Continuous Glucose Monitors Are Revolutionary - I’ve always believed that if I could give myself insulin conveniently and constantly knew my blood glucose, I could control my blood sugar almost as well as a non-diabetic person. Nine years ago, an insulin pump made the first condition come true. Since then I have been waiting for the magic blood sugar machine.

My doctor lent me a MiniMed continuous sensor for three days, but there was no way to see the readings until after the experiment was over. I bought a Cygnus GlucoWatch but it didn’t penetrate the skin, so it skipped too many readings and was uncomfortable. I guess I could have done twenty daily finger-stick tests as a short-term strategy, but I was never really in the mood.

Finally, in October 2006, my dream came true.  I got the Dexcom STS Continuous Glucose Monitoring System.  It gives me glucose readings every five minutes, thereby showing trends in my glucose. I consult it frequently so I can respond to falling blood sugar before symptoms of low blood sugar hit, eliminating many hypoglycemic episodes entirely. In the event that a low does occur, the adjustable low sugar alarm warns or wakes me. The monitor also allows me to recognize high blood sugar sooner, so I correct it before it gets worse and spend less time hyperglycemic. That’s tighter control, but without losing hypoglycemic awareness.

Shock of Discovery

Knowing my sugars continuously is what I always wanted, but it’s been an unpleasant shock to see the wild excursions in my glucose levels. When I did a fingerstick before a meal and another one two hours after eating, I wasn’t aware of the high sugars in between. 

I was equally shocked by how hard it is to avoid highs and lows, even armed with all this new information. I thought that getting the monitor would make it a no–brainer to “stay between the lines.” It’s a lot harder than that.  For one thing, with continuous glucose measurement, treatment decisions become more frequent, so most of the insulin I need is already “on board.” My insulin pump is too old to do “insulin on board” calculations, which prevent “stacking” one insulin dose on top of a previous still active dose.

I’m thinking about reducing carbs rather than trying to quash big post-meal glucose rises with insulin. When driving, I try to anticipate slowdowns and ease the car gently along. For me, food and insulin are like accelerator and brake: I would like to use it to glide toward my target blood sugar without “rollercoastering.” After months of practice with my Dexcom, I’m still struggling to find a satisfying strategy.

Even before the pump, I charted my blood sugar and annotated carbs and insulin on a time-line. Now I have so much more information that it’s a struggle to understand it. I have a hunch that treating my diabetes will get more complicated and obsessive until I reach a new level of understanding. Then treatment will get radically simpler. Perfection is not required, however; even with some highs after meals, my A1c is already dropping.

Faking the Sensor

The insertable part of the Dexcom glucose monitor, the sensor, costs $35 and expires in three days. I learned right away not to throw the sensor away when it expires. You can trick the receiver into starting a new three-day period by pushing buttons to claim you have put in a new sensor. This “fake” sensor must go through a two-hour start–up period, followed by calibration. The sensor often performs better in its second life. I repeat this process for as many as seven three–day periods. This reduces costs dramatically! The receiver will shut down early if there’s too much noise in the measurements, or if the sensor readings are too different from the finger–stick calibrations. Even then, pretending to insert a new sensor usually fixes the problem. If the receiver shuts down early a second time, then I actually replace the sensor. Medtronic MiniMed sensors (which also cost $35 and expire in three days) can also be extended in an alternating seven-day cycle. My success in extending sensor wear seems completely typical for either brand, according to the active discussion group diabetescgms on Yahoo Groups.
 
Experiences

The Dexcom, like all CGMs, measures interstitial fluid rather than blood. It takes slightly longer for blood sugar changes to show up in interstitial fluid. This is one reason the default low–sugar alarm is set to 80 mg/dl, higher than most people would expect. That setting works well for me, and false alarms are infrequent. The vibrating alarm is emphatic enough to wake me; letting it rattle on the night table works better than losing it in the bedcovers.

The Dexcom can display one–hour, three–hour, and nine–hour charts of sugar levels. These stretch and shrink to occupy the same screen area. The one–hour chart always looks flat even when the other screens show levels changing rapidly, so I don’t use it. The three–hour chart works well during the day, and the nine–hour chart is good first thing in the morning. The screen does turn itself off too soon for me.

The sensor is supposed to stay dry, so Dexcom sells adhesive shower patches to cover it. The patches fail often, but blotting the area dry gets the readings back to normal in an hour. Recently, I stopped peeling off the paper ring; it protects the patch edge from rolling back and failing. If a shower patch sticks really well, says one acquaintance, he leaves it on for several days.  However, the adhesive irritates my skin. Once the transmitter accidentally came apart from the sensor, but I was able to plug them back together without a problem. Sometimes there is a bit of dried dirt or blood near the site when I finally remove the sensor.

The Dexcom receiver is a smooth, round piece of plastic that feels really good in my hand. The belt holster that comes with it is clumsy, so I carry mine loose in my shirt pocket. I was afraid I’d lose it when I bend over, so I taped my name and phone number onto its backside.

Way Forward

Dexcom explains that the current device is “first generation” and that improvements are on the way. They are doing clinical trials on a seven–day sensor and previously showed a one–year implantable sensor on their website. While writing this article, I was told that they were going to crack down on users who extend sensor life, by issuing one single–use password with each new seven–day sensor. They explained that the clinical trials were only for three or seven days per sensor, so it would be irresponsible for them to let me extend use beyond what was demonstrated in the trials. I cynically assumed they were trying to shake me down to buy more sensors, but in March they announced that they will eliminate the password. It turned out that they were shutting the sensor down early in an attempt to please the FDA. I’m delighted that they listened to users’ objections. Americans always brag when they get a lot of mileage from a car or a glucose sensor, and faking the sensor is a favorite discussion topic for Dexcom and MiniMed users. I don’t like going through the charade of putting in a fake sensor; it blinds me to the readings for two hours at the most inopportune times! Sensors could eventually last a year.
 
Continuous glucose monitors are an important advance that will lead to longer, healthier lives for diabetics. They’re at least as complicated as pumps, so in addition to basic user education, we need support groups and user bulletin boards. We need better conceptual frameworks to comprehend our glucose swings and to understand how to use and troubleshoot these monitors. 

For an excellent video on the subject of continuous glucose monitors by Dr. Steven Edelman, professor at the UC San Diego School of Medicine, go to http://cme.ucsd.edu/cgmonline/