An obligatory warning: This is a nerdy, regulatory blog post that gets into some of the nuances behind the FDA’s interoperability regulatory pathway, and how that interacts with Tidepool Loop’s regulatory clearance.
Now that Tidepool has been through the submission process with the FDA, we have a lot more insight into the nuances and practical realities of how the FDA’s interoperability pathways works.
In this blog post, we talk more about the details of what it means to be “interoperable,” the subtle distinction between 1st-party interoperability and 3rd-party interoperability, and what that means for past, present and future devices that make up an automated insulin delivery (AID) systems using one or more of the FDA’s interoperability pathways, known as iAGC, ACE Pump and iCGM.
Here is the TL;DR (too long, didn’t read) summary:
(Please note: This is Tidepool, not FDA terminology. We’re using the typical definitions “1st Party” to mean “from the same company”, and “3rd party” to mean “from another company.” Presumably, the “2nd party” is you, the person with diabetes… :)
Once you understand this nuance, you can see that the current FDA-cleared and popular ACE Pump / iAGC combinations commercially available are examples of 1st Party Interoperable ACE Pumps:
Here’s the sad news: There are currently no ACE Pumps on the market in the US that are compatible with Tidepool Loop.
Fortunately, here’s the good news: There are multiple additional ACE Pump Sponsors (manufacturers) who either intend to, or already have, submitted ACE Pumps to the FDA that are capable of being compatible with Tidepool Loop. We can’t yet tell you who they are, but we promise we will share the news with you as soon as we can.
Here’s a quick refresher on how all of this came to be.
In 2015, the FDA knew that there was an impending wave of closed-loop automated insulin delivery (AID) systems. In 2012, they had delivered AID guidance telling Sponsors what they expected to see in these submissions [FDA’s 2012 “artificial pancreas” guidance].
But the Agency also knew that reviewing AIDs as a system, where the submission would need to include everything about the system, was going to be untenable. Reviewing an AID system would mean that all of the components, the controller, the CGM, the pump, would need to be reviewed together each time any component was updated. It just wouldn’t scale, and it would slow down access to these life-saving and life-changing devices.
So very much to their credit, the Agency started talking to manufacturers, including Tidepool, in 2015, about their desire to make an expedited pathway for interoperable AID system components. This approach would make things easier for both industry and the FDA, allowing the components of an AID system to be reviewed independently. It would also, presumably, provide greater choice to the diabetes community, allowing people to separately choose the controller, pump and CGM that were right for them.
At this point, we should also give huge kudos to a few other organizations that helped push the ball forward on interoperability initiatives for diabetes devices at the time:
Finally, and this can not be overstated - the progress made by the #WeAreNotWaiting community, and in particular projects like Nightscout, OpenAPS, AndroidAPS and Loop, demonstrated very clearly to both industry and regulators that diabetes community was really unsatisfied by the proprietary solutions, and that innovation could come from anywhere, delivering powerful, and easier-to-use systems that reduced burden and improved outcomes.
You may be wondering: Is this “1st Party Interoperability” what the FDA intended when it created the interoperability pathways? Yes, in fact, it is. This ACE Pump pathway allows Sponsors (manufacturers) to independently iterate on their controller separate from their pump, without needing to submit an entire new submission for the pump when they update the controller algorithm.
This is a very good thing for the diabetes community: It means that companies that make an insulin pump and that have their own algorithm, like Medtronic, Insulet, and Tandem, can continue to iterate on and improve their algorithm independently from their pump if they choose to utilize the ACE Pump and iAGC regulatory pathways. Insulet and Tandem have each achieved ACE Pump and iAGC clearances.
Is this different from what most people were expecting would be enabled by the interoperability pathway? Well, we don’t know if we can answer for everyone, but it sure was not what we (Tidepool) were hoping for when the concept of the diabetes device interoperability pathways were originally introduced. Our hope was that, in exchange for the less burdensome (read: less work, faster) Class II interoperability pathway, that manufacturers would be required to demonstrate that they had a plan that would allow their ACE Pump to safely interoperate with third-party controllers. We understand and appreciate the 1st party use case, but we think that third-party interoperability is better for the diabetes community overall because it increases choice and opportunities for innovation.
If you are a regulatory nerd and want to follow along at home, the FDA is really good about publishing all of the relevant clearance decision documents, which we’ve helpfully summarized below.
The De Novo Pathway
Typically, if you have a new product that you want to bring to the FDA, the required, default pathway is a rigorous Class-III submission (a PMA, or Pre-Market Application). This usually makes sense - if it’s never been done before, there are a lot of unknowns about the product, there may be a lot of risk to using the new device, and FDA and users won’t have prior experience with the product to understand those risks well.
