Why your device family keeps getting challenged at review

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

This scenario repeats itself in audit after audit. Manufacturers group devices into families without documenting the rationale. They treat it as an administrative convenience. Notified Bodies treat it as a clinical safety question.

The tension is real. And it matters.

Because if the grouping is not justified properly, the entire clinical evaluation can be invalidated. You cannot borrow clinical data from one device to support another unless you can demonstrate that the similarity is clinically relevant.

This is not about saving time. It is about proving that the time you saved did not compromise patient safety.

What MDR actually allows you to group

The MDR does not define device families. It defines equivalence.

Article 61(5) allows manufacturers to rely on clinical data from an equivalent device if they can demonstrate that the two devices are similar in terms of clinical, biological, and technical characteristics.

That is the framework. Equivalence is about data transfer. It is about justifying why clinical evidence from device A can support the safety and performance of device B.

Device families are an extension of this logic. Instead of justifying equivalence one by one, you justify it for a group.

But the burden of proof does not disappear. It concentrates.

If you cannot make that argument clearly, the grouping will not hold.

The three dimensions that define a valid family

Manufacturers often think in terms of product lines. Same brand. Same platform. Same intended use.

That is not enough.

A valid device family must demonstrate similarity across three dimensions: technical, biological, and clinical.

Technical similarity means the devices share the same design principles, operating mechanisms, and performance characteristics. They do not need to be identical. But the differences must not introduce new risks or alter the clinical profile.

Biological similarity means the devices have the same body contact type, duration, and tissue interaction. A device that touches intact skin and a device that penetrates tissue are not biologically similar, even if they look alike.

Clinical similarity means the devices are used in the same patient population, for the same clinical indication, by the same type of users, in the same setting. If one device is used in surgery and another in home care, they are not clinically similar.

I have seen this happen with surgical instruments that differ only in size. The manufacturer assumed size was a minor variation. The Notified Body pointed out that the larger size changed the tissue trauma profile and required separate clinical justification.

The lesson: every difference must be evaluated for its clinical impact. If the impact is negligible, document why. If the impact is significant, the device does not belong in the family.

What you must document to justify the grouping

So how do you prove that your device family is valid?

You start with a table. A comparison matrix.

List all devices in the proposed family. Then compare them systematically across technical, biological, and clinical characteristics.

For technical characteristics, include: materials, design features, energy sources, sensors, software functions, accessories, sterilization methods. Any technical element that could influence safety or performance.

For biological characteristics, include: body contact type, contact duration, invasiveness, tissue type, systemic circulation, implantation. Use ISO 10993-1 as a framework.

For clinical characteristics, include: intended use, indication for use, patient population, contraindications, user training requirements, clinical setting, duration of use.

Then document the differences.

Because differences are not disqualifying. Unexplained differences are.

If one device has a coating and another does not, explain what the coating does. Does it prevent infection? Does it reduce friction? Does it improve visibility? If the coating changes the clinical profile, the devices are not in the same family. If it does not, say why.

If one device is single-use and another is reusable, explain the sterilization validation and the contamination risk. If the risk is the same, document it. If it is different, split the family.

This is not paperwork. This is reasoning made visible.

When you should not create a device family

The pressure to group devices is constant.

Clinical data is expensive. Clinical evaluations take time. If you can evaluate 10 devices as one family instead of 10 separate reports, the efficiency gain is obvious.

But there are situations where grouping is a mistake.

First case: when the devices differ in risk class. A Class IIa device and a Class IIb device cannot be in the same family. The conformity assessment routes are different. The clinical evidence requirements are different. Forcing them into one family creates confusion and audit risk.

Second case: when the devices have different indications for use. A diagnostic device and a therapeutic device are not clinically similar, even if they measure the same parameter. The clinical context is different. The decision-making is different. The evidence required is different.

Third case: when the clinical data available is heterogeneous. If you have strong clinical data for device A and weak data for device B, grouping them does not strengthen device B. It weakens device A. Notified Bodies will challenge the entire family if the evidence for any member is insufficient.

