Regulatory Open Forum

This message was posted by a user wishing to remain anonymous

Hello all, 
which information do you suggest to provide to the IRB for a "non-significant risk" determination?
Our device is a class II device, it is not an implant, it is not used to support or sustain human life, and it is not of substantial importance in diagnosing, curing, mitigating, or treating disease, or otherwise preventing impairment of human health. 
What about the last bullet point of the "significant risk device definition" 21CFR 812.3 (m), i.e.: (4) Otherwise presents a potential for serious risk to the health, safety, or welfare of a subject?
Would the complete risk analysis with the conclusion that "all risks, after risk control measures and application of the risk/benefit analysis, have been assessed as acceptable" be sufficient?

Also, is there any database or any mean to see if predicate devices were required to submit a full IDE? I don't manage to find such a database in the FDA website nor such information (e.g., IDE number) in clinicaltrials.gov.
Thanks in advance for the help.

Good day,

Have you looked at this guidance document?  https://www.fda.gov/media/75459/download which is published by the FDA though gives the same reference to 21 CFR 812 as you mention.  There is no further guidance or information available (though there are some presentations and couple papers out there on this topic could research) but in effect you need to create a risk assessment document for the product.  Go through each of the four bullet points to address each one, i.e. not an implant, not life-sustaining, importance in diagnosis/treating, and risks involved with the device.  There are a number of ways this can be presented to the IRB which I have done a couple different versions over the years.  Just some words of advice, do not give them your risk analysis FMEA - many will not know what they are looking at.  You do want to keep it simple and straightforward so they understand.  A couple times, I provided information which was extracted from the regulatory strategy (which I used later for the actual submission) such as device description, indications for use, intended purpose, patient population, mode of action, operation principle, etc.  Then I had a "easy" risk assessment table where I listed some of the general risks and in particular the residual risks; this all included the hazard, hazard situation, and harm which could result from the risks identified.  If this is a new, novel device you might have difficult staying out of the SR region for determination.  If the product is fairly well-understood, then the risk profile should be stated to go into the NSR area for determination.

To answer your final question, no there is no public database for IDEs or IDE information.  You can request IDEs through the Freedom of Information (FOI) Act, but this will take a very, very long time unless you have specific information like the IDE number and dates of approval.  Sometimes in the Summary for Safety and Effectiveness of the PMA (and sometimes in the 510(k) Summary) the company will list the IDE number used for the clinical studies performed - not always.  This just means you would need to go into the original PMA submission SSED letter to see if this is listed.  Also sometimes if there is a clinical study performed in the US, in the clinical trials database under the NCT number, also sometimes the IDE number might be listed.  This would just be going into each record individually, but there is no requirement to have this listed.

------------------------------
Richard Vincins RAC
Vice President Global Regulatory Affairs
------------------------------

Original Message:
Sent: 07-Dec-2020 12:21
From: Anonymous Member
Subject: non-significant risk study determination

This message was posted by a user wishing to remain anonymous

Hello all,
which information do you suggest to provide to the IRB for a "non-significant risk" determination?
Our device is a class II device, it is not an implant, it is not used to support or sustain human life, and it is not of substantial importance in diagnosing, curing, mitigating, or treating disease, or otherwise preventing impairment of human health.
What about the last bullet point of the "significant risk device definition" 21CFR 812.3 (m), i.e.: (4) Otherwise presents a potential for serious risk to the health, safety, or welfare of a subject?
Would the complete risk analysis with the conclusion that "all risks, after risk control measures and application of the risk/benefit analysis, have been assessed as acceptable" be sufficient?

Also, is there any database or any mean to see if predicate devices were required to submit a full IDE? I don't manage to find such a database in the FDA website nor such information (e.g., IDE number) in clinicaltrials.gov.
Thanks in advance for the help.

I concur with Richard on the FMEA.  When an engineer gives me a FMEA, I always run my finger down the Effects column and highlight those that involve clinical harm to patients and/or other users.  I extract those as the risks that need to be analyzed.  Then I pull out the relevant information (e.g., Severity, Mitigation) from the other columns for those risks.

You can easily dispense with #1 and #2 of the definition. As the RA team member, I'd probe a bit about #3, because that can be subjective.  Even if I were confident it didn't meet this criteria, I wouldn't dispense with it quite so easily as the first two.  I would reiterate intended use and EXPLAIN why it is not of substantial importance for diagnosing, treating, or whichever of these functions it is of unsubstantial importance for.  Then I'd sit down with the information from the FMEA and start in on #4.  I would not rely on the fact that the FMEA spreadsheet rated a risk as not severe.  I would spell out why each of the clinical risks identified in the FMEA has no potential to cause serious harm.

Unless you work with a commercial IRB that specializes only or mostly in medical devices, I would not assume the IRB has a good grasp on NSR.  I would explain it to them in my submission, including the role of the IRB, and cite the definition.  Then, once you have done that, you can smoothly segue into how your device meets all four points in the definition.  I would also make it clear that I am not just seeking approval of the study, but also confirmation by the IRB that it is NSR.

I also concur with Richard about the non-existence of an IND/IDE database, and also recommend looking at the 510(k) decision summaries for your potential predicates.  You never know what they might tell you.  However, unless you can narrow it down to a few competitors, I wouldn't spend a lot of time on this, as they really aren't all that likely to give you useful information.


------------------------------
Julie Omohundro, ex-RAC (US, GS), still an MBA
Principal Consultant
Class Three, LLC
Mebane, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com
------------------------------

Original Message:
Sent: 07-Dec-2020 12:21
From: Anonymous Member
Subject: non-significant risk study determination

This message was posted by a user wishing to remain anonymous

Hello all,
which information do you suggest to provide to the IRB for a "non-significant risk" determination?
Our device is a class II device, it is not an implant, it is not used to support or sustain human life, and it is not of substantial importance in diagnosing, curing, mitigating, or treating disease, or otherwise preventing impairment of human health.
What about the last bullet point of the "significant risk device definition" 21CFR 812.3 (m), i.e.: (4) Otherwise presents a potential for serious risk to the health, safety, or welfare of a subject?
Would the complete risk analysis with the conclusion that "all risks, after risk control measures and application of the risk/benefit analysis, have been assessed as acceptable" be sufficient?

Also, is there any database or any mean to see if predicate devices were required to submit a full IDE? I don't manage to find such a database in the FDA website nor such information (e.g., IDE number) in clinicaltrials.gov.
Thanks in advance for the help.