governance and #accountability. It’s time for NHS leaders, #regulators, and #innovators to step up—before operational convenience trumps patient #safety. If we don’t fix this compliance debt, when, not if, harm occurs, the #legal consequences will be severe and public trust will be shattered. Digital health is the future but only if we make safety the top priority. Read the research: https://lnkd.in/euJdQxfP #NHS #DigitalHealth #PatientSafety #ClinicalRisk #Innovation #Leadership #MedTech #HealthPolicy #regulation

In 2023, the FDA issued draft guidance on "predetermined change control plans for AI medical devices". This new update will allow for modifications to be made without changing the efficacy or safety of the medical device, provided these changes align with the predetermined plan. This is a crucial step in dealing with the fast-paced evolution we see with continuously learning algorithms. We talked to 𝐖𝐞𝐫𝐨𝐧𝐢𝐤𝐚 𝐌𝐢𝐜𝐡𝐚𝐥𝐮𝐤, 𝐒𝐚𝐌𝐃 𝐋𝐞𝐚𝐝 at HTD Health to better understand the challenges in regulating AI algorithms in healthcare. "𝐖𝐡𝐚𝐭 𝐬𝐭𝐞𝐩𝐬 𝐜𝐨𝐮𝐥𝐝 𝐰𝐞 𝐞𝐱𝐩𝐞𝐜𝐭 𝐟𝐫𝐨𝐦 𝐭𝐡𝐞 𝐅𝐃𝐀 𝐢𝐧 𝟐𝟎𝟐𝟒 𝐢𝐧 𝐫𝐞𝐠𝐮𝐥𝐚𝐭𝐢𝐧𝐠 𝐀𝐈? In 2024, the FDA appears to be steering a proactive course to address the complexities of regulating AI in healthcare. A key initiative is the ambitious plan to draft 18 new documents, which signals a robust effort to provide clearer guidelines and standards for stakeholders navigating the realms of AI, cybersecurity, and other digital health technologies. This move reflects the FDA's recognition of the rapidly evolving landscape of digital health and its commitment to fostering a well-regulated environment that both upholds patient safety and encourages innovation. In parallel, the FDA's IT Modernization Plan for 2024 to 2027 is set to enhance the agency's technological framework, aiming to better support clinical trials and sponsors. On the regulatory front, the FDA's intention to categorize more AI tools as devices showcases a move towards a tighter regulatory framework. This initiative aims at ensuring the safety and efficacy of these tools, reflecting the FDA's cautious approach to manage the potential risks associated with rapidly evolving AI technologies in healthcare. Meanwhile, the continuous approval of AI-enabled devices, especially in radiology, highlights the FDA's recognition of AI's potential to enhance diagnostic accuracy and efficiency. By continuing to approve these devices, the FDA is facilitating the integration of AI in healthcare, which is likely to lead to improved patient outcomes and streamlined workflow for healthcare providers." https://lnkd.in/eDK5auu2

