Hello Anon,
Using a second predicate device is not really advisable. The FDA only wants to see the use of one (1) predicate device, but then you can use as many "Reference Devices" as needed to support the substantial equivalence discussion. In the Executive Summary and Substantial Equivalence Discussion sections, outline the Reference Device(s) and how they further contribute to the safety and efficacy of the device. Also using another predicate device or reference device just for number of patients in a study is not really how intended to be compared. The comparison of a predicate device (and support of reference devices) is about the intended use, technology, operation, materials, specifications, and the primary point: safety and efficacy of the device type.
For the clinical trial and number of patients, this revolves around whether the number of patients sufficiently show through the clinical data that the device is safe and effective (performs). You can also "push back" from a Q-Submission meeting response if FDA is looking for a number of patients for the study. You can also submit as many patients or as little patients as needed during the actual 510(k) submission, but there is risk of not being accepted. As stated, there is no "magic number", without knowing the device type, study protocol, and a few other pieces of information it would be difficult to provide further advice. However, the main advice around determining number of patients for a trial does depend on a number of factor including statistical power, patient population, disease prevalence, how the results or clinical outcome is presented, etc. And it is possible the number of patients gets lowered in subsequent studies, because as further knowledge and experience is gained with a medical device type, this number does decrease to a point where a 510(k) submission might not even need a premarket clinical study in order to achieve clearance. Though this is not the only reason as noted above.
------------------------------
Richard Vincins ASQ-CQA, MTOPRA, RAC
Vice President Global Regulatory Affairs
------------------------------
Original Message:
Sent: 15-Nov-2022 16:04
From: Thomas Lawson
Subject: predicate device and # of patients in clinical study
This question is a tad hard to answer without a bit more information, mainly it would help to know why you think you need this new, 3rd device for your argument and whether you do indeed plan to have 2 predicate devices. But, my first reaction is that if clinical data are necessary for your submission, the sample size is driven by expectations of the outcomes and those expectations come from published studies reporting AEs (safety) or clinical benefits (efficacy) when used in a manner similar to what you plan to monitor in your study. I don't know what is meant that safety is not statistically powered-- if you need the data to support your submission, the study has to have a strong level of statistical justification for the sample size to achieve the stated outcomes, even if it is a single-arm study. Which means also that your statistician can assess this 3rd device's data in the sample size calculation, but that is not merely saying they got the indication with 30 subjects so we'll do a trial with 30 subjects as well (for example).
Your strategy needs to be proactive and be that the statistical plan calls for X subjects in order to meet performance goals, endpoints, etc and so that is why the protocol calls for X subjects. I don't think the FDA ever has a "magic number" in mind since protocol outcome goals vary widely. You should go into any meeting with a plan/protocol and ask for the FDA to concur that if you provide data from this protocol it will accept those data as the basis for your new indication submission. The only time you should push back is when your protocol is well-designed with statistical support and the FDA cannot come up with a good argument why it does not concur with your plan/protocol.
------------------------------
Thomas Lawson PhD
Principal Consultant
El Cerrito CA
United States
Original Message:
Sent: 14-Nov-2022 06:20
From: Anonymous Member
Subject: predicate device and # of patients in clinical study
This message was posted by a user wishing to remain anonymous
Hi all,
i am working on a Qsub for the expansion of indication of our cleared device.
I am puzzled concerning the choice of the secondary predicate for this submission (primary predicate = our own cleared device), between:
- the same predicate we used the first time. Pros: the job is already 95% done
- a new device (and potential predicate) that got recently clearance with a clinical trial with much lower number of patients. Cons: new uncertainties. Pros: study with lower number of patients
Both the predicate I originally used and this new device are defendable in terms of substantial equivalence. Clinical trials design are all very similar in terms of design and the primary endpoint is always safety, so the studies are not really statistically powered and I have the feeling that the number of patients is sort of a "magic number" that FDA agrees to be good enough predicate after predicate. The fact that they allowed a lower number of patients for this new device also makes me guess that FDA has lowered a little bit the bar and expectations on clinical data for this specific device regulation over time.
What are your thoughts? Is it possible to push back during the presub meeting on the number of patients, leveraging this third device as precedent case, although we do not choose it as a predicate? Any experience of pushing back based on other cases (that are not the predicate)?
Thanks in advance