Medical Device Regulations

BREAKING! The FDA just released this draft guidance, titled Artificial Intelligence-Enabled Device Software Functions: Lifecycle Management and Marketing Submission Recommendations, that aims to provide industry and FDA staff with a Total Product Life Cycle (TPLC) approach for developing, validating, and maintaining AI-enabled medical devices. The guidance is important even in its draft stage in providing more detailed, AI-specific instructions on what regulators expect in marketing submissions; and how developers can control AI bias. What’s new in it? 1) It requests clear explanations of how and why AI is used within the device. 2) It requires sponsors to provide adequate instructions, warnings, and limitations so that users understand the model’s outputs and scope (e.g., whether further tests or clinical judgment are needed). 3) Encourages sponsors to follow standard risk-management procedures; and stresses that misunderstanding or incorrect interpretation of the AI’s output is a major risk factor. 4) Recommends analyzing performance across subgroups to detect potential AI bias (e.g., different performance in underrepresented demographics). 5) Recommends robust testing (e.g., sensitivity, specificity, AUC, PPV/NPV) on datasets that match the intended clinical conditions. 6) Recognizes that AI performance may drift (e.g., as clinical practice changes), therefore sponsors are advised to maintain ongoing monitoring, identify performance deterioration, and enact timely mitigations. 7) Discusses AI-specific security threats (e.g., data poisoning, model inversion/stealing, adversarial inputs) and encourages sponsors to adopt threat modeling and testing (fuzz testing, penetration testing). 8) And proposed for public-facing FDA summaries (e.g., 510(k) Summaries, De Novo decision summaries) to foster user trust and better understanding of the model’s capabilities and limits.

🔍 𝐀 𝐂𝐥𝐢𝐧𝐢𝐜𝐢𝐚𝐧'𝐬 𝐆𝐮𝐢𝐝𝐞 𝐭𝐨 𝐄𝐯𝐚𝐥𝐮𝐚𝐭𝐢𝐧𝐠 𝐃𝐢𝐠𝐢𝐭𝐚𝐥 𝐇𝐞𝐚𝐥𝐭𝐡 𝐓𝐨𝐨𝐥𝐬 As a GP immersed in the digital health world, I'm often asked by fellow clinicians: "How do we know if a digital health tool is actually any good?" 📋 Here's your practical quality checklist on what to look out for, questions to ask and nuances that can make a big difference! 🔏 𝐑𝐞𝐠𝐮𝐥𝐚𝐭𝐨𝐫𝐲 𝐒𝐭𝐚𝐭𝐮𝐬 & 𝐂𝐞𝐫𝐭𝐢𝐟𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬 ✅ Is it EU MDR, UKCA or FDA certified? (e.g. Ada Assess, our flagship product is MDR Class IIa certified) ✅ Know your classifications! (e.g. Class I medical devices in the EU have the lowest perceived risk. In many cases, the manufacturer can self-certify Class I devices without the involvement of a notified body. This risk class includes products like stethoscopes, bandages, or glasses!) ✅ ISO 13485 certification (shows commitment to quality management) ✅ Data security certifications - ISO 27001, HIPAA compliance, GDPR (show they take data privacy seriously) 🔗 See The Strategic Value of EU MDR & ISO 13485 in Digital Health: https://lnkd.in/eNit_dWa 🥼 𝐄𝐯𝐢𝐝𝐞𝐧𝐜𝐞 𝐁𝐚𝐬𝐞 ✅ Publicly available clinical studies (peer-reviewed is gold standard) ✅ Real-world evidence and performance data ✅ Transparent reporting of limitations and uncertainties ✅ Independent validation studies (not just internal testing) 🔗 See Ada Health's Clinical Studies: https://lnkd.in/ewswDfcV 🏥 𝐏𝐨𝐬𝐭-𝐌𝐚𝐫𝐤𝐞𝐭 𝐒𝐮𝐫𝐯𝐞𝐢𝐥𝐥𝐚𝐧𝐜𝐞 ✅ Active safety monitoring systems ✅ Regular performance updates and transparency reports ✅ Clear processes for handling adverse events ✅ Continuous performance monitoring, especially for AI-based tools 👔 𝐂𝐨𝐦𝐩𝐚𝐧𝐲 𝐀𝐜𝐜𝐞𝐬𝐬𝐢𝐛𝐢𝐥𝐢𝐭𝐲 & 𝐒𝐮𝐩𝐩𝐨𝐫𝐭 ✅ Dedicated medical safety team ✅ Clear channels for clinical feedback ✅ Responsive to safety concerns ✅ Regular updates and communication about changes 🔗 See Prioritising Patient Safety in Digital Health: https://lnkd.in/enUJCghz) 🤝 𝐈𝐦𝐩𝐥𝐞𝐦𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧 𝐒𝐮𝐩𝐩𝐨𝐫𝐭 ✅ Were you involved in designing/evaluating the tool? ✅ Have you been provided with clinical workflow integration guidance? ✅ Training materials and support? ✅ Clear documentation of intended use (and when not to use the product!) 