lkalliance

I'm not yet comfortable with CGM units. As Spike2013 notes above, a CGM is especially useful for detecting trends. Even the most accurate blood glucose test only tells half the picture. I'm at X. Am I at X and falling? How fast? X and rising? How fast? A CGM need not hit the blood glucose with a lot of accuracy...it DOES need to accurately give me a sense of the trend. I can do a finger stick for the "accurate" reading.

I get frustrated because the three pieces of technology I use (glucose meter for blood test, insulin pump for dosing, iPhone for logging) don't all work together as a trio. I can get a meter that will talk directly to my phone, but which then requires me to manually enter the reading into the pump for dosing. I can get a meter that will talk directly to my pump and transmit the reading and save a step, but then I need to remember to log the reading.

(I happen to use Medtronic; it may well be that there is a different brand of pump/meter that will make them all work together, but I'm not interested in changing pump brands)

At the moment I choose to have my meter talk to the pump.

tallest skil said:

The thing about something like this is it has to be perfect. I mean, fucking PERFECT. The fact that it would be used outside a controlled medical context means that it has to be even more accurate. This isn’t “the number of cell reception bars isn’t being reported correctly”, and it isn’t “the phone is slower because the battery is dying”. This is “get it right or you could have a death that people point to you about.”

Never mind that you know the media will be screaming “FAULTY APPLE WATCH CAUSES DIABETIC COMA” simply because the guy who went into a coma was wearing one at the time. And by the time it’s discovered that he ate an entire box of Krispy Kremes and that the Apple Watch was working perfectly, the PR damage will have been done.

Welllllll...the bar for "perfect" isn't infinitely high. Current home-use blood glucose monitors aren't perfect, either: there's some level of reliability they have to reach regularly when compared to lab equipment. I don't remember what that figure is, but I remember it surprised me when I heard it. Within 20% either way, perhaps?

But there is some use even if it's not right on with lab equipment: if it can be validated to be off by a consistent amount, then at least you can get a reliable reading on whether your blood sugar is going up or down over the course of the day, and make adjustments to your daily insulin regimen. This is the value in CGMs currently.

Why would a non-invasive (and from the little biology I know, I don't see how it could be) device require FDA approval, especially over the next few years as the current administration attempts to deregulate everything.  How would it be any different than a blood pressure gauge?  I don't think my gauge is FDA approved.   And unless there's a way to measure blood glucose via saliva, I don't see how it could ever be non-invasive anyway.  Of course there probably would be a way for a device to track the results from urine test strips, but it's the strip that's already doing all the work.  Unless they expect us to urinate on our phones or watches. 

It has to do with indications for use. Although a patient might use a reading from a blood pressure monitor to make personal healthcare decisions, he would more likely take that data to his doctor, I would think. A diabetic makes immediate decisions based on the readings from his meter.

From what I understand, potential non-invasive glucose measurement involves shining light through the skin and watching how that light gets diffracted. Therein lies a big challenge: glucose molecules are very basic things, and light reflected off one doesn't have a distinctive look to it. It could have bounced off of anything, not just glucose. Perhaps they find a way to measure some secondary substance. I don't know if it's still the case, but Continuous Glucose Monitors don't test your blood but rather the interstitial fluid between cells, which has a strong correlation. Strong, but not strong enough to use the data for insulin dosing: you need to actually test your blood for that. A non-invasive CGM would be really good...a non-invasive blood tester that I could use to set a dosage would be AWESOME.