Catching the low-hanging fruit

Today’s CGM Primer: “If you set your CGM’s low threshold alarm at 75 you will never, ever, ever get an alarm when you are 75. By the time you get your alarm the horse is out of the barn and the barn has burned to the ground. You might actually be at 65, 55, or lower.”

----from Beyond Fingersticks


Truth be told, both girls missed it. With two CGMs, I figured I was pretty low-proof. But the stupid little Roche Nano was the first to break the news to me. I did a fingerstick right before lunch and was at 78 mg/dL. Shit.

I checked new Dex and she thought I was at 87. Old Dex thought 97. Well, it was a slow coasting low. Nothing dramatic. That’s how it works with CGM: the blood gets ahead of the interstitial fluid. CGM, like my wife, is frequently running late.

Here’s how it works (with CGM, not with my wife— I have no idea why my wife is always late). Recall that CGMs don’t do blood. They look at sugar in the water between your cells. Where blood goes, that water will follow. Think of it like a roller coaster ride. Every car in the train will follow the track (barring a very unfortunate incident, anyway), but no two cars are at the same place on the track at the same time.

Capillary glucose is in the front car. Alternate-site testing glucose is in the middle car. CGM is sitting in the back car. In a blood sugar low, the fingertips can report the news first. This is important ‘cause your fingertips and your brain are wired much the same. What happens at the tips tells you what’s happening in your mind. Literally. Not enough sugar in your pinkie and your brain is shutting down, too.

While this time lag vexes many CGMers, it’s really not as big a problem as it sounds like. Not generally, anyway. If you’re sitting in the back car of a roller coaster, you can still see a peak coming. Or a drop. You can see the lead car screaming down the track towards the bottom. It’ll get there first. But you’ll get there, too. Soon.

CGM is the same way. We can’t expect the CGM readings and the fingerstick readings to line up in a fast drop. But if we’re paying attention, we can watch the direction the train is going and know what’s about to happen.

Of course, I wasn’t paying attention. And as the low was a slow-moving one, no rate-of-change alarm went off to tell me to pay attention. And truth be told, even though I don’t like being in the 70s, I really wasn’t in any danger, and now had an excuse to raid the sack of chocolate-covered blueberries on the carb table near the nurse’s office. (Anti-oxidants and stuff, it must be healthy, right?)

I pop a few of the yummy morsels in my mouth and start my lunch. Half way through, with the meter showing a mild rise established, I feel a vibration on my waist. The G4 is finally on the case of the mild low. The lead car of my blood sugar roller coaster is already headed back up, but the tail-end car has reached the bottom arc of the track.

The vibration is a mild tugging, like an insistent small child pulling on your hand for attention. Hmmmm…. Old Dex had a vibration that shook the foundations of the universe. A deep-resonate buzz that rattled your teeth and could wake you at night. If you’re awake, you’ll notice G4’s vibration. But asleep?

Can the new girl’s buzz wake me at night, like her older sister does? If not, I’ve got a problem. The Dex CGMs, old and new, use good vibrations as their first-phase alarms, no matter what noise-making options you’ve chosen. They don’t go audio until five minutes later. In the middle of the night, if you sleep through the vibe, you’ll be five more minutes into an emergency before anyone “talks” to you about it.

I may have to review my threshold settings with that in mind…. I may have to set them higher to give myself more time to respond to a low. Oh, and speaking of threshold alarms, this is one area where The Other Guys still trump Dexcom. Both the old Seven Plus and the G4 have very limited options when it comes to threshold alarms. We can only choose one low threshold and one high threshold for the entire day. Med-T systems allow us to program different alarm thresholds for different times of the day, much like basal rates on a pump. Frankly, this is one of the strongest features (along with predictive alarms) that the Med-T system has going for it. I was disappointed that the G4 doesn’t have this option. It seems like such an easy thing to add.

Why does it matter? Well, if you are the mom of a little type 1, you might want a higher nocturnal low threshold to give you more time to deal with a night-time emergency. Or if you’re an over-exhausted college student, you might want a higher upper-end alarm at night. Sleep might be more important than fixing a 200 mg/dL, but when you’re up and about, you might want to keep tighter control.

Anyway, back to the lunch table. I decide not to acknowledge the alarm. I let it ride. I want to hear the new girl’s voice in the real world. I don’t warn my colleagues, who are in a lively discussion about the latest polling numbers.

I squirm in my chair. I’m having a hard time paying attention to the conversation. I’m waiting for the alarm. It’s like Chinese water torture. Hasn’t it been well over five minutes already?

Bee-Bee-Dee-Me-Ree-da-da-da. Electro-Latin-American-Pop kindergarten band. Musical, simple, electronic, unique. Not quite child’s toy. Not quite muffled ring tone. A little bit playful. The G4’s low alarm debuts.

Startled, the nurses are patting their pockets, looking around, reaching for their cell phones. Is that you? Is that me? Only our social worker, mom of a little type 1, turns sharply and fixes her piercing blue eyes on me, “Is that a low alarm?” she snaps, her mamma bear anti-diabetes maternal protective instincts rearing up in an instant.

I laugh at her, and tell her, This is a test of the emergency low blood sugar system. Had this been a real low blood sugar emergency, you would have found your clinic’s diabetes educator lying in a pool of drool on the floor.

I take the Dex off my belt, open the case and show her the pretty red graph. Everyone crowds around to see the latest and greatest.

One of the things Dexcom is bragging on is the G4’s supposed improved performance in lows. It’s supposed to be something like 17% more accurate than the old one. In this case, which was hardly the poster-child for a hypoglycemic event, G-4 did recognize the slow-coasting low. Granted, I had found and fixed it with a the help of an old-fashioned meter before it did, but still…

And the old girl?

Old Dex never caught the slow low at all.