🎯 𝐏𝐫𝐨 𝐓𝐢𝐩: Engage with the company's medical team. A quality digital health partner should welcome clinical dialogue and demonstrate a genuine commitment to patient safety and co-designing products. 𝐑𝐞𝐦𝐞𝐦𝐛𝐞𝐫: The best digital health tools aren't just technically sound - they're built on a foundation of clinical understanding and regulatory rigour. They should enhance, not complicate, your clinical practice. 🤔 What quality aspects do you look for when evaluating digital health tools? Share your experiences below!

Clinical evaluation is the most underestimated challenge in your MDR compliance journey. Most manufacturers focus heavily on QMS documentation and technical files. But it's your clinical evaluation that often becomes the bottleneck. Why? Because it requires both scientific rigor and regulatory precision. And notified bodies are scrutinizing this area more than ever before. The stakes are clear: insufficient clinical data means delayed market access or even rejection. So what does a compliant clinical evaluation actually look like? Here are 5 essential elements every MedTech leader needs to master: Clinical Evaluation Plan (CEP) ↳ This isn't just a document. It's your roadmap for success. ↳ Define specific endpoints that align with your intended purpose. ↳ Remember that vague objectives lead to undefined outcomes. Literature Review Strategy ↳ Simply collecting studies isn't enough anymore. ↳ You need a systematic search methodology with clear inclusion/exclusion criteria. ↳ Document why certain studies were rejected— auditors always ask this. Clinical Data Sufficiency ↳ "Sufficient" clinical data is subjective until you define it. ↳ Create a clear threshold for what constitutes adequate evidence. ↳ Pre-MDR data often falls short of current expectations. Post-Market Clinical Follow-Up (PMCF) ↳ This isn't optional. It's a fundamental part of your clinical evaluation. ↳ Notified bodies expect proactive data collection, not just passive surveillance. ↳ The days of "we have no complaints" as sufficient PMCF are long gone. Equivalence Justification ↳ The bar for equivalence has been raised significantly under MDR. ↳ You need access to technical documentation of equivalent devices. ↳ Without contractual agreements, equivalence claims are increasingly difficult to defend. Clinical evaluation isn't a one-time task. It's a continuous process throughout your device's lifecycle. The manufacturers who succeed are those who integrate clinical thinking from design phase through post-market surveillance. P.S. What's been your biggest challenge with clinical evaluations under MDR? Is it finding sufficient data, justifying equivalence, or something else? ⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡ MedTech regulatory challenges can be complex. But smart strategies, cutting-edge tools, and expert insights can make all the difference. I'm Tibor, passionate about leveraging AI to transform how regulatory processes are automated and managed. Let's connect and collaborate to streamline regulatory work for everyone! #clinicalevaluation #regulatoryaffairs #medicaldevices

