More impressions
A few more things I've noted... and this is just my first day!
1) The Dex seems to be a lagging indicator when a trend goes HIGH, and a leading indicator when a trend goes LOW.
What I've noticed is that the Dex lags behind the meter when a rising BG trend gets established. By the same token, it leads the meter when a dropping BG trend gets established.
I have a theory about this, which is most likely complete and utter nonsense, especially considering my highly limited understanding of this field. I'll share it anyway so you can all make fun of me. It goes like this:
The Dex is subcutaneous whereas meter is blood, and the way the body works is that sugar in the blood gets broken down, then enters the cells through the layer in which the Dex sensor is embedded. This process takes a bit of time. Now see what happens:
In a rising trend, too much sugar is in the blood stream, and the rate of breakdown to glucagon is essentially too slow; thus, it takes it more time to filter through the "Dex layer" and show on the Dex screen. The end result is that the Dex reacts a little slower in confirming this trend, as the breakdown activity "catches up" to the rising trend.
As a side note, if this is right, then it could very well be that the Dex will never be able to show blood sugar levels when no insulin is active whatsoever (such as pre-diagnosis).
In a dropping trend, the reverse occurs. As blood sugars drop, the cells get starved even faster, because too much insulin is breaking down too little sugar much too fast. In a sense the Dex becomes predictive because of its location as a middleman between blood and cell.
Again, if I'm right then people who generally have stable blood sugars will see overall a much higher degree of "accuracy" when comparing the Dex to their meter than those who have bigger and more rapid changes in blood sugar levels. It would also suggest that the Dex on its own will be less useful in times of, say, sickness, when blood sugars get messed up. But it would be more useful if used in conjunction with a meter than a meter alone; see point 2.
Enough pseudo-science. If anyone has some ideas around this I'd love to hear them, and again, what I say above could be and is entirely likely to be complete and utter hogwash.
2) The combination of Dex and meter is extremely powerful in allowing the PROPER reaction in the proper TIMING once a trend gets established. Why?
When its trending lower, the Dex will tell you two things: (1) a hypo is coming even if your meter hasn't caught up to it yet; it might be just enough of a lead to avoid an actual hypo altogether; (2) that your corrective action has succeeded in staving off the hypo when the Dex is still claiming low blood sugars but your meter is already reflecting the future increase (reversing the trend).
It's a little more difficult to explain when the Dex is trending higher. Basically, the velocity of the change on the Dex can give you a good guesstimate of how high you'll go, when compared to the actual values as measured by your meter. In other words, the meter becomes a predictor for the Dex, while the Dex provides an indication of when the rising trend is tapering off (and thus the rise will stop). The combination of both results can provide a pretty tight window estimation of just how high you are about to get and how to correct.
Think of it this way. The Dex is trying to play "catch up". It will therefore hit its highest "slope" on the graph when your actual BG's are still rising. However, the graph will being to taper off when your actual BG's have more or less stopped rising, because the catching up must become more accurate. You can then use your meter to see where you actually are and have a pretty solid indication that you won't rise much anymore. Sure, the Dex may take another 15-20 minutes to get there, but you get to act NOW, and get to act APPROPRIATELY.
I've used both of the above analysis successfully today, in staving off a hypo before it happened and stopping a rise in a manner that avoided what would previously have resulted in a follow-up hypo.
3) Can someone tell me what the deal is with the stupid transmitter? first of all, the range is dismal. But beyond that, why the heck are they even using an active instead of a passive unit? wouldn't it be a lot easier to use a dynamic RFID chip which is then read by the display unit? it would surely make the cost of what is today the transmitter unit a lot lower, because a passive unit is a lot easier to build let alone make waterproof. It would also allow it to be a lot more compact. Am I missing something here?
1) The Dex seems to be a lagging indicator when a trend goes HIGH, and a leading indicator when a trend goes LOW.
What I've noticed is that the Dex lags behind the meter when a rising BG trend gets established. By the same token, it leads the meter when a dropping BG trend gets established.
I have a theory about this, which is most likely complete and utter nonsense, especially considering my highly limited understanding of this field. I'll share it anyway so you can all make fun of me. It goes like this:
The Dex is subcutaneous whereas meter is blood, and the way the body works is that sugar in the blood gets broken down, then enters the cells through the layer in which the Dex sensor is embedded. This process takes a bit of time. Now see what happens:
In a rising trend, too much sugar is in the blood stream, and the rate of breakdown to glucagon is essentially too slow; thus, it takes it more time to filter through the "Dex layer" and show on the Dex screen. The end result is that the Dex reacts a little slower in confirming this trend, as the breakdown activity "catches up" to the rising trend.
As a side note, if this is right, then it could very well be that the Dex will never be able to show blood sugar levels when no insulin is active whatsoever (such as pre-diagnosis).
In a dropping trend, the reverse occurs. As blood sugars drop, the cells get starved even faster, because too much insulin is breaking down too little sugar much too fast. In a sense the Dex becomes predictive because of its location as a middleman between blood and cell.
Again, if I'm right then people who generally have stable blood sugars will see overall a much higher degree of "accuracy" when comparing the Dex to their meter than those who have bigger and more rapid changes in blood sugar levels. It would also suggest that the Dex on its own will be less useful in times of, say, sickness, when blood sugars get messed up. But it would be more useful if used in conjunction with a meter than a meter alone; see point 2.
Enough pseudo-science. If anyone has some ideas around this I'd love to hear them, and again, what I say above could be and is entirely likely to be complete and utter hogwash.
2) The combination of Dex and meter is extremely powerful in allowing the PROPER reaction in the proper TIMING once a trend gets established. Why?
When its trending lower, the Dex will tell you two things: (1) a hypo is coming even if your meter hasn't caught up to it yet; it might be just enough of a lead to avoid an actual hypo altogether; (2) that your corrective action has succeeded in staving off the hypo when the Dex is still claiming low blood sugars but your meter is already reflecting the future increase (reversing the trend).
It's a little more difficult to explain when the Dex is trending higher. Basically, the velocity of the change on the Dex can give you a good guesstimate of how high you'll go, when compared to the actual values as measured by your meter. In other words, the meter becomes a predictor for the Dex, while the Dex provides an indication of when the rising trend is tapering off (and thus the rise will stop). The combination of both results can provide a pretty tight window estimation of just how high you are about to get and how to correct.
Think of it this way. The Dex is trying to play "catch up". It will therefore hit its highest "slope" on the graph when your actual BG's are still rising. However, the graph will being to taper off when your actual BG's have more or less stopped rising, because the catching up must become more accurate. You can then use your meter to see where you actually are and have a pretty solid indication that you won't rise much anymore. Sure, the Dex may take another 15-20 minutes to get there, but you get to act NOW, and get to act APPROPRIATELY.
I've used both of the above analysis successfully today, in staving off a hypo before it happened and stopping a rise in a manner that avoided what would previously have resulted in a follow-up hypo.
3) Can someone tell me what the deal is with the stupid transmitter? first of all, the range is dismal. But beyond that, why the heck are they even using an active instead of a passive unit? wouldn't it be a lot easier to use a dynamic RFID chip which is then read by the display unit? it would surely make the cost of what is today the transmitter unit a lot lower, because a passive unit is a lot easier to build let alone make waterproof. It would also allow it to be a lot more compact. Am I missing something here?
posted by Blinkered at 8:33 PM
2 Comments:
RFID is a newer technology -- remember that a medical device is 10+ years in the making.
Yeah, I suppose that's fair enough Kath, but for heaven's sake, can they hire an RFID guy to make it happen? it will also solve many of their expensive overhead relating to bad sensors and such.
Post a Comment
<< Home