I have reviewed families that started with three similar devices and grew to include twelve. The original rationale was sound. But each addition introduced new features, new indications, new risks. No one updated the comparison matrix. No one questioned whether the family still held together.

By the time the Notified Body reviewed it, the family was indefensible.

How device families interact with PMCF and vigilance

Here is where manufacturers often miscalculate.

They think that if the devices are in the same family, they can run one PMCF study and one vigilance analysis for all of them.

That is partially correct. And partially dangerous.

You can design a PMCF plan at the family level. You can collect data that applies to all devices. But you must still track device-specific outcomes.

Because if one device in the family shows a higher complication rate, you need to know which one. You need to analyze whether the difference is random or whether it signals a design issue.

If you aggregate all data without device-level tracking, you lose this visibility. You dilute the signal. You delay the detection of safety issues.

The same logic applies to vigilance. Incident reports should specify which device variant was involved. Trend analysis should include device-level breakdowns. If you report everything as “device family X,” you cannot identify which variant is driving the incidents.

I have seen manufacturers surprised when a Notified Body asked for device-specific PMCF data after they submitted a family-level report. The manufacturers thought the family concept meant full aggregation. It does not. It means shared evaluation with individual monitoring.

What Notified Bodies focus on during review

When a Notified Body reviews a device family, they ask three questions.

First: Is the grouping logic documented and defensible? Can the manufacturer explain why these devices belong together in clinical terms?

Second: Are the differences between devices analyzed for clinical impact? Has the manufacturer considered whether variations in size, material, or function change the risk profile?

Third: Is the clinical data representative of the entire family? If the clinical studies were conducted on device A, does the evidence apply to device B?

If the answer to any of these questions is no, the family will be challenged.

The most common issue I see is the third one. Manufacturers group devices together, but all the clinical data comes from one or two variants. The others are included based on assumption, not evidence.

Notified Bodies do not accept assumptions. They accept reasoning. They accept gap analysis. They accept bridging arguments. But the argument must be explicit.

This is not bureaucracy. This is patient safety.

If you claim that device B is clinically equivalent to device A, but device B has never been studied, you must explain why the absence of direct data does not create uncertainty about safety and performance.

If you cannot explain it, you cannot claim it.

The efficiency calculation you should actually make

So should you create device families or not?

The answer depends on the quality of your grouping logic.

If your devices are genuinely similar in ways that matter clinically, grouping them is efficient and defensible. You reduce duplication without compromising rigor.

If your devices are similar in some ways but different in others, you need a more nuanced approach. Group the truly similar ones. Evaluate the outliers separately.

If your devices are grouped for convenience rather than clinical logic, do not expect the grouping to survive review. You will spend more time defending a weak family than you would have spent evaluating the devices separately.

The real efficiency is not in reducing the number of reports. It is in aligning your documentation with the clinical reality.

Because when the clinical evaluation reflects the actual risk profile, Notified Bodies have less to challenge. When it does not, they have everything to challenge.

I have worked with manufacturers who split an oversized family into three smaller, better-defined families. The amount of documentation increased slightly. The time to approval decreased significantly.

Because the Notified Body could see the logic. They could verify the claims. They could approve with confidence.

That is the goal. Not fewer pages. Clearer reasoning.

What this means for your next submission

If you are planning to group devices into a family, start with the comparison matrix. Build it early. Use it to test whether the grouping makes sense before you write the clinical evaluation report.

If the matrix reveals significant differences that you cannot explain away, reconsider the grouping. Better to submit two focused reports than one sprawling family that collapses under review.

If you already have a device family in your technical file, audit it. Check whether the grouping logic is documented. Check whether new variants have been added without updating the rationale. Check whether the clinical data still supports the entire family.

Because Notified Bodies are not getting more lenient. They are getting more systematic.

And device families are one of the first places they look for weak reasoning.

If your family is solid, you have nothing to worry about. If it is not, you will find out during review. Better to find out now.

Peace,
Hatem
Clinical Evaluation Expert for Medical Devices
Follow me for more insights and practical advice.