🤩 A milestone for #MedTech: Digital technical documentation will enter EU legislation This week, the European Commission published a proposal to simplify the european laws MDR and IVDR for #medicalDevices and in vitro diagnostic devices. 🌟 For the first time ever, the draft includes explicit provisions on the digitalisation of technical documentation and conformity assessment (new MDR Article 52b / IVDR Article 48b). 👉 This is a true novum for the medical device industry and a strong signal that that the #digitalisation of regulatory processes and related #technicalDocumentation plays an important role for the industry. --- 😁 Personally, this proposal makes me genuinely happy: I have been working on solutions for these topics for many years now. Seeing these concepts reflected in a concrete legislative proposal is highly motivating and reinforces my belief that this work truly matters. Not only am I proud of my work at the avasis solutions GmbH (avasis Group), but above all I am proud of our collaborative work in the non-profit Medical Device Knowledge Units (MDKU) e.V. association: More than five years ago, we began to put into practice the idea of a unified data model for technical documentation of medical devices. We will soon publish DIN SPEC 91509 with a first release. Our goal has always been to enable structured, interoperable and reusable technical documentation - digital by design, not a collection of static PDFs. Because that is "real digitalisation" and the foundation for a useful application of AI. ❇️ From documents to data! ❇️ --- Seeing this principle now reflected in a legislative proposal confirms that the direction was right - and that collaborative, pre-competitive work can help prepare the ground for future regulation 🥳 I am more motivated than ever to continue contributing to this transformation and to help ensure that these new legal provisions can be translated into practical, scalable and industry-ready solutions. --- 👉 How do you see the future of digital technical documentation under MDR & IVDR? I’d love to hear perspectives from manufacturers, notified bodies and regulators!

FDA just posted this draft guidance updating Quality Management System requirement for Premarket Authorization (PMA) submissions. Most of the document lays out how the requirements from each clause in ISO 13485 and the revised 21 CFR part 820 should be communicated in the submission package. The thing everyone really wants to know, however, is how the FDA will handle submissions delivered prior to 2 February 2026, where the review period crosses the effective date. Here is your answer in section IV of the draft guidance: "On and after February 2, 2026, FDA will be evaluating the documents and records included in marketing submissions to determine whether there is conformance with the requirements of the QMSR. A gap analysis or another type of comparative analysis may assist FDA in determining when documents and records created prior to the QMSR effective date are submitted to FDA. Additionally, on and after February 2, 2026, FDA inspections of device manufacturers that are evaluating CGMP, including PMA preapproval inspections, will evaluate compliance with QMSR requirements. In doing so, it may help FDA to make that determination by providing a gap analysis or a comparative assessment. " I've attached the full draft guidance for your reading pleasure.

