?? "Curious about the claims your Class I software can make under MDR?" ? Let’s explore what medical claims might fit into Class I according to Rule 11 of Annex VIII MDR. A full analysis takes time, but this guide will offer clarity on potential pathways, strategies, and approaches for companies and regulatory professionals. ? Here are some exclusion criteria: if any of these apply to your claim, achieving Class I status is unlikely. Let’s examine the criteria: ? ?? Diagnostic or Therapeutic Purposes: If your software’s purpose involves direct diagnosis or therapy (e.g., image analysis for stroke detection), it’s unlikely to be Class I. Check MDCG 2022-5 section 1.2.3 for the definition of diagnosis. ? ?? Supporting Diagnostic/Therapeutic Processes: If your software aids these processes by providing crucial information (e.g., treatment recommendations for heart attack patients), it doesn’t qualify as Class I. ? ?? Monitoring Vital Signs: If your software monitors vital physiological processes (e.g., brain function, blood pressure, body temperature), it will not be classified as Class I. ? If these don’t apply, consider these factors for final classification: ????????????????????????????????? ?? Rule 11 Insights: Software for monitoring physiological processes is Class IIa. There’s debate about what constitutes a physiological process. For instance, is a pathological tremor a physiological process or its effect? ? ?? Prevention, Prediction, Measurement: While diagnosis, therapy, and physiological monitoring rule out Class I, activities like prevention (e.g., preventing chronic diseases), prediction (e.g., estimating disease risk), or anatomical/physiological measurement might still fit. ? ?? Conception Support: An easier category is software supporting conception, as explicitly stated in MDCG 2021-24 as Class I. ? ? Clinical Benefit: If your Class I device lacks any clinical benefit, passing a conformity assessment will be challenging. ? What examples of Class I software do you have? ? #medicaldevices #mdr
What’s your immediate thoughts on Apple Watch and their vital sign monitoring (i.e. respiration). Neither the watch itself is considered MD nor is their sleep analysis based on monitoring respirations considered MD. But if it was categorised as MD then instant class IIa (minimum)?
Great insights Tibor Zechmeister. What strategies do you recommend for companies when they’re on the borderline between Class I and Class IIa classifications?
Tibor Zechmeister Imagine you're a Young and dynamic Team, developing a cool Software...and suddenly it's a medical device :D
This is going to be fun for the wearables market too.
"Class I, activities like prevention (e.g., preventing chronic diseases), prediction (e.g., estimating disease risk), or anatomical/physiological measurement might still fit." -> Last is a Class Im and required a Notified Body.
@Clinical Benefit: If a direct clinical benefit for that patient is missing, their might well be an indirect clinical benefit. We have that case for a clinical research device, for which benefit comes via physiological knowledge gain.
Tibor Zechmeister thanks to explain clearly the situation of software as medical device in eureupean union. This is the same for software as in vitro diagnostics medical device. A lot of start up shall read your post because they always think ??we can start with a class 1 or A device and after we move to an higher class of risk??, this is not a relevant regulatory strategy in my mind.
Great breakdown of the criteria for Class I software under MDR Tibor Zechmeister! It's a complex area, but this guide offers much-needed clarity. How do you see the debate around what constitutes a 'physiological process' evolving, especially with advancements in AI-driven health tech?
I am missing the criteria where the software is intended to provide information which is NOT used for taking decisions with diagnosis or therapeutic purposes. Unfortunately MDR does not define 'therapeutic purpose'. How would you define it? Do you have other Class I software examples?
CEO & Co-Founder at FormlyAI | Medtech certification for startups made easy | ex-FDA
1 个月I hate that we don't have a solid definition of "monitoring". Obviously having a self-made diary log of your health information in your app and reading that data directly from a wearable are different. When does it cross the line into monitoring? I feel you could argue either way.