You ever read a regulation and feel like it was written to confuse you on purpose? On the flip side, yeah. ISO 14155:2020 is one of those documents that sounds dry as toast — but if you're running a clinical study on a medical device, it's the rulebook you can't ignore Nothing fancy..
Counterintuitive, but true.
Here's the thing — the iso 14155:2020 gcp standard applies to clinical investigations of medical devices in humans. Not drugs. Here's the thing — not supplements. Devices. And that distinction matters more than most people realize until they're three months into a study and someone flags a compliance gap.
People argue about this. Here's where I land on it.
What Is ISO 14155:2020
So what is this thing, really? ISO 14155:2020 is the international standard for good clinical practice (GCP) in clinical investigations of medical devices. It's the updated version of the 2011 edition, and it got a serious overhaul to line up better with how modern device studies actually run And that's really what it comes down to..
Think of it as the operating manual for doing a device trial the right way. It covers how you design the study, get ethics approval, protect the people in it, collect data, and report what happened. It's not a law by itself — but in a lot of places, regulators treat it like one because they've adopted it directly or folded it into their own rules.
Medical Devices, Not Drugs
This is the part most folks mix up. The standard is built for devices — things like implants, diagnostic kits, surgical tools, wearable monitors, software as a medical device. If it's regulated as a device and you're testing it on people, ISO 14155 is your GCP reference Took long enough..
Drug trials answer to ICH-GCP. Think about it: they can have learning-curve effects for surgeons. Also, they can be modified mid-study. They interact with anatomy in ways a pill doesn't. Device trials don't fit neatly there, because devices behave differently. ISO 14155 was written for that reality Worth keeping that in mind. Less friction, more output..
Investigational Device vs Approved Device
The standard applies whether you're testing something brand new or studying an already-approved device for a new use. If the investigation is systematic and involves humans to assess safety or performance, you're in scope. Even post-market clinical follow-up studies can pull from these principles, though those have their own ISO (that's 16142, if you're curious).
Why It Matters
Why should you care? Worth adding: because a device trial done without ISO 14155 discipline falls apart fast. Not maybe — fast.
I've seen teams treat device studies like engineering validation with a few patients tacked on. Even so, the moment you put a device in or on a person, you're dealing with consent, adverse events, and data integrity. That's a mistake. Skip the standard's structure and you'll struggle to get past an ethics committee, let alone a regulator.
This changes depending on context. Keep that in mind.
And here's what most people miss: ISO 14155:2020 actually helps you. Because of that, the 2020 update clarified risk-based monitoring, added detail on ethical considerations, and made the link to ISO 14971 (risk management) explicit. Here's the thing — the old version was clunky. In practice, that means you can build a smarter study instead of a bloated one Still holds up..
Turns out, regulators in the EU, through the MDR (Medical Device Regulation), point straight to ISO 14155 for clinical evidence. If your investigation doesn't meet it, your CE mark pathway gets rocky. In the US, the FDA doesn't cite it by name, but their device GCP expectations overlap heavily. So knowing this standard is knowing the global baseline Still holds up..
How It Works
Alright, let's get into the guts. How does ISO 14155:2020 actually shape a clinical investigation? It's not one big rule — it's a web of requirements across the study lifecycle.
Planning and the Clinical Investigation Plan
Everything starts with the Clinical Investigation Plan (CIP). It's device-specific. This isn't a protocol borrowed from a drug study. You define the device, the intended use, the endpoints, the patient population, and the statistical approach Easy to understand, harder to ignore..
The standard says the CIP has to show the investigation is scientifically sound and ethically justified. You can't just hope the data works out. You need a clear hypothesis and a method that can actually test it That's the whole idea..
Risk Management Link
Here's a detail that separates 2020 from the old days: ISO 14155 now expects you to tie the clinical investigation into your ISO 14971 risk management file. That means known device risks show up in your monitoring and your consent forms. If a pacemaker lead has a known fracture risk, your CIP better say how you'll watch for it Not complicated — just consistent..
In practice, this stops teams from treating clinical risk as someone else's problem. It's your problem, documented and tracked That's the part that actually makes a difference..
Ethics and Informed Consent
The standard is strict on ethics. It has to be informed, voluntary, and documented. You need approval from a recognized ethics committee before a single patient is enrolled. And consent? No coercion, no fine print nobody reads Worth keeping that in mind..
