Guest Column | May 24, 2018

Determining Significant Vs. Nonsignificant Risk In Medical Devices

By Jon Speer, greenlight.guru

Risk Management & Design Controls

One of your important tasks as a medical device manufacturer is to classify your device correctly. This means knowing the difference between significant and nonsignificant risk, and where your device falls on that classification, although it’s not always a clearly-defined process.

I recently spoke with Mike Drues, president of Vascular Sciences, whose background in medical technology, compliance, and working with the FDA, Health Canada, and other international agencies make him an expert on significant vs. nonsignificant risk. We discussed those classification differences and how they impact your device development process.

Significant Risk vs. Nonsignificant Risk

There are a number of ways in which we separate or classify medical devices, including:

  • The “Class I, II or III” system. This is theoretically based on risk, although there are different definitions of risk that can be used.
  • The idea of significant risk (SR) or nonsignificant risk (NSR). The definition of risk here is really not well-defined. The ramifications of your decision are important in terms of what you have to do and who you have to notify during your development process.

When discussing risk, Mike points us toward three “buckets" that he favors in determining classification:

  • Probability of direct harm — For example, the probability of direct harm to the patient or even a caregiver.
  • Probability of harm from not using the device — What other options are available if the device isn’t used? This is a PMR requirement.
  • Probability of harm from the wrong information — This covers things like cancer detection and mobile medical apps. What happens if the device gives false indications?

Consider all three of these buckets when making an SR vs. NSR evaluation. Be sure to document your decisions — it’s not a clearly defined process, but you need to be able to show why you arrived at a particular conclusion.

Who Gets To Determine Risk?

I’ve worked with companies that have thought they could conduct a clinical trial and reliably conclude NSR. It’s unfortunate that many seem to try this because they want to avoid dealing with the FDA or having to submit an IDE.

The manufacturer gets to make the risk determination, according to regulations. That may seem a bit like putting the fox in charge of the henhouse, but Drues asserts that companies shouldn’t use this power as an excuse to try avoiding SR classification. Liaising with the FDA could be a key opportunity to work together. Further, Drues strongly suggests taking even NSR devices to the FDA, prior to trials, as a courtesy. This is not to ask for permission or “Is this ok?”, but to keep the agency in the loop and ensure everyone is on the same page.

We’ve both seen companies assume their device is NSR and go to trials, only to have the FDA say afterwards, “We need to see X,” or, “Hey, we don’t see this device as NSR at all.” This can lead to real trouble for companies — in extreme cases, you won’t just be dealing with the FDA, but the Justice Department, as well.

Where To Go For Guidance

There are a few guidance documents that you can turn to. Besides getting to know the guidance, though, it’s on the device developer to observe due diligence throughout development. If you’re in a category with similar devices, these are a great place to look for any precedents. If nothing is even close, it’s still going to be your own determination at the end of the day. Use some form of analysis and document it.

Drues is a huge fan of using subject matter experts, such as clinicians or engineers, to give their view. This does not mean asking them to write a letter praising your device, but to provide input, based on their own assessment, of how they would classify your device. These opinions are great to use as letters to file, which can be used as evidence to show you didn’t forget or hide anything from the FDA.

Making Contact With Regulatory Bodies

The pre-submission process probably is the most appropriate method to make initial contact with the FDA. If your device truly falls in the grey area between SR and NSR, you might request a pre-submission meeting dedicated to the topic. Again, you have no obligation to talk to the FDA about NSR devices — that’s Drues’ recommendation. However, you do need to present your device to Institutional Review Board (IRB).

The IRB is like backup for the FDA. It exists at the level of the local institution, where you conduct your clinical trial, and its most important role is to protect the safety of test subjects locally. If you haven’t gone to the FDA, it is the only group between you and the patient. In our view, the IRB becomes even more important to devices classified (or likely to be classified) as NSR devices, because it should look at things that would otherwise be checked by the FDA.

The bottom line is that choosing to work on an NSR device in the hopes of avoiding the FDA isn’t a shortcut. In fact, given that you still need to show the same information to the IRB, it involves the same amount of work. It’s our responsibility as medical device professionals to be able to present our case. How will you objectively show that the product is safe for clinical trial?

Some think that avoiding the FDA will speed the process to market, but this isn’t really the case. The IRB is not subject to requirements akin to the FDA ones, which require the process to keep rolling. This means review can take a very long time, especially if the IRB lacks the FDA’s subject matter expertise.

Clinical Trials

Would you feel comfortable using this device on a family member or friend? If the answer is “Yes,” then you’re probably good to go; if you’re not sure, then it’s probably not ready. Whether a device is SR or NSR, it should be substantially ready to go before any clinical trials. You can’t expect to keep going back and iterating when peoples’ health and lives are at stake.

There’s a growing (and controversial) trend of U.S.-based companies running clinical trials outside of the U.S. There’s a perception in many cases that they’ll be able to get clinical experience faster outside of the U.S., or that, for SR devices, they can avoid the IDE and the process with the FDA. They’re trying to avoid regulatory “hurdles,” but to us it should still come back to that question: Would I use this on my own family members?

The answer should tell you whether it’s a good idea to conduct clinical trials anywhere. Why are we doing this study? Are we planning on giving the information to the FDA or not?

There are some scenarios where companies want a “quick and dirty” human study to determine which device prototype they’re going to pursue. They’re never planning to share this information with the FDA. In cases where they want to submit data to the FDA, they’d better complete all appropriate processes as they would in the US. Historically, the FDA hasn’t solicited data procured outside the US, but they have accepted it recently. This is something Drues would add in the pre-submission, just as a “by the way, this is what we’re doing, so you’re aware.”

We would both add here that, no matter how you’re planning on using the data, going outside of the U.S. is not a substitute for good science and good clinical practices. It’s not a strategy to avoid the FDA. Finally, note that the 2015 FDA guidance Acceptance of Medical Device Clinical Data from Studies Conducted Outside the United States was withdrawn in February this year, and no longer represents FDA’s current thinking on the topic.

Final Thoughts

Knowing how to classify your device as either SR or NSR is not a clearly defined task. There are implications for each decision made as you proceed through your development process, and the burden is on you to understand and provide what the FDA will require of you.

The idea that classifying as NSR will help you to avoid the FDA interaction is false. For one thing, you might find they question you on it at pre-submission, especially if you went ahead with clinical trials. Secondly, you still have to put in the same work for local IRBs.

Our previous advice bears repeating: Always follow good science and good clinical practices, and ask yourself, “Would I use this on my own family?”

About The Author

Jon Speer is the founder and VP of QA/RA at Greenlight Guru, a software company that produces the only modern quality management software solution exclusively for medical device companies. Jon is a medical device industry veteran with over 20 years experience, having helped dozens of devices get to market over his career in a variety of roles, including product development, project management, quality, and regulatory. He is a thought leader, speaker and regular contributor at numerous leading industry publications. He also is the host of the most-downloaded podcast in the industry, The Global Medical Device Podcast.