📆 It's been a year since my last post identifying key peer-reviewed papers on regulation of novel AI (such as LLMs) that have informed, provoked or both! (https://lnkd.in/dDVgSHbt) 🃏 The subsequent months have been nothing short of frenetic with regards to evolutions in AI and there's been no shortage of perspectives on regulating such technology in the digital health setting. Here are a few of the key reads 📖 that have hit my radar since. 1️⃣ The Regulation of Clinical Artificial Intelligence by David Blumenthal and Bakul Patel. NEJM AI 2024;1(8) (https://lnkd.in/dnfVur83) They argue that existing regulatory frameworks are just about adequate for Pre-Generative AI but with Generative AI the shortcomings of these frameworks become apparent. One possible direction would be to treat these GAI-based clinical applications not as devices but as a new type of clinical intelligence: that is, to regulate them less like devices and more like clinicians. 2️⃣ Navigating the European Union Artificial Intelligence Act for Healthcare by Felix Busch, Jakob Nikolas Kather, Christian Johner et al. Nature Portfolio Digital Medicine volume 7, Article number: 210 (2024) (https://lnkd.in/dvQRxJSS) A comprehensive and accessible guide to the most important aspects of the AI Act, including practical examples from the medical field. 3️⃣ Regulate Artificial Intelligence in Health Care by Prioritizing Patient Outcomes by John W Ayers, Nimit Desai, Davey M Smith. JAMA 2024;331(8):639–640. (https://lnkd.in/dXiAE8BU) A strong push for patient outcomes to feature in the regulation and procurement of AI tools - "patients' feeling, function, and survival". AI technologies are too new to rely on regulating processes & the products are at times inscrutable, so let’s focus on outcomes. 4️⃣ Congress Must Update FDA Regulations for Medical AI by Scott Gottlieb JAMA Health Forum. 2024;5(7):e242691. https://lnkd.in/d-iz3pHT Gottlieb argues that the FDA should formally shift from a product-based to a firm-based approach for regulating AI, focusing on manufacturers' quality systems rather than dissecting individual AI products. This would potentially allow (and require) third-party certifications for low risk devices. 5️⃣ A future role for health applications of large language models depends on regulators enforcing safety standards by Oscar Freyer, Isabella Wiest, Jakob Nikolas Kather & Stephen Gilbert. The Lancet Digital Health. Volume 6, Issue 9, September 2024, Pages e662-e672 (https://lnkd.in/dfmVzNb4) What is the value of regulation without effective enforcement?! ☝🏾 Have all bases been covered? ☝🏾 What would you add to the mix?

Is the Evolution of Functionally Aggregated DHTs essentially an Ecosystem Challenge? The authors observe a phenomenon of "aggregated intended purposes" of digital health technologies (DHTs), or "device-aggregates," increasingly being applied in groups of clinical tasks and sub-tasks, from the perspective of regulatory approval. At the highest level, 'super device' aggregates or device suites may be 1) coupled to form loosely defined parts of digitally integrated care pathways, such as hospital-at-home, or 2) cascaded serially. Other pathways are participatory care and patient navigation pathways, and AI-powered anticipatory care pathways are important. This two-article analysis is significant because it highlights the gaps and key issues of regulatory, HTA and reimbursement aspects of data-coupled collaborative innovation. 🔷 Regulatory: Authors note the evolution from passive to active groupings. From cascaded effects to networked, interconnected devices with dynamic dependencies and combined effects that need to be regulated as such. The emergent "super devices" reduce human intervention, necessitating airtight regulation, especially considering the inclusion of non-MDs which are deregulated. Interpreting EU regulations, the “lead” manufacturers of super-MDs (SMD) would be responsible to obtain approval for all components, which could be impractical given their non-manufacturer status for some. 🔷 Reimbursement: Gathering cost-effectiveness evidence introduces new complexities. These include the absence of comparators and the complex estimation of initial investments. Ongoing performance monitoring might solve part of the problem but in the absence of evidence ecosystem standards this will be highly impractical. 🔷 Inclusive evidence: In addition to regulating emergent system properties that arise in interactions, building, testing and evaluating super-MDs in primary care and public health settings and pathways is a limitation. Part two observes the following modalities: 1️⃣ Single manufacturer develops and seeks approval for SMD/components to perform a specific function.   2️⃣ Multiple manufacturers develop approved components brought together and placed on the market by a single commercial entity. 3️⃣ Multiple manufacturers develop approved components brought together and placed on the market as a service provided by a single commercial entity. 4️⃣ Multiple entities brought together flexibly and dynamically and possibly also automatically. As (4) points to a collaborative innovation ecosystem, an overarching challenge emerges: the requirement for regulatory and HTA pathways built on evidence sandboxes and regulated evidence ecosystems, leveraging data frameworks for data governance such as IEEE’s P3493.1™.   PART-1 https://lnkd.in/dv78qpnK PART-2: https://lnkd.in/dVrCN24w #HealthcareInnovation #DigitalHealth #InnovationEcosystem #MDR #SaMD #RegulatoryPolicy #HTA

