The certification system under load

Conformity assessment is easy to misunderstand from a distance. To outsiders it can look like paperwork standing between a finished product and a market. Inside a Notified Body, it is something else: scarce technical and clinical expertise applied to large, evolving evidence packages under strict rules, with real consequences for patients, manufacturers, and the institutions that sign the certificate.

That work has always been demanding. Over the last several years it has also become structurally overloaded. The Medical Device Regulation (MDR) and In Vitro Diagnostic Regulation (IVDR) raised expectations for clinical evidence, documentation, and ongoing surveillance. Transition deadlines were extended more than once. Applications and certificates have both grown, but not in lockstep. Average timelines stretch measured in many months, often beyond a year. Manufacturers report diverted resources and a weaker preference for launching first in Europe.

The system is not failing because rigor went out of fashion. It is straining because demand, documentation volume, and assessment complexity have outgrown the way expert capacity is organized.

Demand and throughput

By end of 2025, MDR applications still exceeded certificates among Team-NB designated members (25,978 vs 13,953), a persistent throughput gap (Team-NB Survey 2025). The cumulative gap is a system-load signal, not a simple backlog - certificates doubled in 2023 and 2024 before slowing to about two-thirds growth in 2025, yet applications remained well ahead and can also be withdrawn, refused, duplicated, or still in process.

Designation capacity has expanded: Team-NB notes MDR designated Notified Bodies rising from 18 in 2020 to 52 in 2025 (Team-NB Survey 2025). More certificates are real progress. They do not erase multi-year transition pressure, queues, or long calendar times. In 2025 the association also reported the first decline in conformity-assessment staffing in more than a decade.

Time is the capacity metric people feel

Most new MDR certifications take well over a year - about 70% at 13 months or more in Team-NB's 2025 distribution (Team-NB Survey 2025). For new (not renewal) certificates, only about 31% average under twelve months; the largest band is 13-18 months (48%), with 22% at 19 months or longer, essentially unchanged despite more designated bodies and more certificates.

Manufacturer-side surveys tell a compatible story from the other end of the table. MedTech Europe’s 2024 IVDR & MDR survey finds average times for medical-device quality-management-system assessment around 19.5 months and for technical documentation assessment around 21.8 months (MedTech Europe 2024). Those averages include phases that are not pure document review, pre-review and certificate issuance among them, which is part of the point: the clock patients and manufacturers experience is a system clock, not only an auditor’s reading time. No newer manufacturer survey has superseded those averages as of this writing.

Transition extensions are a symptom

When the Union extended certain MDR and IVDR transitional provisions through Regulation (EU) 2023/607, it was not celebrating success. It was acknowledging that the system could not absorb legacy transition on the original timetable without unacceptable risk to device availability. Staggered deadlines into the later 2020s bought time. They did not invent new expert hours, nor did they make technical documentation easier to navigate.

Policy can rephase a cliff. It cannot, alone, reorganize how scarce reviewers spend their days. Extensions are therefore best read as evidence of structural load, not as proof that the load has been solved.

What the strain costs

Load is not only an operational inconvenience for assessment teams. It reshapes where innovation is launched and how manufacturers allocate attention. MedTech Europe reports that, since the Regulations applied, the choice of the EU as first-launch geography has fallen substantially for medical-device manufacturers: on the order of about one-third for large companies and about one-fifth for SMEs in their respondent set (MedTech Europe 2024). The same survey links regulatory burden to pressure on innovation activities and to the difficulty of securing qualified regulatory staff.

Those findings do not imply that standards should be weakened. They imply that when assessment capacity and predictability erode, the costs show up as delayed access, diverted R&D attention, and geographic strategy that routes around Europe, even when European patients and clinicians would benefit from earlier availability of safe devices.

• Patients and health systems face delayed or uncertain access when certificates lag demand.
• Manufacturers face multi-year planning horizons, rising compliance cost, and weaker first-launch preference for the EU.
• Notified Bodies face queues that cannot be cleared by designation counts alone while every file still requires accountable human judgment, and staffing that can contract when submission volume dips.

What this diagnosis is not

It is tempting to collapse the problem into a single villain: “too few Notified Bodies,” “too much regulation,” or “manufacturers who submit poorly.” Each of those has partial truth and none is the whole system. Designation has increased. Requirements were deliberately strengthened after public-health failures under older regimes. Incomplete and hard-to-navigate files are a documented capacity tax, but they do not excuse treating expert reviewers as infinitely elastic. Nor does a temporary easing of NB workload mean the assessment environment has scaled: time bands and completeness friction remain sticky.

The more precise diagnosis is this: rigor remains non-negotiable, while the information and workflow environment around assessment has not scaled with the load. When applications outpace certificates, when average times sit past a year, and when policy must extend transitions to protect availability, the system is telling us that usable expert time is the binding constraint.

Expanding capacity without diluting rigor is not a slogan for faster rubber stamps. It is the only coherent response to a world that needs both safe devices and an assessment system able to keep pace.

This first piece establishes the load. The next essays explain where expert hours go, how binding timelines change the operating environment, and how variance and generative tools compound the constraint.

The standard cannot fall. The system still has to move. Understanding the load is the first obligation of anyone who claims to help it do both.