*** For a limited time Diabetic Investor is offering a 20% discount for new subscribers to Diabeticinvestor.com. To take advantage of this incredible offer contact David Kliff directly at firstname.lastname@example.org with the word INCONTROL in the subject line. ***
Recently Beta Bionics, the makers of the iLet artificial pancreas, reached a significant milestone: raising $1 million through online public fundraising. Beta Bionics is also one of the few companies I’m aware of that’s structured as a public benefit corporation. With this type of corporate structure, the main goal is NOT to maximize profit. As the name implies, it’s run for the benefit of the public.
Diabetic Investor is not surprised that the company reached this milestone, as an artificial pancreas is one of the sexiest pieces of diabetes technology. It is no longer a question of if an artificial pancreas will exist, but of when it will become commercially available. With every major insulin pump company working in this area, along with many startups, patients will not just have one system, but several they can chose from.
While I applaud the advancements in this area, I remain skeptical there is a sustainable, profitable business here. This may not be a problem for Beta Bionics, but it is for everyone else. The most pressing question from a business perspective is just what patient demand will be. Yet, there are also some therapeutic questions which also need to be addressed. Before we examine these issues let’s be clear about one thing: when these systems work as designed, there is no question patients benefit from this technology.
However, what happens when these systems do not work as designed? These systems deliver insulin, which when delivered incorrectly can be — and is — lethal. Sometimes people seem to forget that an artificial pancreas is really a series of interconnected devices which are run by sophisticated algorithms. There is the insulin pump which delivers the insulin, the continuous glucose sensor which gathers the glucose data, and the algorithm that processes all this information, controlling the amount of insulin and when it is delivered. For the moment, there is also a conventional glucose monitor which is used to calibrate the continuous sensor.
The simplest way to think of an artificial pancreas is like a chain on a bicycle, a series of links, if you will. When the chain operates as intended with no broken links, or does not become decoupled from the gears of the bike, the bike runs smoothly. However, when just one link is broken, functions improperly or becomes detached from the gears, the whole system comes to a halt. Now, the makers of these artificial pancreases will say that’s what they want to happen: they want the system to basically shut down as a failsafe.
But what happens when the system processes incorrect data or malfunctions in other ways? As sophisticated as these devices are, they do fail. If you don’t believe that, check out the MAUDE database on the FDA web site.
I also believe the term “artificial pancreas” is somewhat of a misnomer, as it implies the system does everything and the patient does nothing. The patient has many interactions with the system, and as the late Al Mann used to say,“The more chances you give a patient to make a mistake, the more likely it is they will make a mistake.” What happens, for example, if the continuous sensor falls off during exercise; or what happens if the sensor is calibrated with inaccurate data? Let’s say the pump receives the correct instructions and then inaccurately executes those instructions. Frankly there are a host of possible issues, and there is no way, no matter how much thought is given to them, all can be anticipated and planned for.
The artificial pancreas is a sophisticated medical device, but it is a medical device and medical devices, fail, malfunction, break down, etc. They do not work 100% of the time with 100% accuracy.
Let’s move beyond this issue and look at another: what problem do these systems solve? The default response is that a patient using an artificial pancreas will achieve better outcomes. Well, the default response is incorrect. I could make a very strong argument that better outcomes can be achieved with existing technology. Take a “smart” insulin pen, CGM and app, connect them all up, and what is it? Effective insulin dosing without all the way-cool whiz-bang technology. A system which I would argue actually is safer, not to mention much cheaper, than an artificial pancreas.
Now if the response is that an artificial pancreas is more patient friendly and makes their lives easier, I can semi-buy that response. It all depends on how smart the system is, and how quickly it learns. Long ago I raised somewhat of a ruckus when I stated that insulin pump therapy is a lifestyle choice, not an outcomes choice; that based on numerous clinical studies, patients can and do achieve solid outcomes following multiple daily injection (MDI) therapy. If pumps really produced better outcomes, why is it that still, after more than 20 plus years, only 30% of type 1 patients use an insulin pump?
Are insulin pumps not smarter than they were 20 years ago? Are they not more sophisticated than 20 years ago? Do they not now connect with continuous sensors? Is not insurance coverage more widespread? Is it not true that physicians are more comfortable and aware of insulin pump therapy? Why, then, do the vast majority of type 1 patients and insulin using type 2s prefer not to use an insulin pump? The answer cannot simply be insurance coverage, out of pocket cost to the patient, or learning how to pump.
Could it be that some — and by all publicly available information, actually the majority — of insulin using patients just don’t want to use an insulin pump? That they feel more comfortable, perhaps safer, not being attached to a machine which delivers a lethal drug?
I am not by any means questioning the value of insulin pump therapy, or the value of an artificial pancreas. I am questioning just how many patients will use this way-cool, whiz-bang, very sexy technology. I am questioning whether there is a sustainable business model here. Diabetic Investor frankly has been shocked by some of the estimates companies are using to justify their move into the artificial pancreas arena, seemingly ignoring some very real issues and believing almost blindly, if they build it, patients will use it.
The other day Diabetic Investor had a nice conversation with a diabetes device veteran who, like me, can remember when the insulin pump market consisted of two companies, MiniMed and Disetronic, reminiscing about how back then the market was growing at 25% plus each quarter.
Fast forward to today — and how many insulin pump companies are there? How many are working on an artificial pancreas? And the insulin pump market is growing at what rate? And an artificial pancreas will change this, why?
I am not trying to take anything away from Beta Bionics, Bigfoot, Medtronic, Tandem, Insulet, Roche, or Animas (a unit of Johnson and Johnson). Nor am I in any way questioning the need for an artificial pancreas, there is a segment of the type 1 patient population that will use such a system. But after looking at this segment of the insulin pump market, and listening to all the explanations for how many patients will actually use such a system, I still cannot make a business case for the artificial pancreas.
Maybe I’m right and the market isn’t that big; maybe there’s another reason I can’t make that case. Time will tell. What’s definite is that lots of money is being used right now to produce such a system, money that could be better spent elsewhere and produce a real return on investment.
What I see happening her is almost blind faith that if it is built, patients will use it; and payors will pay for it. It seems to me that there is some fuzzy math when estimating the size of this market, and that estimates are put together to justify the amount being spent on R&D, rather than the potential return on investment.
It’s not a popular statement, but there is a reason it’s called the business of diabetes. And from a business perspective, I cannot make the case for the artificial pancreas.