5 𝐁𝐮𝐫𝐧𝐢𝐧𝐠 𝐐𝐮𝐞𝐬𝐭𝐢𝐨𝐧𝐬 𝐍𝐨 𝐎𝐧𝐞 𝐢𝐧 𝐃𝐢𝐠𝐢𝐭𝐚𝐥 𝐇𝐞𝐚𝐥𝐭𝐡 𝐖𝐚𝐧𝐭𝐬 𝐭𝐨 𝐀𝐧𝐬𝐰𝐞𝐫 (𝐁𝐮𝐭 𝐖𝐞 𝐂𝐚𝐧’𝐭 𝐈𝐠𝐧𝐨𝐫𝐞) 🚀 1️⃣ 𝐃𝐚𝐭𝐚 𝐏𝐫𝐢𝐯𝐚𝐜𝐲 𝐯𝐬. 𝐀𝐜𝐜𝐞𝐬𝐬𝐢𝐛𝐢𝐥𝐢𝐭𝐲: 𝐂𝐚𝐧 𝐖𝐞 𝐄𝐯𝐞𝐫 𝐇𝐚𝐯𝐞 𝐁𝐨𝐭𝐡?   Healthcare generates 30% of the world’s data, but breaches hit the sector hardest. How do we balance seamless data sharing for care coordination with ironclad security? With AI mining patient data for insights, are current regulations (GDPR, HIPAA) enough—or are we gambling with trust?   2️⃣ 𝐓𝐡𝐞 𝐈𝐧𝐭𝐞𝐫𝐨𝐩𝐞𝐫𝐚𝐛𝐢𝐥𝐢𝐭𝐲 𝐌𝐢𝐫𝐚𝐠𝐞: 𝐖𝐡𝐲 𝐀𝐫𝐞 𝐖𝐞 𝐒𝐭𝐢𝐥𝐥 𝐒𝐢𝐥𝐨𝐬?  Decades into the digital health revolution, systems *still* don’t talk to each other. EHRs, wearables, and telehealth platforms operate in parallel universes. Is this a technical failure—or a refusal to prioritize patient outcomes over proprietary gains?  3️⃣ 𝐇𝐞𝐚𝐥𝐭𝐡 𝐄𝐪𝐮𝐢𝐭𝐲 𝐨𝐫 𝐃𝐢𝐠𝐢𝐭𝐚𝐥 𝐄𝐥𝐢𝐭𝐢𝐬𝐦?   Digital health promises democratization, but 40% of low-income patients lack broadband for telehealth. Are we building tools for the privileged—or designing for the 85-year-old with a flip phone? How do we close the gap before it becomes a chasm?  4️⃣ 𝐓𝐡𝐞 𝐈𝐧𝐧𝐨𝐯𝐚𝐭𝐢𝐨𝐧 𝐯𝐬. 𝐑𝐞𝐠𝐮𝐥𝐚𝐭𝐢𝐨𝐧 𝐓𝐮𝐠-𝐨𝐟-𝐖𝐚𝐫  AI diagnostics can save lives, but FDA approvals take years. Europe’s new AI Act adds more guardrails. Are regulators stifling progress, or is “move fast and break things” a disaster in healthcare? Where’s the line between safety and stagnation?  5️⃣ 5. 𝐓𝐡𝐞 𝐄𝐧𝐠𝐚𝐠𝐞𝐦𝐞𝐧𝐭 𝐏𝐚𝐫𝐚𝐝𝐨𝐱: 𝐓𝐞𝐜𝐡 𝐂𝐚𝐧’𝐭 𝐅𝐢𝐱 𝐇𝐮𝐦𝐚𝐧 𝐁𝐞𝐡𝐚𝐯𝐢𝐨𝐫  90% of health apps are abandoned within months. Patients ignore medication reminders; clinicians resent algorithm-driven workflows. Are we overestimating tech’s power to change habits—and underestimating the human element?  Your turn: Which challenge keeps you up at night? #digitalhealth #AIRegulation #MedTech #Innovation #ThoughtLeaders Let’s debate—drop your take below.  👇 👇

