Most companies treat PMCF surveys the same way they treat a design history file. Every question gets the same round of legal review. Every response scale got the same statistical scrutiny. Every survey gets the same three-week build cycle, no matter what device is behind it.
That instinct comes from a good place. Uniform process feels safer than one-off judgment calls. (My auditor brain loves a checklist too.) But post-market clinical follow-up, or PMCF, under the EU Medical Device Regulation covers a portfolio of activities, each carrying a different amount of regulatory weight. Portfolios reward smart allocation over habit.
If you run clinical operations across more than one product line, and your team has finite hours to spend on survey design, sample size justification, and response validation, you will run out of rigor before you run out of surveys. The real question before you build the next one is which one earns the better treatment first. And which one can survive a lighter pass.
Here's the pattern we've seen time and time again, and hear about on the podcast when clinical leads describe their own PMCF programs. Survey rigor gets allocated by calendar, not by consequence. Whichever survey is due next gets the team's current attention, regardless of whether it is holding up a Class III implant's entire clinical evidence package or supplementing a mountain of registry data for a low-risk accessory.
A team building a PMCF plan for the first time usually starts with a template. Same question bank, same Likert scale, same recruitment approach, copied across devices with wildly different regulatory exposure. Building it that way is efficient. It is also blind to where the actual compliance risk lives, because a template does not know which device it is attached to.
In one instance, a QA director had her team run the exact same 12-question satisfaction survey across a portfolio of four devices spanning three risk classes. The Class I device did not need a survey at all under its post-market surveillance (PMS) plan. The Class IIb device needed one badly, and it was the shortest, least validated instrument in the batch. Nobody had decided that on purpose. It was just the order the devices came up on the shared calendar.
The cost of that kind of drift shows up later, and it shows up in the worst possible place: a notified body review, a clinical evaluation report (CER) update, or a complaint investigation where someone finally asks how solid that survey data really is. By then the rigor gap has already become a finding, and findings come with timelines you don't always choose.
I think about this as a rigor budget: a fixed amount of design time, statistical support, and legal review capacity across a reporting cycle. These four questions decide how it gets spent.
A survey that is one of several PMCF inputs (a registry, complaint trend data, a literature review, a formal clinical investigation) has its flaws diluted by everything else in the file. A survey that is the only PMCF evidence for that device has every design weakness show up unfiltered in the CER. Solo evidence earns first claim on your best people.
The Medical Device Coordination Group's guidance document, MDCG 2020-6, lays out a clinical evidence hierarchy, and a retrospective, uncontrolled survey sits near the bottom of it. For a Class IIb or Class III device, that low rung usually is not enough on its own, so the survey needs prospective design, validated instruments, and a justified sample size to climb higher. A low-risk accessory has more room to work with and less ground to make up.
A survey feeding a CER update or a notified body audit in the next reporting window earns priority over one that will not be evaluated for two years. Rigor invested six months before anyone looks at the data pays off. Rigor invested two years early, on a survey that gets redesigned twice before then anyway, often does not.
Data behind an active safety signal gets read line by line, by your own team and by regulators. Data behind a lesser known device does not get that same scrutiny, at least not yet, which buys you room to phase in rigor rather than front-load it.
Scoring does not need to be elaborate to be useful. Give each survey one point for every question above it answers with a clear yes, so a survey that is the sole evidence source, sits low on the evidence hierarchy for its device class, feeds an imminent CER update, and sits behind an active safety signal lands a four out of four. That survey should get your best people this cycle. A survey scoring zero or one is your first candidate for a lighter, faster design pass.
Run that scoring exercise across a real portfolio and a pattern usually appears pretty quickly. The survey getting the least attention on your team's calendar is often the one that would fail hardest under real scrutiny, and the one getting the most polish is often the one that could have skipped a rigor cycle entirely.
If you're the one signing off on headcount and timelines rather than running the surveys yourself, this ranking exercise is a resourcing conversation before it's a clinical one. Every hour your clinical team spends perfecting a low-priority survey is an hour they didn't spend on the survey that will actually get scrutinized at your next notified body audit. That trade is invisible until an auditor finds it, and by then it looks like a quality problem instead of a scheduling one.
The fix doesn't cost anything extra, because it's a reallocation you can make without asking for more headcount: move your best statistician, your sharpest legal reviewer, and your longest design timeline to the surveys that scored highest on the four questions above, and run the low scoring ones on a lighter, faster process. The standard does not change, but the hours behind it do. That is what a resource-constrained clinical team should be doing anyway, matching effort to actual exposure instead of habit or calendar position.
It also gives you a defensible answer when someone from a notified body's team asks why one survey in the file looks more rigorous than another sitting right next to it. Random is a hard story to tell in that room. A four-question rationale, written down before anyone asked, is an easy one.
The mechanics of a well-designed survey matter next, once you know which survey has actually earned them.
→ Bonus resource: PMCF surveys: when they work and when they don't walks through the practical design choices, sample size, instrument validation, and recruitment that turn a high-priority survey into evidence a notified body will actually accept.
And if your ranking exercise turns up a survey drifting toward questions a survey should not be answering, exploratory endpoints, off-label use patterns, anything closer to hypothesis generation than confirmation, that is a different problem with a different fix.
→ Bonus resource: PMCF survey vs. clinical investigation covers exactly where that line sits and what crossing it costs you.
None of this triage gets easier if every survey lives in its own spreadsheet, built from scratch, scored from memory each reporting cycle. I've watched the teams that handle this well run their surveys through a single validated system, so PMCF studies, registries, and ad hoc case series all sit in one place. Ranking them by real compliance value becomes a conversation you can have in an hour instead of an archaeology project that eats a week.
That matters more as your device portfolio grows. A company running one device can hold the whole picture in someone's head. A company running eight, across four risk classes and three notified bodies, cannot, and that's where an ungoverned rigor budget turns into an audit finding waiting for a date.
If you're wondering which survey matters most, it should always be some version of the four questions above. Rank first. Design second. The rigor you have is finite. Spend it where the regulation actually looks hardest, and let the rest run lighter without apology.
If you're building out your PMCF survey process, these related guides go deeper on the specific pieces:
If you want to see what that looks like inside a system built for it, take a look at Greenlight Guru's clinical electronic data capture (EDC) platform.