A look ahead

Yes the diabetes device world is still buzzing about the approval of Dexcom’s (NASDAQ: DXCM) G6 system and money continues to flow into the space as evidenced by Bigfoot’s latest funding round. Equally exciting is the buzz surrounding just how impactful all these new way cool whiz bang insulin dosing algorithms are becoming. While it may have seemed like a pipe dream only a few short years ago it is now possible to imagine a world when patients will not just have a closed loop insulin delivery system but one that does not require a patient to count carbs.

At the same time its also encouraging that we are on the cusp of redefining what is and what is not good control. For ions HbA1c has been the gold standard for measuring control and with good reason it’s a very easy number to understand, 7 or above bad 7 or below good. Yet thanks to CGM technology and the excellent work done by people like Dr. Irl Hirsch we now know that time in range is equally important as A1c. We have no idea how this will play out but the good news is we may finally have different definitions of good control for different patient groups. This may be a little confusing at first but on balance we see this as very good thing.

Lost in all this excitement is another hidden benefit of all these way cool whiz bang cloud enabled insulin delivery systems in particular insulin pumps. Now we all know that sensor augmented systems are becoming the standard for insulin pumps. It goes without saying given all the great algorithms out there hybrid closed loop systems will be the follow on standard. These systems offer more benefits than just great control, benefits that will make the business of insulin pumps easier and less labor intensive. Simply put in the not so distant future it’s going to get a lot cheaper to run an insulin pump company, allow us to explain why.

1. Training will become simple. With the system doing all the heavy lifting all the patient needs to know is how to fill a reservoir (something that should be replaced by pre-filled reservoirs) and how to insert the infusion set and sensor.

2. Support will be proactive rather than reactive. Today patients call AFTER a problem occurs while in the future it’s possible to imagine a system not just sending glucose data to the cloud but also hardware diagnostic data too. The system could detect problems before they happen and with the newer pumps send instructions via the cloud to fix the problem or automatically send hardware before the old hardware goes bad.

3. Support will be automated. Yes pump companies do offer automatic re-order of pump supplies now but in the future we can see a system that knows when a patient needs supplies and just ships what’s needed. The patient won’t have to do much other than cover whatever their co-payment is. No apps, no web sites and most importantly no calls to customer service.

4. Hardware updates will be replaced by software updates. Tandem (NASDAQ: TNDM) is doing this already and will be followed by Medtronic (NYSE: MDT) and Insulet (NASDAQ: PODD). One of the most expensive items for any pump company is the cost to upgrade patients to new hardware platforms. Something Medtronic will face when the replacement for the 670G is ready. Hardware is very expensive to replace, software on the other hand can be simply downloaded just like an app on a smartphone.

Now comes the bad news for insulin pump companies these technological advancements will drive costs lower, much lower. As we have been stating with these new algorithms the most important elements of a closed system are the algorithm, a realizable and accurate CGM plus a pump. The pump basically just does what it’s told to do. As we noted the other it really doesn’t matter which pump it is the hardware is just a commodity.

This will in turn create a new business model for pump companies as supply revenue will replace hardware revenue as the main profit center. We can now imagine a pump company giving away the hardware for FREE and existing solely on pump supply revenue. It goes without saying that as the market leader Medtronic could easily transition to this model. This model becomes more problematic for Tandem given the size of their installed user base. Insulet has a bigger issue given they don’t sell supplies. Insulet would benefit from lower support costs but without revenue from the continual sale of pods they cease generating revenue.

This scenario also brings into play another possible business model which is to go 100% at risk and earn revenue based on outcomes. Theoretically a company could walk into a payer with a simple proposition; give us all your insulin pump patients and you only pay us when we achieve these outcomes. Again this is a very costly way of doing business however there is no reason the data generated by these systems cannot be shared with the payor. Rather than get paid for selling hardware and pump supplies the company could be paid based on metrics like HbA1c, time in range, etc.

A third scenario which seems more plausible would be to split the difference. The pump company agrees to an all in price for the hardware and supplies and then earn bonuses based on these outcome metrics. This would please payors as it would eliminate the huge upfront expense of a new piece of hardware. As we noted the other day one reason payors hate new pump patients is the initial start up cost.

No matter which revenue model a company chooses we believe its not a question of if the model will change but when it will change. Thanks to advanced technology many of the costly support functions associated with running an insulin pump company can either be eliminated or drastically reduced. Which is not unlike what happened in banking, when was the last time anyone did business with a bank teller. Now with Uber and Lyft when was the last time anyone called for a taxi. There are other examples but the simple point is technology can eliminate the need for costly humans.

Technology does not get sick, it does not take days off, demand healthcare coverage or pay raises. And the fact is us humans while we may not like it all the time we are pretty good at adapting to technology.