Medical device regulation: how classification decides the cost of everything else

Classification is a technical argument with commercial consequences

Class is determined by intended purpose, duration of contact, invasiveness, whether the device is active, and how it interacts with the body — not by how simple the engineering looks. A change of intended purpose in marketing copy can move a product between classes without a single design change. Because the route to market, the evidence burden and the involvement of an independent assessment body all follow from class, the determination is made early, documented with its reasoning, and revisited whenever claims change. It is also the point where an inexperienced manufacturer most often gets contradicted.

The quality system reaches back into design

Device quality systems, commonly built to ISO 13485 and to regulator-specific expectations, cover design control as well as production: user needs traced to requirements, requirements traced to verification, design reviews recorded, risk management running through the whole lifecycle rather than filed once. Engineering teams arriving from other sectors find the traceability burden unfamiliar and often build it retrospectively, which produces a design history that looks assembled because it was. Building the file as the project runs costs far less than reconstructing evidence for decisions taken by people who have moved on.

Clinical evidence and why the literature route is narrowing

Demonstrating that a device performs and is safe rests on clinical data, whether generated directly or drawn from equivalent devices and published sources. Reliance on equivalence has become harder in several jurisdictions, particularly for higher-risk products, which changes the economics of legacy portfolios: a product sold for years may need evidence it never had. Manufacturers with wide catalogues face a portfolio decision rather than a regulatory one, because generating data for every legacy variant can cost more than the variants earn. That triage is worth doing deliberately rather than by deadline.

Traceability, labelling and the ability to find units in the field

Device regimes expect a manufacturer to identify products through distribution and to know where units went, supported by unique identification carried on labelling and in registration databases. Operationally that means serial or lot control through production, correct data submitted and maintained in registries, and distribution records that survive a distributor changing systems. Where firms come unstuck is in the field: units transferred between hospitals, servicing done by third parties, and implants recorded in patient systems the manufacturer cannot see. The strength of the distribution record determines how targeted a field action can be.

Post-market duties do not pause between submissions

Once devices are in use, obligations continue: collecting complaints and performance data, reporting serious incidents to authorities within defined periods, updating documentation and risk files as experience accumulates, and taking field action where needed. Regulators examine whether the surveillance was active rather than passive. Since classification rules, assessment routes, timelines and reporting duties differ between the United States, Europe and other markets, and this page is orientation only, work from the relevant regulator's own material and take the determinations to a regulatory professional. Complaint handling and the reporting decision need a named owner who is reachable when the plant is closed.

Frequently asked questions

We make components for a device maker. Are we regulated too?

Often the finished-device manufacturer holds the regulatory position while you are treated as a supplier, which means your obligations arrive contractually: specification control, change notification, record retention, right of audit and material traceability. That said, some components and accessories are regulated in their own right depending on intended purpose. The line is product-specific and worth confirming with a regulatory adviser rather than inferring from what your customer tells you they need.

How much does a design change affect our regulatory position?

It depends on whether the change touches safety, performance or intended purpose. Changing a supplier of a non-contacting fastener sits differently from changing a material in contact with tissue, a sterilisation method or software behaviour. Most quality systems handle this with a documented change assessment that decides whether verification, validation, updated documentation or notification to an authority is needed. Building that assessment step into engineering change control keeps the decision from being made informally on the shop floor.

Does one market's approval help in another?

It helps with the evidence rather than with the authorisation. Testing, clinical data and quality system documentation are generally reusable, and some programmes allow audit reports to be shared between participating regulators, which reduces duplicated inspection. The authorisation itself remains territorial, with its own classification rules, submission format and local representation arrangements. Plan market entry sequence around where the evidence you already hold goes furthest, and confirm each route with the regulator concerned.