𝗡𝗮𝘃𝗶𝗴𝗮𝘁𝗶𝗻𝗴 𝘁𝗵𝗲 𝗪𝗮𝘁𝗲𝗿𝘀 𝗼𝗳 𝗠𝗲𝗱𝗶𝗰𝗮𝗹 𝗗𝗲𝘃𝗶𝗰𝗲 𝗖𝗹𝗮𝘀𝘀𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻 🚢 Ever found yourself pondering over the classification of a medical device? Perhaps while holding a seemingly simple product, like eye drops and wondering about its regulatory standing. You're not alone. Understanding the classification of medical devices is crucial but it can often seem like deciphering hieroglyphs. Here's a brief guide to help demystify the process: In many jurisdictions, including those regulated by the FDA (U.S. Food and Drug Administration) or under the MDR (Medical Device Regulation) in the EU, medical devices are categorized based on their risk level to patients and users. These range from Class I (low risk) to Class III (high risk). Factors influencing classification include invasiveness, duration of contact with the body and whether they are active or non-active devices etc. A place to start is with a strong intended use statement, many I review are often Tenuous and not very clear Now, regarding eye drops: They could be classified differently based on their intended use and mechanism of action. For example, if they provide lubrication only with no therapeutic claims, they might be considered lower risk compared to those claiming to treat or prevent eye conditions. Understanding these classifications isn't just bureaucratic compliance; it's about ensuring safety and efficacy in products that impact health outcomes.

Post-market surveillance doesn’t start – or stop – with the PSUR. It’s a living ecosystem that runs continuously, far beyond a periodic report. Throughout the lifecycle of a medical device, data should be actively and systematically collected. Not just when the PSUR is due. This includes sales data, user profiles, clinical experience, complaints, literature, trend reports, usability surveys, and more... All this data feeds into the PSUR (mandatory for class IIa devices and above). But the goal of PMS isn’t just to write a report. It’s to take action based on the data as it becomes available. That’s what makes PMS proactive. Waiting for a reporting deadline before reassessing your risks is reactive. Instead, risks should be continuously reviewed, signals identified early, and actions taken in real time. This includes vigilance activities. When an unacceptable increase in risk is identified, a FSCA must be issued. This is followed by a Field Safety Notice (FSN) to inform users, often including updated instructions. And if an incident occurs, a Manufacturer Incident Report (MIR) must be submitted to the competent authority – following strict timelines: → Serious public health threat: within 2 days → Death or unanticipated serious deterioration: within 10 days → Other serious incidents: within 15 days None of these are optional. They are key tools within a truly operational PMS process. Once data is gathered, it must lead to analysis and concrete actions. → Were new risks identified? → Could known risks be re-evaluated? → Do trends reveal emerging issues? If the answer is yes, manufacturers must conduct a root cause analysis, update the risk assessment, implement or adapt risk controls, and update relevant documentation. That includes the technical documentation, IFU, clinical evaluation, risk management file, and more. Because in the end, PMS is not a document. It’s a decision-making process based on real-time evidence. Need more about Post-Market Surveillance ? Grab our PMS templates (plan + PSUR), take advantage of : → Pre-written paragraphs and tables → Collection and analysis methods → Clear links to MDR 2017/745 and ISO 20416 → Concrete examples with a medical device Save time and be efficient in your PMS → https://lnkd.in/ezMXxxMc

WHOOP’s FDA Warning Letter Now Anchors a Class-Action Lawsuit The latest update in the WHOOP case underscores how quickly regulatory risk can escalate into legal and commercial exposure. ArentFox Schiff's recent analysis breaks it down well: 🔗 https://lnkd.in/gWvHbWEp FDA’s original warning letter focused on WHOOP’s Blood Pressure Insights feature — marketed as wellness but functioning as a diagnostic device without appropriate clearance. Now, that same letter is being used as the foundation of a class-action lawsuit alleging misleading claims and regulatory noncompliance. Key lessons for the wearable and digital-health sector: - “Wellness” is no longer a safe harbor when the product generates physiological metrics that look, feel, or behave like clinical measurements. Derived or algorithmic outputs matter just as much as raw sensors when FDA evaluates intended use. - A single regulatory misstep can compound rapidly into litigation, reputational harm, and market-access barriers. - AI-enabled insights amplify the risk, especially when consumers may interpret them as diagnostic or actionable medical guidance. As more consumer devices move into areas like cardiac risk, blood pressure, sleep apnea, and glucose approximation, the line between wellness and regulated medical devices is tightening fast. This is a critical moment for wearable-tech innovators to reassess: ✔ Intended-use statements ✔ Data-labeling and claims ✔ Algorithm transparency ✔ Sales and marketing practices Clear regulatory strategy early on isn’t optional anymore — it’s a safeguard against costly downstream consequences. #FDA #Wearables #DigitalHealth #RegulatoryAffairs #AIinHealthcare #MedTechCompliance #SaMD

