I be, you be, we all be IOB… now I need to pee… (sorry, rhymes tend to run amok)

Now back on Day Two of this little project I mentioned that the Snap pump lets you choose how you want it to track insulin on board, called IOB, or sometimes “active insulin.” And I suggested that you should all send thank you cards to Asante for doing this, and promised to talk in more detail about this feature later. So today is later! But before we get into the details, we need to spend a minute making sure everyone understands what IOB is all about. And it’s all about pancakes.

More specifically, it’s about stacks of pancakes.

Yep, that American gut-busting tradition of piling good things on top of each other is the perfect analogy for one of the most wicked risks of modern insulins: Stacking. Here’s the deal: all of the ’Logs—Humalog, Novolog, Apridralog (OK, I took the liberty of renaming that last one, but it is an analogue insulin)—are classified as rapid-acting insulins.

Hah! Rapid, my ass. Twenty minute onset, two hour peak, four-to-six hour duration of action: What’s rapid about that? Nothing at all. They only get the title by being the fastest of the slow. The insulins that preceded the analogues were called short-acting. ’Logs are faster than short, and so the get the honorary title of rapid, but don’t make the mistake of thinking they come anywhere close to the speed of real insulin in the body of a sugar-normal. Or anywhere close to the speed we dFolk need.

But back to the pancakes. Even if you ate a whole stack of pancakes, it’s likely that you’ll be feeding your face again sometime in the next 4-6 hours. Which also means you’ll take more insulin (at least you’d better had). The problem is, some of the insulin from that first stack of pancakes is still in your system, still active, still able to lower blood sugar. If you test your blood sugar and take insulin for your sugar and a second round of pancakes without considering that some of the earlier insulin is still in play, you can stack the insulin doses, and end up with too much insulin in your body.

Insulin stacking causes spectacular hypos.

The other classic example of stacking crashes is someone checking their BGL and saying, “Wow, I’m too dern high”—forgetting that they took 10 units three hours before—and throwing more fuel on the fire with an hour or more of old insulin action yet to play out.

Traditionally, using syringes or pens, there’s no good way to keep track of when you dosed, and how much insulin may still be in play. [Modern shooters can now use the marvelous RapidCalc Ap on a smart device.] But a pump has the computing power to track this insulin, and take the insulin on board into consideration when crunching the numbers and recommending an insulin dose.

“Wonderful,” you say. “So what’s the problem?” The problem is that no one agrees on the best way to track the active insulin in our bodies. Now the world has changed, but back in the day there were two very different ways to look at IOB, and they were called the Medtronic Way and the Cozmo Way. Medtronic pumps traditionally only counted correction bolus insulin on the theory that any insulin given for food would be perfectly cancelled by our superior carb-counting skills. The Cozmo way realized that was ridiculous, none of us really knows what the fuck we’re doing, and counted all insulin. I confess to being a fan of the Cozmo way of looking at IOB.

In today’s world, OmniPod’s blue-controller uses the Med-T way, as does the Ping. Ironically, the new Med-T pumps no longer use the classic Med-T way. The t:slim uses the Cozmo way, as does, as I understand it, the new black-controller OmniPod pumps. To be honest, I’m not sure how the “new” Roche pump calculates IOB. I tend to forget it’s out there. Have any of you ever seen one in the wild?

I digress. In truth, there are plusses and minuses to either method. And as long as you understand what your system is doing, and it makes sense to you, you’ll be fine. That said, from a clinical perspective, I’ve seen a lot more pumpers get into trouble with the Med-T-type IOB systems than with the Cozmo-type systems. In my view, playing with insulin is playing with fire. And if I’m playing with fire, I really want to see all the fire—not just half of it.

When OmniPod first came out, I was disappointed to learn that they had chosen the Med-T method of IOB tracking. When one of their senior folks came out to set me up to be one of their pump trainers, I asked why they did that. I don’t know if it’s true or not, but the answer I was given was that they polled a bunch of endos and 51% of them preferred the Med-T method. If it was that close, why didn’t you just make it a menu option? I asked.

She just shrugged one shoulder.

Well, the good folks at Asante did just that. In the menu, the choices are called Correction or All. In the Snap pump, this menu is buried in the Smart Bolus setup menu. Go Home>Menu>Bolus Menu>Bolus Setup>Smart Setup>IOB Mode. Getting there takes a lot more button pushes than it looks like, as you have to skip over many other options to reach home plate. But I have no qualms about how deeply it’s buried in the menus because it’s unlikely you’ll ever change it once you set it the way you believe IOB should be tracked.

