Let's Talk Risk! with Dr. Naveen Agarwal

Imagine you are running a pivotal clinical trial for a novel implant. The data comes back, and it is rough: 80% of your patients have suffered a serious adverse event, and 40% have developed acute kidney injury.If you are sitting in the regulatory or risk management seat, you are likely drafting the project’s post-mortem. In a traditional risk management paradigm, you are preparing to tell the executive team that the device failed to meet any traditional safety threshold.But what if the FDA didn’t just approve this device, but approved it specifically because the sponsors mathematically proved that patients were willing to tolerate a higher level or risk to gain access to this device?This scenario completely dismantles the way the MedTech industry has historically viewed safety and effectiveness. As professionals, we are trained to treat clinical thresholds as objective, immutable laws of physics—a line in the sand where an adverse event rate either passes or fails. However, with the FDA’s finalized guidance issued on March 30, 2026, safety is no longer just a raw numerical threshold; it is now a quantifiable variable relative to the validated preference of the end user.So, how does a manufacturer mathematically prove that a severe safety profile is acceptable, and how does the FDA reconcile approving it?🎧Click Play above to listen to a brief audio summary about this case and lessons QA/RA and Clinical professionals can apply in practice using the newly released FDA Guidance.In this episode, we discuss:* The fundamental difference between Patient Reported Outcomes (PROs) and Patient Preference Information (PPI)—and why conflating the two leads to flawed regulatory submissions.* The exact mechanics of how a rigorously designed Discrete Choice Experiment (DCE) rescued the alfapump system from regulatory rejection.* How to utilize the Q-submission program to negotiate mathematical models with the FDA before collecting a single data point.* Strategic traps to avoid, including the “subpopulation matching problem” that can engineer a massive off-label use issue for your pipeline.Thanks for reading Let's Talk Risk!. If you liked this post, share with others.Note:The audio summary was prepared using Google NotebookLM, an AI-enabled research tool. Here are a few key resources used for this analysis:* FDA Guidance: Patient Preference Information - Voluntary Submission, Review in Premarket Approval Applications, Humanitarian Device Exemption Applications, and De Novo Requests, and Inclusion in Decision Summaries and Device Labeling (Issued March 30, 2026).* P230044, Sequana Medical N.V., alfapump® System, Approved December 2024. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit naveenagarwalphd.substack.com/subscribe

Summary“AI brings novel risks and addressing them will require a more creative approach to human factors engineering.”In this Let’s Talk Risk! conversation, host Naveen Agarwal speaks with Jonathan Kendler, a seasoned medical device human factors and UI design consultant, about how artificial intelligence is changing the practice of usability engineering in MedTech. Jonathan brings a rare perspective to the discussion, blending decades of experience in user interface design, human factors, and medical device development.The conversation explores a central tension facing the industry today: while AI introduces new interaction patterns, trust challenges, and potential safety risks, the core responsibility remains unchanged - understand the user, understand the context of use, and design for safe, effective human interaction. Jonathan also shares where he sees real promise, including AI-supported training, surgical planning, and stronger cross-functional collaboration across risk, quality, regulatory, engineering, and human factors teams.Listen to the full 30-minute podcast or jump to a section of interest listed below. Chapters00:00 Introduction02:53 Why AI changes the environment, but not the fundamentals of human factors05:46 Transparency, over-reliance, and helping clinicians stay vigilant09:36 AI-strengthened training and the shift from training to competence13:19 Promising AI use cases, especially in surgical planning and clinician support15:38 The future of human factors: creativity, collaboration, and new evaluation methods23:30 Jonathan’s upcoming workshop on advanced human factors and use-related risk26:45 Final takeaways: stay grounded in the user, fundamentals, and teamworkIf you enjoyed this podcast, consider subscribing to the Let’s Talk Risk! newsletter.Suggested links:LTR: Start Treating Human Factors as Science, Not Art.LTR: Understanding Human Factors for AR/VR Applications in MedTech.LTR: LTR Risk Coach - AI-Powered Decision Support Tool.Key Takeaways* AI-enabled devices do not eliminate the need for strong human factors fundamentals; they make them more important.* One of the biggest emerging risks is lack of transparency around what the AI is doing and how users should interpret its output.* Accuracy alone is not enough; teams must design for the reality that AI can fail and users must remain appropriately vigilant.* AI may create meaningful opportunities to improve training, making it more interactive, adaptive, and competence-focused.* Some of the most promising applications of AI are those that strengthen clinician judgment rather than replace it.* Human factors professionals will need more creativity to anticipate novel use errors, misuse, and complex real-world interactions.* Conventional usability testing may not be enough for AI-enabled systems, especially when risks like habituation emerge over time.* The best outcomes will come from closer collaboration across human factors, engineering, risk management, quality, regulatory, and safety teams.KeywordsHuman factors, usability engineering, AI in medical devices, machine learning, transparency, clinician vigilance, use-related risk, AI training, competence, surgical planning, usability testing, cross-functional collaborationAbout Jonathan KendlerJonathan Kendler is a medical device human factors and UI design consultant with more than 20 years of experience supporting the development of healthcare products, including dialysis machines, infusion pumps, glucose monitors, and patient monitoring systems. He is Principal Consultant at Curiolis and previously co-founded Wiklund Research & Design, which was later acquired by UL.Jonathan has authored numerous articles on usability engineering and co-authored Usability Testing of Medical Devices and Designing for Safe Use. He has also taught graduate-level courses in user interface design at Tufts University and delivered workshops and lectures on usability engineering across North America, Europe, and Asia.Let’s Talk Risk! with Dr. Naveen Agarwal is a bi-weekly live audio event on LinkedIn, where we talk about risk management related topics in a casual, informal way. Join us at 11:00 am EST every other Friday on LinkedIn.DisclaimerInformation and insights presented in this podcast are for educational purposes only, and not as legal advice. Views expressed by all speakers are their own and do not reflect those of their respective organizations.Parts of this article were created using AI-generated content, which was subsequently reviewed, edited, and fact-checked by the author to ensure accuracy and alignment with our standards. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit naveenagarwalphd.substack.com/subscribe