A Little More Cyborg

Thursday, July 20, 2006 | 3 comment(s)

Right now, I feel like I'm a little bit closer to being a real-live cyborg.

This morning, I got set up with Minimed's CGMS® System Gold™ (sic). In addition to my insuln pump, I'll be wearing this second, larger, non-detachable machine on my opposite hip for the next 72+ hours.

First impressions: This thing is bulky. It’s probably about the size of a blackberry, but rather than the thin tubing I'm used to with the pump, the wire running from it to me is more like something you'd plug into a stereo (I'm wired for sound!). Also, since I can't disconnect from it and it’s not water proof, I have to seal it in a zip lock bag contraption while showering (I can't wait
Just So You Know
That there parenthetical tidbit was loaded with sarcasm. I wouldn't want you to actually think that I'm looking forward to this experience.
to try that out tomorrow morning).

It is purported to be measuring my blood sugar every 20 seconds, and then averaging them every 5 minutes to produce 288 data points
Show Your Work
24hrs * 60min/hr = 1440min/day
1440min/day / 5min = 288 data points per day
of blood sugar readings per day. This is an order of magnitude more than my current self monitoring.

By most standards, I'm an obsessive
My name is Kevin, I've been diabetic for 27 years now, and it’s been 36 minutes since I last tested my blood sugar.
glucose tester. I currently test approximately 12-20 times per day. I record all these readings in an Excel spreadsheet which automatically plots out a graph of my daily blood sugars (see below). Looking over some of my recent charts and thinking about the data that I'm going to be getting back from this CGMS experiment, I started to think about my A1c conundrum.
In Case You Missed This
I've been puzzled about the disconnect between my average meter readings and my most recent A1c reading(s). You can read about it here in all my yammering glory.

Okay, Okay
I'll admit it. This is a pretty atypically good day.
Also, if anyone would like a copy of the Excel file I use, I'm more than happy to share it.

So, I always thought that I was getting a picture that was pretty darn close to what I'll be getting back from the CGMS data. I mean, there isn't a whole lot of room for missed spikes in here. But while looking at this chart and thinking about my A1c conundrum, a light bulb went on over my head. The average reading coming out of my meter is biased downward because when I have a low blood sugar, I tend to test multiple times in a very short period of time.

You see, on July 18th, I tested 21 times. When I calculate the average, each reading contributes 1/21st to the overall result. In reality, however, if I were to only give each reading 15 minutes worth of weight, each would only be contributing 1/96th
Showin' My Work Again
24hrs * 60min = 1440min/day
1440min / 15min = 96 15min increments per day.
of the day. So, I'm clearly overweighting my low blood sugar readings simply because I'm testing so frequently around them. When I have low blood sugars and I test in 15 min increments, I'm pushing my average lower and lower, even though I'm not really spending a whole lot of time down in that range. This is just a hypothesis, but I think it's a pretty good one.

Hopefully the data I get back (in two weeks or something (almost useless) like that) will help me figure this out for sure. I should be able to compare the calculated average from my readings with the calculated average from the continuous readings. And since I've already had a low blood sugar while on this thing and tested within 15 mins of each other, I should be biasing my average downward and should see this when I get the data back.

(I'm really, really sorry if you've fallen asleep by now. Don't worry, I completely understand - I have this effect sometimes).

3 Comment(s):

Blogger floreksa said...

I think your theory is dead-on. I've created an excel chart (looks frighteningly like yours). Basically mine had hourly inputs for readings, so even if I tested 5 times for 1 low, I only inputted 1 of the readings.

Give that a shot.

Oh, and I HATED wearing the CGMS. That "stereo wire" was so stiff (and short!), that I just couldn't figure out what to do with it. Didn't help that they gave me one with a broken clip so evertime I moved it would go flying.

Blogger Scott K. Johnson said...

Aha! I think you're on to something here Kev!! Nice!

It will be really interesting to see how things correlate between your super nice Excel tracking and the CGMS.

That does not sound comfortable to wear. It's going to be a LONG 72 hours eh?

I do think that this is the next big wave of technology in DB management - and it's going to mature and get better, easier to use, more comfortable to wear, more affordable, all that good stuff.

I'm anxious to hear what "light bulbs" this will pop on for you.

Blogger Minnesota Nice said...

Kevin, are you like a genius or something? We could all chip in and pay your tuition to medical school and then you could be our online doctor.1

Post a Comment || Go Home