𝗛𝗲𝗿𝗲'𝘀 𝗺𝘆 𝟳-𝘀𝘁𝗲𝗽 𝗽𝗹𝗮𝘆𝗯𝗼𝗼𝗸 𝗳𝗼𝗿 𝗲𝗻𝘀𝘂𝗿𝗶𝗻𝗴 𝘀𝗺𝗼𝗼𝘁𝗵 𝗿𝗲𝗴𝘂𝗹𝗮𝘁𝗼𝗿𝘆 𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻𝘀 𝘁𝗵𝗮𝘁 𝗜'𝘃𝗲 𝗿𝗲𝗳𝗶𝗻𝗲𝗱 𝗼𝘃𝗲𝗿 𝘆𝗲𝗮𝗿𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗠𝗲𝗱𝗧𝗲𝗰𝗵 𝗾𝘂𝗮𝗹𝗶𝘁𝘆 𝗮𝗻𝗱 𝗿𝗲𝗴𝘂𝗹𝗮𝘁𝗼𝗿𝘆 𝘀𝗽𝗮𝗰𝗲: 𝟭. 𝗦𝘁𝗮𝗿𝘁 𝘄𝗶𝘁𝗵 𝘁𝗵𝗲 𝗲𝗻𝗱 𝗶𝗻 𝗺𝗶𝗻𝗱 - 𝟭𝟴-𝟮𝟰 𝗺𝗼𝗻𝘁𝗵𝘀 𝗯𝗲𝗳𝗼𝗿𝗲 𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻 • Map your regulatory strategy to your commercial goals • Identify your regulatory pathway early (510(k), De Novo, PMA) • Build testing protocols based on predicate devices when applicable 𝟮. 𝗗𝗲𝘀𝗶𝗴𝗻 𝘆𝗼𝘂𝗿 𝗤𝘂𝗮𝗹𝗶𝘁𝘆 𝗠𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁 𝗦𝘆𝘀𝘁𝗲𝗺 𝗳𝗼𝗿 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 • Implement ISO 13485 principles from day one • Focus on the 7 critical SOPs that impact submissions most • Avoid the common trap of documentation overload (I've seen startups with 200+ SOPs when 35-40 would suffice) 𝟯. 𝗩𝗮𝗹𝗶𝗱𝗮𝘁𝗲 𝘆𝗼𝘂𝗿 𝘁𝗲𝘀𝘁𝗶𝗻𝗴 𝗺𝗲𝘁𝗵𝗼𝗱𝗼𝗹𝗼𝗴𝘆 𝗯𝗲𝗳𝗼𝗿𝗲 𝗲𝘅𝗲𝗰𝘂𝘁𝗶𝗻𝗴 • Pre-validate test methods with 3-5 pilot runs • Engage with testing labs that have FDA submission experience • Document protocol deviations properly (we found 63% of submissions get delayed due to inadequate deviation management) 𝟰. 𝗟𝗲𝘃𝗲𝗿𝗮𝗴𝗲 𝗽𝗿𝗲-𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻 𝗺𝗲𝗲𝘁𝗶𝗻𝗴𝘀 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗰𝗮𝗹𝗹𝘆 • Schedule Q-Sub meetings 9-12 months before planned submission • Prepare focused questions (limit to a few critical issues) • Follow up with written summaries within the allocated time 𝟱. 𝗕𝘂𝗶𝗹𝗱 𝗮 𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻 "𝘄𝗮𝗿 𝗿𝗼𝗼𝗺" • Assemble cross-functional team (R&D, Clinical, Quality, Regulatory) • Create submission trackers with accountability metrics • Hold twice-weekly stand-ups in the 90 days before submission 𝟲. 𝗖𝗼𝗻𝗱𝘂𝗰𝘁 𝘁𝗵𝗶𝗿𝗱-𝗽𝗮𝗿𝘁𝘆 𝘀𝘂𝗯𝗺𝗶𝘀𝘀𝗶𝗼𝗻 𝗿𝗲𝘃𝗶𝗲𝘄 • Have external experts review 100% of your technical documentation • Use submission management platforms like RADAR or MasterControl • Schedule review 45-60 days before planned submission date 𝟳. 𝗣𝗿𝗲𝗽𝗮𝗿𝗲 𝗳𝗼𝗿 𝗶𝗻𝘁𝗲𝗿𝗮𝗰𝘁𝗶𝘃𝗲 𝗿𝗲𝘃𝗶𝗲𝘄 • Anticipate FDA questions with "pre-mortem" analysis • Have subject matter experts on standby during review period • Create response templates for common deficiency categories I learned these lessons the hard way. Early in my career I worked at a company where we had three submissions rejected due to inconsistent test data formatting. Now we use standardized data presentation templates that have cut our Additional Information requests by 72%. 𝗧𝗔𝗞𝗘𝗔𝗪𝗔𝗬: Regulatory success is about methodical preparation and strategic execution. The companies that view regulatory as a strategic function rather than a compliance burden consistently outperform their peers in time-to-market by an average of 7 months If you're preparing for an FDA submission in the next 12-18 months, I'd be happy to share our pre-submission checklist. Just message me directly

Big AI news for the MedTech sector! The European Commission released its new health package yesterday, which includes a significant amendment to how the AI Act would apply to medical technologies. Notably, the proposal would shift the Medical Devices Regulation (MDR) and the In Vitro Diagnostic Regulation (IVDR) from Annex I, Section A to Section B. While this may look like a technical adjustment, the implications could be substantial. Industry stakeholders have long highlighted friction between the AI Act and existing MedTech regulatory frameworks, and this change could reshape how those tensions are managed. Under Section A, products must meet the AI Act’s high-risk obligations immediately in parallel with MDR/IVDR requirements. By contrast, placing these rules under Section B means compliance would be channelled primarily through the sector-specific regime, with the AI Act applying as clarified through secondary legislation.