But sometimes a Sponsor says “Hey, we think we can show that this isn’t so risky, so instead of the default Class-III pathway, we can make a case that this product should not be classified as a Class III device.” Sometimes, the FDA encourages Sponsors to take this approach, since they know it’s less work for everyone (FDA included) and the new product really isn’t such a high risk device. Now, FDA can down-classify a device on its own, but that requires a LOT of regulatory time and effort on FDA’s part, including a public comment period, so FDA encourages the “de novo” pathway, described below, as a faster way to accomplish the same thing.
The Sponsor that does this for a new type of device submits “a de novo petition”, from the Latin meaning “from the beginning; anew”. In these cases, the de novo petition requests the FDA to create a new product category, which establishes a Class I or Class-II classification for the new device type. If FDA classifies the device as Class II, which is a less rigorous (read: easier and faster) pathway than the default, higher risk Class-III pathway, this is where those Special Controls come in; we discuss those in more detail below.
That’s what happened in 2018 when the FDA issued de novo authorization to Dexcom for the G6 as an iCGM, and in 2019 to Tandem for both the t:slim X2 ACE Pump and Control IQ Technology iAGC.
More kudos…
In addition to giving kudos to the FDA for encouraging Sponsors to use the de novo pathway, we should also pause and give a shout out to Dexcom for pursuing the iCGM de novo pathway for the Dexcom G6 (the first iCGM), and similarly to Tandem for pursuing the de novo pathway for the t:slim X2 (the first ACE Pump) and Control-IQ Technology.
These de novo petitions paved the way for all following submissions for the same kind of devices as Class -II devices. It was extra work on the part of Dexcom and Tandem to do this, and that extra work benefited the entire diabetes community.
But wait, there’s a small catch. With these de novo decisions, the FDA also laid out Special Controls for these new device types - the extra requirements that each Sponsor must conform with for their product to be classified the same as the new product.
Example: The first device of a given product type is given “de novo authorization.” If the de novo authorization says that the device is Class II, which almost always requires a 510(k) submission, then every device of the same type that comes later will then use the 510(k) pathway, which allows future device sponsors to say “my device is ‘Substantially Equivalent’ to a specific, prior device of the same type (‘the predicate device’)”. FDA may also define exactly what must be included in the 510(k) submission to demonstrate that substantial equivalence - those Special Controls.
So where are these Special Controls that you speak of for each product category? Ah, well, they are detailed in the de novo reclassification order, and ultimately in the regulation for that specific device type. The FDA also publishes a “Decision Summary” for some submissions; this is FDA’s documentation of the information they relied on to make the regulatory decision for that submission.
Without further ado, these are the FDA de novo authorizations and decision summaries that established each of the diabetes device interoperability components:
iCGM - DEN 170088 (Dexcom G6)
ACE Pump - DEN 180058 (Tandem t:slim X2)
iAGC - DEN 190034 (Tandem Control-IQ Technology)
As new devices continue to be cleared by the FDA, here are some key things to look for that indicate whether or not the sponsor has done the work to make their device 3rd party interoperable. (Emphasis added by us…)
In the Decision Summary for the Dexcom G6 (DEN170088), you’ll find this in the Indications for Use (page 2):
The Dexcom G6 System is also intended to autonomously communicate with digitally connected devices, including automated insulin dosing (AID) systems. The Dexcom G6 System can be used alone or in conjunction with these digitally connected medical devices for the purpose of managing diabetes.
and later on page 5:
The Dexcom G6 System is an interoperable connected device that can communicate glucose readings and other information wirelessly and securely to and from interoperable electronic interfaces; including compatible AID systems. The G6 CGM system is designed to communicate with interoperable devices in several ways, such as described below:
and importantly on page 36:
Interoperability:
A plan and approach for interoperability were provided according to the FDA Guidance “Design Considerations and Pre-market Submission Recommendations for Interoperable Medical Devices - Guidance for Industry and Food and Drug Administration Staff” and determined to be adequate to support and clearly specify expectations, requirements, and interface specifications to potential interoperable devices.
In addition, their plan covered their approach to working with connected device companies regarding contractual approaches, interfaces for data communication and exchange, and post-market reporting procedures and responsibilities (e.g., who is responsible for investigating and reporting complaints, malfunctions, and adverse events).
We see the same language in the ACE Pump de novo for the Tandem t:slim X2 (DEN180058) we see (on page 13):
Interoperability:
A plan and approach for interoperability were provided according to the FDA Guidance “Design Considerations and Pre-market Submission Recommendations for Interoperable Medical Devices - Guidance for Industry and Food and Drug Administration Staff” and determined to be adequate to support and clearly specify expectations, requirements, and interface specifications to potential interoperable devices.
