I could call today’s guest post a “straggler” from my Summer Reading series, but that might downplay its impact. Good things take time, as noted today by Samantha Katz, who along with a fellow graduate student at Northwestern University, won the the $10K Grand Prize in the 2009 DiabetesMine Design Challenge. 
She was subsequently hired by Medtronic as a Global Product Manager for their evolving insulin pump systems (! – and she served as a judge in the 2010 Design Challenge). Now that she’s an “industry insider,” I asked Samantha to comment on the state of new insulin pump technology — and I found out that’s she’s learned a few things:
A Guest Post by Samantha Katz, of Medtronic Diabetes
I have to admit that I was pretty naïve when I won the 2009 DiabetesMine Design Challenge. The concept of an iPhone – which already contained so many great technologies like GPS, Bluetooth, and Monacle Augmented Reality – controlling someone’s insulin pump and even continuous glucose monitoring system seemed pretty simple to me. I assumed the reason something like it was not yet on the market was design-related, and I figured the LifeCase & LifeApp System would be on the market in, say, two or three years, right?
Wrong. It has been over a year now, and for various reasons personal to me and Eric LifeCase & LifeApp is no closer to fruition. But even if we had chosen to pursue developing the system and somehow had corralled enough funding and talent to help us, there is no way we could have the product on the market in such a short amount of time.
That’s because insulin pumps are complicated machines that safeguard health. They are extraordinarily complex to develop, especially if they have integrated continuous glucose monitoring, and no matter how many colors they come in or how flashy their marketing is they will never be pure consumer devices. When working in Medtronic’s booth at ADA this past June, I would often hear impatient chatter about Medtronic’s patch delivery system or Dexcom’s delayed integration with Animas and OmniPod. All I wanted to do was remind people that this is not your PDA or iPod we are talking about. An insulin pump is a device that keeps you alive.
Because of that, insulin pumps also benefit from the oversight of the FDA and other regulatory authorities around the world. I say “benefit” because without regulation there would be little trust between patients and physicians and pump manufacturers, and that trust is what motivates physicians to prescribe pumps and entrust us with their patients. This trust also keeps people with diabetes on insulin pump therapy, which, according to several studies, is the best therapy for managing diabetes.
However, the FDA has very strict guidelines regarding the clinical data that is required before a device is approved. At Medtronic, safety is our top priority and therefore we have been working closely with the FDA to launch the most technologically advanced, safest pumps possible. I truly wish I could put every last bit of technology in our insulin pumps because I know that someone out there could benefit, but sometimes that is not feasible until the required data is available.
While this takes time, there are several signs that our industry is making strides. When taking a break from booth duty at ADA, I had the chance to check out AgaMatrix’s WaveSense Direct Connect Cable and Debiotech’s Jewel Pump and Android app. I am thrilled to see these advancements in the vein of my LifeCase & LifeApp System, and I know that small steps with the FDA in the short term will set the precedent for further advancement in the long term. All of the diabetes device manufacturers are working feverishly to automate insulin delivery and ultimately close the loop, and I believe that this sense of urgency sprinkled with a healthy respect for impeccable safety will get us there – perhaps it will just take a little longer than my initial estimate of two or three years.
For more on iPhone integration, see this neat Timeline: the iPhone as a Medical Tool from MobiHealthNews.
 Strides){kind=link}
A good point, that it takes a lot longer to successfully develop something that is life-saving, vs. something that is convenient/neat/a way to pass the time. Thanks for taking the time to write this post.
As a sw eng, I have always viewed this whole line of controlling pumps with a consumer device ridiculous. These devices are not built with any of the same safety, reliability, battery life, etc requirements.
An open platform is fundamentally unsafe. What happens when your latest greatest app has a memory leak and corrupts your pump sw? What unrelated sw should be loaded during testing? Which ones should be running? What about firmware updates to the phone? What about when the battery runs out? The list is endless.
And who is responsible for all this testing? Not the FDA. So is it the developers, and federal agencies and consumers should just trust them?
A reliable, purpose-built, closed system, that can be properly and fully tested, with minimal functionality, and very limited ((no) sw/fw upgrades is the only solution.
The rest is hype.
“All of the diabetes device manufacturers are working feverishly to automate insulin delivery and ultimately close the loop….”
Music to my ears – hope it’s true! Thanks, Samantha!
I think FDA is right to ask for testing of unproven diabetes technologies, but I am not nearly as sanguine; the agency seems to want zero risk, and that is simply not possible. To wit – the Veo is approved in the EU and we’re said to be still two years away in the US, where so many patients would be helped by a device that could reduce severe hypoglycemia. Better, simpler, more powerful technologies will serve patients and public health well and while safety is paramount, we also don’t want to see innovation stifled, given how much help patients still need. Thank you for the column!
To MG: It’s not that ridiculous! Most insulin pumps can be controlled on the pump itself, so if your iphone stops working as a controller you can always just take the pump out of your pocket. I’ve left my one-touch Ping controller at home before and while that left me without a Glucose tester, I was in no danger and was able to control the pump. Plus, what does my ping controller actually do for me? It reads my glucose and suggests a bolus. Then I have to dial up a dosage manually anyway. Are you saying an integrated iPhone case/glucose tester can’t do that safely? Even if it was totally one-way communication from the pump to the iphone (so the phone doesn’t actually control the pump, just receives info from it), this system would be light-years better than my crappy Ping controller.
All your hypotheticals about testing seem to assume that the entire insulin delivery system would fail if the controller failed. That’s ridiculous!
Your “solution” of “a closed system, with minimal functionality, and very limited ((no) sw/fw upgrades” is insulting and depressing. How about an OPEN system, with TONS of functionality, and REGULAR software updates? Just make it the secondary controller in the system. Why can’t the FDA and people like you wrap your head around this idea and allow it to happen?
Todd said: “Why can?t the FDA and people like you wrap your head around this idea and allow it to happen?”
Well, because if a “secondary controller” has the ability to dose insulin, if has to be rigorously tested. Failure doesn’t just mean leaving it at home or running over it with your car, it also means being effected by EM radiation near a microwave or some other intentional or unintentional interaction between applications on the device. That’s just to get started on the reasons why the 2009 prize winner was a pipe dream, albeit a creative one.
I’m just as frustrated as anyone else by the slow pace of innovation with insulin pumps, if you can even call it innovation compared to what I’m used to in silicon valley. But I don’t think we should be so willing to abandon the safety net provided by the FDA. They don’t have a perfect system, but I wouldn’t trust Apple (let alone a 3rd party app developer) or anyone else with my life without some serious regulatory body as a gate keeper.
Pking:
Good point. But we could still move forward with my other suggestion of having an iPhone accessory that just RECEIVED data from the pump and didn’t actually control it. In fact, if I had a case that served as a glucose meter and came with some innovative software for tracking (and didn’t talk to the pump at all) I’d still be way better off than I am now. I know the FDA has to regulate meters as well, but it seems like an established meter manufacturer could pull this off relatively quickly.
It is certainly good to finally see the promising use of technology for more worthwhile purposes than just the leisure market. It may not be as profitable for the makers, but surely more use of these advancements could be produced and developed more timely.
Many insulin pump makers are talking about eventually offering iPhone or Android solutions to control the pump (including http://www.pancreum.com), but the FDA is going to make it very tough for them.