As the #healthcare industry continues to explore the transformative potential of large language models (#LLMs), one area that remains critical yet underleveraged is the role of #standardized #ontologies such as SNOMED International, LOINC, and #RxNorm. While #LLM excel at parsing unstructured clinical narratives, they often generate outputs with high variability—making #interoperability and reproducibility a challenge. That’s where standardized medical ontologies come in. By applying these coding systems as a normalization layer on top of LLM-generated outputs, we can enhance semantic consistency, data reliability, and #EHR integration. These ontologies can help bridge the gap between free text and structured data—unlocking the full potential of LLMs in clinical decision support, population health, and quality reporting. Of course, these ontologies are not without limitations—but their foundational role in standardizing terminology and reducing downstream ambiguity cannot be overstated. SNOMED CT and other standards offer a roadmap toward safer, more interoperable #AI in healthcare. #Healthinformatics #ClinicalInformatics #dataanalytics #Data

🔥Robotics in Healthcare. The Real Constraint Is Not Technology.🔥 What the Robot Report 2026 tells us, indirectly: 🔹Healthcare robotics is ready. 🔹AI vision works. 🔹Autonomous navigation works. 🔹Robots can already deliver, disinfect, transport, and assist. Yet scaled deployment inside hospitals remains limited. Why? Because healthcare is not a warehouse. It is a regulated, human-dense, risk-intolerant environment. 👉 The constraint is no longer capability. It is permission. ⚛️ The Healthcare Problem (As Seen on the Ground) In hospitals, every autonomous action raises questions: 📍 Who authorises a robot to move through a ward 📍 How does it prioritise emergency traffic 📍 What happens near vulnerable patients 📍 Who is accountable when systems intersect with humans 🟣 Without clear orchestration: 📌 Robots become pilots, not infrastructure 📌 Risk officers say no 📌 Clinicians lose trust 📌 Scale stalls ⚛️ The Missing Layer in Healthcare Robotics Hospitals do not need more robots. They need a clinical-grade governance layer. A system that ensures robots: ☑️ Move only where clinically appropriate ☑️ Yield to human care workflows ☑️ Respect infection control zones ☑️ Adapt to real-time hospital priorities 💎This is the Ambient Permission Plane, applied to healthcare. 🔹 What a Permission Plane Means in a Hospital: In practical terms, it enables: 👉 Context-aware movement Robots behave differently in ICU corridors, public areas, and sterile zones. 👉 Clinical priority orchestration Emergency movements override routine logistics automatically. 👉 Risk and liability clarity Every action is logged, governed, and auditable. 👉 Trust by design Nurses and clinicians do not fight the system. They trust it. This turns robotics from a novelty into hospital infrastructure. 🔹 Why This Matters Now Three forces are converging: 1. Labour shortages are structural Hospitals cannot hire their way out. 2. Operational pressure is rising Throughput, safety, and cost control must improve simultaneously. 3. Regulatory scrutiny is increasing Autonomous systems must be explainable, accountable, and safe. 💎 Robotics adoption in healthcare will not be won by the smartest robot. It will be won by the safest orchestration layer. 💎 Healthcare-Focused Conclusion 💎 👉 The future hospital will not ask, “Can this robot move.” 👉 It will ask, “Should it move now, here, and under these conditions.” That is the difference between automation and autonomy. And that is where the next decade of healthcare robotics will be decided. QUIKBOT TECHNOLOGIES is now in the right space and position to orchestrate this through our Ambient Permission Plane with QuikSync.

