When creating a 510k, which Guidance Document Wins?

[SSchoepel]

We are getting ready to put together our 510k and I have been pointed to several guidance documents. We have a class II PACS we are developing. It is software-only. We are going to submit as a moderate level of concern.

I've got these two that are sort of "competing":

Guidance for Industry: Guidance for the Submission Of Premarket Notifications for Medical Image Management Devices

(broken link removed)

(This ones says software is of a minor concern but we aren't taking that to mean it's a minor concern product. And, that's not my issue.)

Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices

(broken link removed)


The FDA makes the comment about "the least burdensome approach" in many of their documents. The first guidance above would give me the least burdensome approach when documenting our software for submission. And, it is most specific to our product. The other one is more general and requires more information.

Which one wins? The product-type specific one or the more general one?

Thank you,
S. Schoepel

 

[MIREGMGR]

Identify the reviewer, or the head of the reviewer group, and ask them.

My experience is that reviewers have broad latitude to choose among competing rules. A senior reviewer or a reviewer group leader may even go off toward their personally preferred interpretation that can only be discerned to be required by having their knowledge of multiple rules, precedents and whatever else they think is relevant. A junior reviewer probably will tend to choose what they think is the most conservative, CYA approach so that they can't later be second guessed. So, ask up front what they're going to want.

 

[SSchoepel]

I got back a very brief message from the reviewer who said to follow the PACS guidance but note that the software guidance changes the level of concern to moderate. And, that this is a common mistake that companies make in not submitting a proper 510k.

Which means I could interpret his remarks either way: follow the PACS or follow the software guidance.

Regards,
S. Schoepel

 

[MIREGMGR]

Is it a junior reviewer? If so, is that person hinting that you should do both, so that the reviewer can comfortably note that there is no territory that you haven't covered, that the reviewer's supervisor later could criticize the reviewer for not having gigged?

If on the other hand it's a senior reviewer, keep asking questions until you understand what they want.

 

[SSchoepel]

It is a third-party reviewer who has been "in the game" for at least a decade. But he did state that his expectations are similar to the FDA...I wonder if I can call the FDA.

 

[MIREGMGR]

I wonder if I can call the FDA.

Why are you using a third party if there's uncertainty as to what is required/expected by that third party and they aren't clearly telling you?

If it's going to be a guessing game (i.e. "his expectations are similar to the FDA", but he's not clarifying exactly what those expectations are...?), you'll lose any benefit of speed, which I would think would be the reason for going to a third party.

Yes, you can talk to FDA. Start with DICE, which used to be DSMICA. They may not have an answer for you, but they should be able to point you to someone that can help you figure out who your FDA reviewer group will be. Then you can talk to them directly.

Some of course will be more helpful than others, but it's still the most authoritative way to go.

 

[robertjbeck]

You should follow Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices, but make the software moderate level of concern when following this guidance document. it makes a difference in the documentation you'll put in the 510(k). I think this is what your third-party reviewer meant and it's correct, regardless of what the PACs guidance states and regardless of what you think "least burdensome" means.

 

[SSchoepel]

I agree. But, I'd like to challenge you as to why it's the way to go "regardless of what the PACs guidance states." Since I will get challenged, I'm curious to see your reasoning in case there is another angle that I hadn't considered.

Thanks.

 

[robertjbeck]

I think you can find this type of answer elsewhere in elsmar .. my reason is simple: FDA does not generally accept a minor level of concern. I can't say I blame them because the documentation required (based on that guidance) for a minor level of concern is not adequate for anyone to assess the safety and effectiveness of the software.

A second reason is that the documentation required for a moderate level of concern should not be a giant problem, unless the company is a mess. the big sticking point is often requirements specifications vs. stories or use cases. another is traceability. I've seen companies where traceability consisted of reading a several hundred page report to find the tests that had failed. based on my experience, when companies do not track this well, there will be problems with the software that were missed.

 

[SSchoepel]

Point 1: Yup, we already know we're going for moderate level of concern so it shouldn't be a surprise. I just think it's more than some of the people remember having to do the last time they started a new medical device years ago.

Point 2: Since it's a startup ALL the documentation has to be created and linked, so everyone's hoping that some of it could be done later instead of sooner.

Thank you for your replies. It's always helpful to double-check the logic.