MDR Rule 9: The Energy Test That Decides Between Class IIa and IIb

Annex VIII, Rule 9 of the EU Medical Device Regulation governs active therapeutic devices — those intended to support, modify, replace, or restore a biological function through energy administered to or exchanged with the body. The default classification is Class IIa. Teams get the escalation wrong constantly, because they reason from the device's category or its reputation — “defibrillators are always IIb,” “physiotherapy units are always IIa” — instead of applying the test Rule 9 actually sets: whether the energy exchange could be potentially hazardous, judged by its nature, density, and site of application.

What counts as an active therapeutic device

MDR Annex I, Chapter I, Section 2.4 defines an active therapeutic device as any active device used, alone or combined with others, to support, modify, replace, or restore biological functions or structures with a view to treatment or alleviation of an illness, injury, or disability. That definition does the first filter: a device intended only to diagnose or monitor — not to treat — falls to Rule 10 instead, and an implantable device falls to Rule 8, which runs a different default and a different set of escalation conditions entirely. Rule 9 picks up what's left: non-implantable active devices whose purpose is treatment through administered or exchanged energy.

The energy test, not the device category

The regulation's own language is narrow on purpose: a device is Class IIb if its characteristics are such that it may administer energy to, or exchange energy with, the human body in a potentially hazardous way, taking into account the nature, the density, and the site of application of the energy. That is a case-by-case clinical and engineering judgment, not a lookup table. An ultrasound physiotherapy unit and an electrosurgical generator are both “active therapeutic devices administering energy” — one normally sits at IIa, the other almost never does, because the energy density and the tissue effect at the application site are different by orders of magnitude. The classification file needs to show that reasoning, not just assert the conclusion a competitor's device reached.

• Nature of the energy. Thermal, mechanical, electrical, or radiant energy carries different hazard profiles even at comparable intensities.
• Density of the energy. The same energy type at a higher density raises tissue-damage risk and pushes the device toward IIb.
• Site of application. Energy delivered near critical structures, or systemically rather than locally, changes the hazard calculus even when the device itself is unchanged.
• Ionising radiation for therapy. Devices emitting ionising radiation for therapeutic purposes are IIb regardless of the three-factor test — the exposure modality decides it outright.

A classification rationale that cites the device's category instead of the energy's nature, density, and site of application has not actually applied Rule 9 — it has assumed the answer. Why the file has to show the test, not just the result

Where Rule 9 stops

Rule 9 does not travel alone. Devices that control, monitor, or directly influence the performance of an IIb active therapeutic device inherit that IIb classification themselves, even if their own energy exchange with the patient is minimal — the regulation treats the controlling relationship as the hazard. And Rule 9 only applies once two upstream questions are settled: that the device is active and therapeutic under the Annex I definition, and that it isn't an active implantable device, which Rule 8 governs under its own default and escalation logic. An active device that is neither diagnostic, implantable, nor therapeutic in the Rule 9 sense — and doesn't fall under Rules 10 through 12, 14, 19, 21, or 22 — defaults to Class I under Rule 13's catch-all, which is a different outcome entirely.

None of this is exotic, but it is the kind of reasoning a notified body will test line by line. A classification rationale built on “devices like ours are usually IIa” does not survive that scrutiny; one built on the energy's nature, density, and site of application does. If your EU MDR classification strategy still leans on category precedent instead of the Annex VIII test, that is the gap to close before your technical documentation goes to a notified body — the same discipline matters across active cardiovascular and energy-based therapeutic devices, where Rule 9 questions come up most often.