⚠️ Post-Market Surveillance: Guidance That Can Actually Help! Building a compliant and effective PMS system today is not just about collecting data. It’s about knowing what kind of data matters, how to structure your activities, and how to demonstrate control proactively. There’s no single manual. And will depend on the different juristictions. But there are key guidance documents that many regulatory and clinical teams find extremely useful when setting up or improving PMS processes. After last week's post on Clinical Evaluation Guidance, I decided to create a quick roundup of PMS guidance and documents worth knowing 👇 📄 𝗠𝗗𝗖𝗚 𝟮𝟬𝟮𝟯-𝟯 𝗥𝗲𝘃.𝟮 Q&A on vigilance terms and PMS concepts under MDR/IVDR 📄 𝗠𝗗𝗖𝗚 𝟮𝟬𝟮𝟮-𝟮𝟭 Guidance on Periodic Safety Update Report (PSUR) according to MDR 📄 𝗠𝗗𝗖𝗚 𝟮𝟬𝟮𝟬-𝟳 PMCF Plan Template: outlines structure and elements expected in PMCF plans under MDR 📄 𝗠𝗗𝗖𝗚 𝟮𝟬𝟮𝟬-𝟴 PMCF Evaluation Report Template: guidance for presenting results of PMCF activities 🇬🇧 𝗠𝗛𝗥𝗔 𝟮𝟬𝟮𝟰 𝗡𝗼. 𝟭𝟯𝟲𝟴 Amendment introducing formal PMS requirements in Great Britain, aligning closely with MDR 🇬🇧 𝗠𝗛𝗥𝗔 𝗣𝗠𝗦 𝗚𝘂𝗶𝗱𝗮𝗻𝗰𝗲 Practical interpretation of PMS requirements in Great Britain 🇺🇸 𝗙𝗗𝗔 𝟮𝟭 𝗖𝗙𝗥 𝗣𝗮𝗿𝘁 𝟴𝟮𝟮 Outlines mandatory postmarket surveillance requirements in the U.S 🇺🇸 𝗙𝗗𝗔 𝟱𝟮𝟮 𝗚𝘂𝗶𝗱𝗮𝗻𝗰𝗲 Clarifies how to fulfill section 522 obligations for PMS in US 🌐 𝗜𝗠𝗗𝗥𝗙 𝗠𝗗𝗖𝗘 𝗪𝗚/𝗡𝟲𝟱𝗙𝗜𝗡𝗔𝗟:𝟮𝟬𝟮𝟭 Guidance on design, implementation and scope of PMCF studies 📘 𝗜𝗦𝗢/𝗧𝗥 𝟮𝟬𝟰𝟭𝟲 International standard for establishing a post-market surveillance system aligned with MDR expectations 🌍 𝗪𝗛𝗢 𝗣𝗠𝗦 𝗚𝘂𝗶𝗱𝗮𝗻𝗰𝗲 Global guidance on organizing post-market and market surveillance systems for medical devices and IVDs 🧭 𝗠𝗘𝗗𝗗𝗘𝗩 𝟮.𝟭𝟮/𝟭 𝗥𝗲𝘃. 𝟴 Considered obsolete under MDR, as per MDCG 2023-3 Rev.2, but still a helpful framework for vigilance and PMS concepts 📌 Using the right guidance doesn’t replace your PMS process but it helps teams align, reduce gaps, and prepare stronger documentation for audits and keeping compliance. Staying up to date with guidance evolution and authorities news is key to staying aligned with what authorities expect from your PMS system! 👉 Which document do you rely on most when building or updating your PMS system? #MedBoard #PMS #PostMarketSurveillance #MedTech #MDR #MDCG #ClinicalAffairs #RegulatoryAffairs #RegulatoryIntelligence #PSUR #Vigilance #IMDRF #ISO20416 #Compliance #PMSPlan #PMCF