This article from July, 15 reports on a closed-door workshop organized by the Stanford Institute for Human-Centered Artificial Intelligence (HAI) in May 2024, where 55 leading policymakers, academics, healthcare providers, AI developers, and patient advocates gathered to discuss the future of healthcare AI policy. The main focus of the workshop was on identifying gaps in current regulatory frameworks and fostering support for necessary changes to govern AI in healthcare effectively. Key Points Discussed: 1.) AI Potential and Investment: AI has the potential to revolutionize healthcare by improving diagnostic accuracy, streamlining administrative processes, and increasing patient engagement. From 2017-2021, the healthcare sector saw significant private AI investment, totaling $28.9 billion. 2.) Regulatory Challenges: Existing regulatory frameworks, like the FDA's 510(k) device clearance process and HIPAA, are outdated and were not designed for modern AI technologies. These regulations struggle to keep up with the rapid advancements in AI and the unique challenges posed by AI applications. 3.) The workshop focused on 3 main areas: - AI software for clinical decision support. - Healthcare enterprise AI tools. - Patient-facing AI applications. 4.) Need for New Frameworks: There was consensus among participants that new or substantially revised regulatory frameworks are essential to effectively govern AI in healthcare. Current regulations are like driving a 1976 Chevy Impala on modern roads, and are inadequate for today's technological landscape. The article emphasizes the urgent need for updated governance structures to ensure the safe, fair, and effective use of AI in healthcare. The article describes the 3 use cases discussed: Use Case 1: AI in Software as a Medical Device - AI-powered medical devices face challenges with the FDA's clearance, hindering innovation. - Workshop participants suggested public-private partnerships for managing evidence and more detailed risk categories for different AI devices. Use Case 2: AI in Enterprise Clinical Operations and Administration - Balancing human oversight with autonomous AI efficiency in clinical settings is challenging. - There is need for transparent AI tool information for providers, and a hybrid oversight model. Use Case 3: Patient-Facing AI Applications - Patient-facing AI applications lack clear regulations, risking the dissemination of misleading medical information. - Involving patients in AI development and regulation is needed to ensure trust and address health disparities. Link to the article: https://lnkd.in/gDng9Edy by Caroline Meinhardt, Alaa Youssef, Rory Thompson, Daniel Zhang, Rohini Kosoglu, Kavita Patel, Curtis Langlotz

🚨 Your “wellness app” might actually be a medical device 🚨 Take this example 👇 An app displays lab test results (such as vitamin levels or biomarkers) and then suggests supplements or lifestyle changes to restore those values to normal. The founders say: “We’re not treating or diagnosing, we’re just guiding. Sharing what other users have done, or what the latest science says.” But regulators in the EU and the US may not see it that way. 🇪🇺 EU (MDR) Under Rule 11 of MDR, software that monitors, prevents, or alleviates disease is considered a medical device. If your app interprets diagnostic data and recommends actions, it’s likely MedTech, possibly Class IIa or higher. 🇺🇸 US (FDA) FDA takes a risk-based approach. General wellness apps are fine if they stay generic: “eat healthy, exercise.” However, once recommendations are based on personal diagnostic results, the FDA may consider it a medical device. 👉 The lesson: you can’t “opt out” of regulation by calling your product lifestyle. If your app is designed to influence medical decisions, regulators will likely classify it as a medical device. ⚠️ Why this matters: Regulatory classification determines your evidence requirements, liability, and pathway to market. Misclassifying your product early risks years of rework and delays. Intended use defines reality, not what you call it. Luckily, it doesn't have to be that costly or troublesome to get through regulatory approval. You just have to know who to talk to... #MedTech #SaMD #DigitalHealth #MedicalDevices #HealthTech