I finally solved a *really* hard problem at work. And it was 100 degrees today. Those two
things, along with lying out on my roof, and driving on the interstate at dusk (still 84 out) with all the
windows down in a town where usually I'm wearing a leather jacket even in july... was positively heavenly.
Happiness,
annoying levels of perkiness just oozed out of me today. Heaven. Then I went to church... it's almost
like being home. Happy sigh of content...
So someone recently asked why insulin
pumps can't monitor you instead of just giving insulin... and I started to answer them and decided why not
just write it all down.
So what follows is a generally unresearched vent...
1. Absorbtion rate of subq insulin is sloooooooooooooow. You may think,
what with lispro and all, 5-15 minutes pretty fast! Problem is NOT FAST
ENOUGH. Let's think about exercise, and a device that monitors blood
glucose AND administers insulin. At 5pm, you are 140 and start
exercising. At 5:15 you are 45. (This is how fast I generally drop at
the start of exercise if I've had any lispro in the previous 5-6 hrs).
By the time the device can monitor your blood sugar and make a decision
to cut the drip rate it's too late! Cutting the rate at 5:15 or even
5:10 won't produce an effect for at least 15 minutes. By 5:30, I'd be
passed out. With an action-reaction time of 15 mintutes to 1.5 hours
(when lispro seems to kick in for me), there is too much room for error
and delayed effects. Add to that the fast absorbption of food... You can
go from 100 to 300 in a half hour, THEN the insulin is delivered,
perpetuating the yo-yo effect.
2. Make it perontial (sp? liver cavity) would solve the speed problem --
prone to infection, and
infection there kills quick. It's a permanent open port to your body.
Swimming, water sports, not a good idea. Hygiene of the utmost
importance. Personally, not worth the risk.
3. The act of monitoring itself. Still in development. Most promising
developments are the
href="http://www.minimed.com/doctors/md_products_cgms_features.shtml">
CGMS by Minimed (Medtronic), the
href="http://www.glucowatch.com">Glucowatch, and an
upcoming device by SpectRx. I cannot comment on the SpectRx device, but
can share other's opinions on the first two, both of which require 4
tests a day as calibration. If we all tested that much, we wouldn't
really need these devices. Glucowatch: only measures
within a certain range, cannot measure with any sweat (rules out living
in the South! or the Midwest! or exercising!), causes skin irritation on
most people. Takes 2-3 hours to calibrate (with a finger stick) and one
'pad' (worn inbetween the skin and the device) only lasts 12 hours. CGMS
is invasive, requires a larger catheter than the pump needle itself,
causes much irritation and the device itself is as large as the insulin
pump, with no immediate display of results. Okay, so they aren't there,
but it will happen someday, I will grant the researchers that. But
leaping from a not yet mainstream technology to making it a close loop
with an insulin delivery device? eek.
4. Software. Closed loop (meaning the device measures and responds
without human interference). Bugs. How many bugs do I as a software
developer deal with constantly? Would I entrust my life to a piece of
software? I do already in some respect, but that is still a major
leap... Dereference a null pointer and you die.
I've officially turned into my mother. I don't know when it
happened, I suspect it was when I moved into my first 'nice' place. I
clean. I still don't cook... aha! That means I am not yet my mother!
I've been dealing with skin, then hair, some more hair, then some skin at
work. Then sometimes a coworker will stop by to pontificate pensively on
hair. I'm waiting for the same thing to happen with ray traced shiny
glitterly metallic paint or fast spinning tires. The day will come.
Stupid people. Can't live with them. Can't live with them.
I had a pretty close call tonight at the intersection of Broadway and
51st. There are two left turn lanes. As I came down the hill, in the
leftmost of the two lanes, I (gee, don't know why) fully anticipated being
able to turn left. A car in the rightmost of the two left turn lanes
turned left / pulled a u-turn IN FRONT OF ME. I slammed on the brakes
horn, everything flying towards the front of the car and the guy behind
me, who for some reason probably also anticipated being able to, I don't
know, TURN LEFT, went over the median to avoid hitting me.
Unreal. I had some more philosophical thoughts I had planned on blogging
but nothing is left in the noggin' after that....