The FDA has just finalized rules about what's going to be regulated when it comes to medical data sharing tools and mobile apps, and those "in-the-know" in the diabetes world are pretty darn excited about it. They’re using phrases like:
"Landmark." "Milestone." "A victory for people with diabetes." "Just the beginning."
The enthusiasm stems from the fact that FDA is clearly trying to pave the way for open systems and innovation. They appear to have really heard the message from our D-community’s public comments on the draft guidance last year.
They’re officially taking a "hands off" approach to any apps not considered to have a medical function -- beyond tracking, organizing and viewing data. Fortunately, much of mobile med tech falls into this "low-risk" category and therefore won’t warrant overly critical regulations or review.
The agency will still mandate oversight on apps that do perform a medical function, such as controlling an existing medical device, or “transforming a mobile platform into a regulated medical device.”
But FDA experts themselves say the guidance “will allow system developers to improve interoperability with other devices” and “will create an impetus for the development of new technologies to better use and display... data.”
Of course, some key diabetes advocates have been critical in pushing FDA to adopt this approach, dating back to even years before the #WeAreNotWaiting movement emerged.
The new final Medical Device Data Systems (MDDS) guidance has been released in a pair of documents -- a 44-page report and an 8-page brief – which are a little... tough to understand, if you aren't hip to all that's going on. So we reached out to some of those key advocates who know what they're talking about, and below are their reactions.
Note also that FDA is holding a public webinar for “industry and other interested stakeholders” to discuss the new guidance on Feb. 24 at 2pm Eastern. Prior registration is not necessary.
Anna McCollister-Slipp, a longtime type 1, data entrepreneur, and FDA patient advocate, says:
This is good news for us … incorporating the feedback we all provided (on) the draft MDDS guidance from this past summer.
This is part of a cluster of things we've been advocating for, for quite some time. Really, November 2012 at the DiabetesMine Innovation Summit was when things started changing. So much of this movement started then, and there's been a spontaneous order to all of this once we found each other and started connecting.
We played a very big role in this, going back to my participation/appointment to the (FDASIA Workgroup (Food and Drug Administration Safety Innovation Act) for ONC (Office of the National Coordinator for Health IT). This is part of what I was pushing for as part of that process.
When the draft guidance came through, we pulled together a group of people from the DOC (Diabetes Online Community) to talk with FDA, got the Nightscout/CGM in the Cloud crowd to participate, got other Diabetes Advocates members to chime in, and so on. The number of comments for this guidance were about 50-fold that of the original rulemaking four years earlier.
Changing these refs is a great example of how the DOC came together through informal networks to do things that are important to us, and this wasn't driven by ADA or JDRF. This was the DOC, and this is a story about what the community has done together.
Howard Look, D-Dad and CEO of non-profit cloud developer Tidepool, says:
In short, this is pretty awesome. The FDA continues be extremely pragmatic and forward-thinking. They recognize that the mobile and digital health revolution is here, and that minimizing the regulatory burden for low-risk applications and devices is best for patients.
The FDA clearly listened to feedback from the community. There are specific recommendations from both Bennet and my comment documents, which were taken up by the CGM in the Cloud group, that made it into the final guidance. It's so wonderful to see how engaged the FDA is with the diabetes community. It's clear that voices from Nightscout, CGM in the Cloud, Tidepool and the rest of the #WeAreNotWaiting community are being heard. Dexcom also gets huge credit... see below.
I’ve highlighted what I think is the important stuff in the document. In short:
- No controlling devices (no surprise there)
- It's all about INTENDED USE. If you say your product is intended for active monitoring, it doesn’t fall under MDDS review
- "Active Patient Monitoring" is NOT considered MDDS. But it specifically says that showing the most recent blood glucose values is OK, including showing the delay time from the last value: "remotely displaying information such as the most recent blood glucose value or time-lapse between blood glucose measurements is not considered active patient monitoring."
- Anything with ALARMS is considered "active patient monitoring" and not MDDS: "devices that provide active patient monitoring include: ... A device that receives and/or displays ... alarms or alerts ... and is intended to alert a caregiver to take an immediate clinical action"
It’s important to note that the regulatory team at Dexcom, led by Andy Balo, has done a TON of work that we will all now benefit from: A year ago, Dexcom Studio and Dexcom SHARE were both considered accessories to a Class III medical device, and therefore would require an FDA PMA (pre-market approval) submission.
Last year, Dexcom filed and got approved a “de novo” application that reclassified Dexcom Studio software as a Class I/Exempt device, so requiring only general controls.
More recently, Dexcom filed and got approved another de novo that reclassified active monitoring on a secondary display (the new Dexcom SHARE app) as Class II/Exempt. That means that similar devices/software do NOT need to file a 510K (though they are still considered to be at Class II risk). From a regulatory documentation standpoint, that is no different from a Class I device general controls – essentially making the whole process easier for all.
Ben West of the Nightscout Foundation says:
I'm currently working on applying these new regulations to Nightscout, and working closely with the FDA, often connecting several times per week, to get this done. They've been really great to work with, super-sharp, responsive, fast, and practical. Contrary to the usual story, instead of delaying things, they seem to be joining the #WeAreNotWaiting movement, at least in-so-much as a regulator can.
In the short-term, it's a huge landmark that FDA has lowered the burden, and repeatedly indicated a desire to work with us. Long-term, this is another interim step in the right direction towards restoring "fidelity" to our therapy.
Long-term, I don't think anyone really cares about monitoring or numbers; these are mere tools we use to better obtain the accurate or "right" amounts of insulin. We still need better ways to properly administer insulin, free from side effects.
Nightscout's R&D is clearly headed towards tools to help with that, including #OpenAPs and other "advanced" projects. Getting Nightscout in its current form approved by FDA could help open the doors for these other projects to flourish quickly.
Once we get our process registered and approved for FDA, we will be able to augment and extend that work to also get other projects approved. My hope is that this will provide a healthy forum for the diabetes hacking community to work with FDA, as well as for innovations in therapy, especially dosing insulin properly, to receive research and an audience in the "market."
FDA has asked about the possibility of accelerating the pace of Nightscout towards approval. Honestly, the biggest thing I'm worried about it the medical device tax.
Going forward, in terms of wider DOC and FDA interactions, we can expect more workshops centered around bolus wizards and closing the loop in a variety of ways.
I recommend reading Nancy Leveson's Engineering a Safer World to get an idea of how this might play out. Some practical questions perhaps include: do we want an "eating soon” mode? (instead of anticipating unknown carbs, simply correct to the lower-end of safe range) or dynamic basal (using CGM feedback to automatically reduce/increase temporary basal rates within safe constraints)? These are a little bit more aggressive, but also much more practical than some of the features we see slowly coming on the market now.
Beyond the Guidance
This all comes at a time when the Diabetes Community is awash in news about new mobile health tools, which sets the stage for further tech integrations that we can only dream of. For example, just this past week, we also heard more about how Dexcom's working with Apple on a smartwatch to track glucose.
The possibilities seem endless… so huge thanks to FDA for moving forward with progressive thinking on mobile health! We can't wait to see what the innovators and designers make happen now that they have the ability to do so.