In addition, their plan covered their approach to working with connected device companies regarding contractual approaches, interfaces for data communication and exchange, and post-market reporting procedures and responsibilities (e.g., who is responsible for investigating and reporting complaints, malfunctions, and adverse events).
and on page 16 (in the Benefit/Risk section):
… The t:slim X2 insulin pump with interoperable technology pump could also be incorporated into a system with a AID controller algorithm. The use of this device in conjunction with an interoperable AID controller algorithm and a continuous glucose monitoring device could allow users to receive the benefits associated with closed loop insulin therapy, for example as described in the Summary of Safety and Effectiveness Data for P160017. The design of this device as enabled for external control is anticipated to facilitate innovation in digitally connecting compatible devices together to meet user needs, and is expected [to] increase the safe and effective therapeutic choices commercially available to patients using insulin pump therapy.
The uncertainty of the benefits and the risks associated with the use of this device is reduced by the special controls including requirements for device design, which includes predetermined communication specifications and validation plans that potential interoperable devices must meet before this device may be used in a combination with those components. It is additionally reduced by device labeling, which clearly describes pump performance parameters, so that developers of potential connected devices (including AID algorithms) may determine whether this device is capable of meeting the needs of any potential system which might incorporate it as an interoperable component. The interoperability plans incorporated into the design of this insulin pump, which serves to define the requirements for devices it would be interoperable with, sufficiently addresses these uncertainties and adequately assures that the uncertainty related to anticipated benefits is acceptable for its intended use and intended use population.
The Omnipod 5 ACE Pump Decision Summary has similar language:
Interoperability:
A plan and approach for interoperability were provided according to the FDA Guidance “Design Considerations and Pre-market Submission Recommendations for Interoperable Medical Devices - Guidance for Industry and Food and Drug Administration Staff” and found to be adequate to support and clearly specify expectations, requirements, and interface specifications to potential interoperable devices.
In addition, the plans covered Insulet’s approach to working with third party manufacturers of digitally connected devices regarding contractual issues, interfaces for data communication and exchange, and post-market reporting procedures and responsibilities (e.g., who is responsible for investigating and reporting complaints, malfunctions, and adverse events).
This language is presumably in reference to integration with iCGM devices; the Omnipod 5 ACE Pump is compatible with the Dexcom G6 iCGM. The Omnipod 5 ACE Pump decision summary does not include similar language to the Tandem t:slim X2 decision summary regarding “developers of potential connected devices (including AID algorithms)”.
However, the Omnipod 5 510(k) decision order (K203768) includes in the Device Description on page 5:
Currently, the Omnipod 5 ACE Pump is compatible with the Omnipod 5 iAGC, whose software is pre-installed on the Pod and the App. Future alternate controllers (iAGCs) may be established for use with the Pod, in which case the software modules of the Omnipod 5 iAGC would be disabled.
To our friends at Tandem and Insulet, we look forward to a future in which third party alternate controllers (iAGC) like Tidepool Loop are compatible with your awesome ACE Pumps!
If you made it this far, congratulations! You’ve definitely earned your Regulatory Nerd Merit Badge!
Hopefully this added detail helps you better understand the current situation regarding interoperability between iCGM devices (like the Dexcom G6), ACE Pumps (like the Tandem X2 and Omnipod 5), and iAGC devices (like Tandem’s Control IQ, Insulet’s SmartAdjust, and Tidepool Loop).
Here at Tidepool, we are very proud that Tidepool Loop is a truly interoperable iAGC device, designed from the ground up to enable interoperability with both third party iCGM devices and third party ACE Pumps, including, as it says in the Tidepool Loop 510(k) decision summary:
Interoperability:
A plan and approach for interoperability were provided according to the FDA Guidance “Design Considerations and Pre-market Submission Recommendations for Interoperable Medical Devices - Guidance for Industry and Food and Drug Administration Staff”. The plan specified expectations, requirements, and interface specifications for potential interoperable devices. In addition, the plan covered Tidepool’s approach to working with connected device companies regarding contractual issues, interfaces for data communication and exchange, and post-market reporting procedures and responsibilities (e.g., who is responsible for investigating and reporting complaints, malfunctions, and adverse events).
These plans mean that over time, Tidepool can continue to add new iCGM devices AND new ACE Pump devices to Tidepool Loop, and that no new FDA filing is required! We think that’s exactly how interoperability should work.
Regarding the FDA’s goals for interoperability, we think Dr. Courtney Lias, former branch chief for the diabetes device group at the FDA said it well:
“So what we want to do is incentivize manufacturers to open up systems so that you don’t have only closed systems available to patients so that healthcare providers and patients can discuss the specific apps, the specific CGMs, the specific pumps, the specific glucose meters that works best for that patient and the potential to mix and match these different devices we believe will be better for patients.”
Dr. Courtney Lias, FDASource: MedCity News “When it comes to diabetes devices and apps, FDA prefers mixing and matching”
Cheers!
Howard & The Tidepool Team
We'll keep you updated with all the need to know news and announcements for Tidepool and Tidepool Loop
Donate today to support Tidepool's mission to improve the lives of People with Diabetes through affordable, accessible and interoperable technologies
Want to get the latest news and updates about Tidepool and Tidepool Loop? We promise to keep it interesting in your inbox.
