A question I keep getting from manufacturers is whether they can put two, five, or twenty device variants in the same technical documentation file under EU MDR. The short answer is that the regulation does not give you a number. It gives you a logic. Once you understand that logic, the right answer for your own portfolio usually becomes obvious.

There is also a commercial layer that nobody likes talking about. Every technical documentation file you submit gets reviewed by your Notified Body as a separate package, and MDR review cycles are longer and more expensive than MDD ever was. So manufacturers have a real incentive to consolidate, and reviewers have a real incentive to push back when the consolidation does not hold up. Most of the friction in early MDR submissions sat in exactly that gap.

What the regulation actually says

Annex II and Annex III of Regulation (EU) 2017/745 describe the content of technical documentation. They tell you what has to be in the file. Neither of them sets a maximum number of devices.

The boundary you are actually working against sits in Annex VI Part C, which defines the Basic UDI-DI. The Basic UDI-DI is the primary grouping identifier in MDR, and it has three criteria: same intended purpose, same risk class, and same essential design and manufacturing characteristics. If two devices share all three, they sit under the same Basic UDI-DI and can usually share a technical documentation file. If even one of those three breaks, you are looking at a separate Basic UDI-DI and, in most cases, a separate file.

MDCG 2018-1 expands on how to assign Basic UDI-DIs in practice, and MDCG 2019-13 gives examples of acceptable groupings for CE marking documentation. Both are worth reading before you commit to a grouping strategy, because they reflect how Notified Bodies have been told to evaluate this.

When grouping usually works

Size and dimensional variants are the easiest case. Catheters of different lengths or diameters, implants of different sizes, instruments offered in different gauges. As long as the materials are the same, the manufacturing process is the same, the sterilization method is the same, and the clinical use is the same, these typically stay in one file. The clinical evaluation can cover the whole range, the risk file can list the variants as configurations, and the verification testing can be done on worst-case representatives with a documented rationale.

Minor design variations that do not change the intended purpose or the safety profile also tend to be acceptable. A reusable instrument with two handle styles, an electrode in two configurations of the same active surface, accessories that share a controller. The trick here is showing that the variation is actually minor. If you cannot point to a specific risk assessment confirming the change has no impact on performance, biological safety, or sterility, you do not have a grouping argument yet.

Where I see grouping fall apart

The most common reason a Notified Body splits a file is that the manufacturer grouped devices with different sterilization methods. Ethylene oxide and gamma irradiation are not interchangeable from a biological safety or process validation perspective, and the Basic UDI-DI logic treats sterilization as part of the essential characteristics. So even if everything else lines up, two sterilization routes means two Basic UDI-DIs.

Materials in patient contact are another sticking point. If one variant uses an animal-derived component and another does not, the biological evaluation, the risk file, and often the regulatory pathway diverge. The same applies when one variant contains a substance covered by Annex I section 10.4 (CMR or endocrine-disrupting) and another does not.

Intended purpose drift is something I see a lot. It usually shows up as a manufacturer trying to keep a diagnostic variant and a therapeutic variant in the same file, or grouping devices used in different anatomical sites with different clinical risk profiles. The clinical evaluation cannot reasonably cover both without splitting the analysis, and once you split the clinical evaluation, the rationale for one technical file collapses.

Mode of action is a quieter one but it comes up with active devices. Two pumps that look similar mechanically but use different control algorithms, or two energy-based devices delivering through different physical principles. These are usually separate Basic UDI-DIs even when the housing and the user interface look identical.

What a clean grouping rationale looks like

When the grouping holds, the technical documentation should make it obvious. I look for a section, usually near the beginning of the file, that lists every variant covered, gives the variant identifier, and walks through the three Basic UDI-DI criteria for each. The clinical evaluation should reference the same list. The risk management file should reference the same list. The verification protocols should reference the same list and state which variant was tested and why that variant represents the worst case for each test.

When manufacturers do this well, the review goes faster because the reviewer is not chasing the question “which variant does this evidence apply to” through five different documents. When they do not, that question becomes the entire deficiency letter.

A note on commercial impact

Splitting a file is not just about paperwork. Each additional technical documentation file means a separate Notified Body assessment, separate review fees, a separate surveillance plan, and a separate trail of CAPA and post-market data to maintain. For a manufacturer with twenty device variants, the difference between two well-grouped files and ten loosely-grouped files is a meaningful cost over the certificate lifecycle.

That said, forcing devices into a file where they do not belong tends to backfire. The first major deficiency, or the first post-market signal that affects only one variant, will surface the underlying issue and the file will get split anyway, just later and with more friction. So the better commercial decision, usually, is to do the grouping analysis carefully up front and accept the split where the criteria require it.

Conclusion

The MDR does not put a number on how many devices can go in one technical documentation file. It puts logic in. Same intended purpose, same risk class, same essential design and manufacturing characteristics, and you usually have a single file. Break any of those, and you have more than one. Notified Bodies expect to see the rationale written down, applied to every variant in scope, and supported by clinical and risk evidence that genuinely covers the whole range.