For devices, there's a wrinkle: sometimes the investigator has to know which patients got the device vs a comparator, but the patient doesn't. Which means that's allowed under certain conditions — but the standard spells out when and how. Here's the thing — most guides skip this nuance. It's right there in the text Not complicated — just consistent..
Conduct and Monitoring
Once the study runs, ISO 14155 expects risk-based monitoring. You don't have to visit every site every month if your risk profile says otherwise. But you do have to justify it. Source data gets verified. Deviations get logged. Adverse events get classified and reported on time.
And the sponsor owns a lot of this. Not the hospital, not the PI alone — the sponsor. That surprises smaller device companies who assume the clinic handles compliance.
Data Handling and Reporting
The standard covers data capture, retention, and the final report. Your data has to be traceable and complete. The clinical investigation report (CIR) is the deliverable regulators read, and ISO 14155 tells you what it must contain. Skip a section and you'll hear about it That's the part that actually makes a difference. And it works..
Common Mistakes
Let's talk about where people trip. Because honestly, this is the part most guides get wrong.
One big miss: treating ISO 14155 like a checklist instead of a system. Also, teams print the table of contents, tick boxes, and miss the connections between risk management and the CIP. The standard is meant to be read as a whole.
Another: using drug-style protocols. But i know it sounds simple — but it's easy to miss. Day to day, a drug protocol with "device" swapped in won't cover usability, training of users, or device-specific failures. ISO 14155 expects you to think like a device person.
Then there's consent. That's why people write a generic form and call it done. Plus, the standard wants device-specific risks explained in plain language. If your implant can set off airport metal detectors, say so. Patients deserve that, and the standard agrees Still holds up..
And the last one — poor adverse event handling. But a subtle vision change from a surgical tool's lighting? Investigators under-report because they don't link it back to the device risk file. And less so. A rash from a wearable strap is easy. Plus, device events aren't always obvious. That's a compliance hole No workaround needed..
Practical Tips
So what actually works when you're facing this standard?
First, read the real document. Not a summary. The 2020 version is readable compared to older standards. You'll save money by understanding it before hiring a consultant to explain your own study to you.
Second, build your CIP and your ISO 14971 file in the same room. Literally or figuratively. When the risk file and the investigation plan talk to each other, audits go smooth.
Third, train your sites on device GCP, not just generic research. A PI who's done drug trials may miss device nuances. Give them the short version: this is different, here's why Still holds up..
Fourth, plan your monitoring before enrollment. Write the rationale. If you go risk-based, show the math. Regulators like a sponsor who thought ahead.
Fifth, keep the consent form human. Real talk — patients sign faster when they understand. Use the standard's requirements as a floor, then write like a person.
FAQ
Does ISO 14155:2020 apply to in vitro diagnostics? It applies to clinical investigations of medical devices, and IVDs are devices. But studies done in labs without human interaction follow different paths. If the IVD is used on human samples in a clinical performance study, the standard's principles often
apply by analogy, even when local regulations carve out specific exemptions. The key is whether the investigation involves human subjects or human-derived specimens in a way that informs safety or performance claims—if yes, the spirit of ISO 14155 should guide your setup Easy to understand, harder to ignore..
Can a single investigator-initiated study skip the sponsor responsibilities? No. Under ISO 14155, the sponsor duties don’t vanish just because the PI is also the funder. The standard expects clear role separation on paper, even if one person wears both hats. Document who does what, or the audit will assume no one did Simple, but easy to overlook..
What’s the biggest documentation trap post-study? The close-out report. Teams finish enrollment, lock the database, and sit on the final summary for months. ISO 14155 expects a defined timeline for study completion and reporting. Treat close-out like a deliverable, not a cleanup task.
Conclusion
ISO 14155 isn’t a hurdle to clear—it’s the operating manual for running device studies that actually protect patients and hold up under scrutiny. The teams that do well aren’t the ones with the biggest compliance budgets; they’re the ones who read the standard, linked it to their risk work, and trained their people to think like device investigators from day one. Get the fundamentals right, keep the paperwork human, and the rest of the clinical path gets a lot quieter.
People argue about this. Here's where I land on it.