Just to be clear: Correction = Med-T way; All = Cozmo way.

In addition to the differences in how the pumps of yesterday tracked IOB, there were also big differences in how they reported this tracking to the user. On older Med-T pumps, a lot of this happened under the hood. You couldn’t see it, couldn’t get to it, and couldn’t interact with it. Even on today’s Revel pump, viewing the IOB is no easy task. On the old Cozmo pumps, on the other hand, IOB was center-stage. There was a IOB status screen, and you could even leave the IOB screen as your default when you turned on the pump!

Tandem took a feather from Cozmo’s hat and even upped that ante: the slick Applesque t:slim really shines amongst other modern pumps when it comes to IOB. You don’t even have to fully turn on the pump to see your IOB! Just run your finger across the top of the pump to press the only analog button on the machine, and the unlock screen appears. At the bottom of the unlock screen (and on the main menu as well) is your IOB in a little status banner along the bottom of the screen, both the units in play, and the time remaining.

So how does Snap compare? Not very favorably. Viewing your IOB on the Snap is no snap. It takes three clicks: One to wake the pump, then the right button to get to Status, then the right button again to view your IOB.

 Uh…. Wait a sec. Something is missing here... I see how much insulin is in play, but I don’t see how long it’s going to last. Most of us IOB fans are used to thinking about both volume and duration. Let me give you an example. Let’s say I check my blood sugar and it’s 197 mg/dL. I look at my IOB and see that there’s 1.4 units on board that will last for another 45 minutes. OK, I’m probably safe from a hypo. I might even end up a bit high by the time the inulin runs its course. On the other hand, what if my blood sugar was 109, instead? Oh, crap! I’m gonna crash!

Naturally, I grilled the Snappers about why this key piece of critical insulin guidance data is missing. Their engineers seemed amazed that any of us would care. They viewed it as a house of cards, and pointed out that most pumps reset the countdown if even a tiny amount of insulin is added to the pile. One of them challenged me this way: What if you had, say, 4 units on board with only 30 minutes left? Then you added a quarter of a unit. Now the IOB time re-sets to four hours. It will show you have 4.25 units running over four hours, which is misleading. The large volume is gone in half an hour. Only the little splat left will last longer.

He’s probably right, damn it. But I still don’t like it.

I’m used to making decisions with two facts: volume and time remaining; and I’m getting old enough that I’m getting stuck in my ways and I don’t want to have to change, damn it! Intellectually, I get it. I get that if one variable is misleading, my decisions won’t be the best. But in the trenches, having some indication of time as well as volume is a big help. It puts my IOB into a useful perspective.

The engineers—who clearly don’t have diabetes—also thought it wasn’t that big a deal to check your bolus history if you wanted to know how much came from when. Are they kidding? Who has time for that??!

For me, a common nocturnal ritual (along with brushing my teeth, kissing the boy goodnight, and making sure my alarm clock is set) is to think about my IOB and the time remaining from my dinner bolus while analyzing the risk/benefit ratio of a bedtime correction bolus. Yes, the pump should be able to tell me if a correction is safe or not, but I know things the pump doesn’t know. I also have CGM on my team. So not only do I know my blood sugar, I know my trend. Am I stable? Am I falling? Am I rising? If I combine that knowledge with how much insulin is on board and how long it will last, I’m smarter than my pump. If I’m drifting upwards and the pump thinks a correction is a bad idea, I’ll override it and give myself more insulin anyway. I know I need it. Other times, Mr. Pump might think I could use a correction based on blood sugar and IOB, but I know better because my CGM trace lines are gliding smartly downwards.

The bottom line here is that time does matter. It’s an integral part of the IOB decision-making process in the trenches, where real diabetes therapy decisions take place. While it’s true that a pile of little boli would mess up the true accuracy of a typical IOB timer, I suspect that most of us pumpers don’t really take boli all that often in the first place. If I’m right about that, the conventional timers work fine for most people, and are certainly better than nothing, even if they are a technical house of cards.

But I’m willing to concede that the Asante engineers have a point, so my challenge to them is this: As you guys think the timing is unrealistic; come up with a solution. Don’t ignore the variable of time. Perfect it instead. Why not combine time and volume with an impact percentage measure? Using the example above, it shouldn’t take a rocket scientist to write a piece of code that could tell me:

IOB 4.25 units. 94% exhausted in 30 minutes; 6% exhausted in 4 hours.

I’d be happy, you’d be happy, we’d all be happy. Happy, happy, happy.

Monday: Oh, Snap! I have no idea what I’m going to write about next week… I’d better snap to it and figure out what the Snap I should talk about next.