Medtronic CGMS

by Drewgolden (A US based member of our forum)

I know the CGMS systems are very new, and over here in the States it’s fairly hard to get insurance companies to pay for them.

A CGMS, or a continuous glucose monitor, is a blood glucose monitor with a small sensor that is inserted under the skin. This monitor automatically checks blood glucose levels every few minutes. A small pager like device wirelessly gets data every few minutes and displays the results on a small screen.

There have been a number of studies conducted here, and of them, the conclusions are that the best success with the systems are strictly with motivated adults. Other groups were noted as having marginal or poor outcomes with the technology.

Further, the studies have indicated that the best success in the motivated adults is when:

* Pre-sensor classes occur.
* Only patients who are controlled diabetics use the technology (A1C of 7.5% or better)

From the studies, they have found:

* A direct relationship to A1C improvement and frequency of sensor use
* .7 to .9% A1C improvement when worn >70% of the time.
* Less hypo/hyper glycemic exposures occured

For further information, I would direct you to Dr Ira Hirsch’s presentation “Real-Time Glucose Monitoring: Revolutionary or Another Fad” – available in the iTunes Store, or on the web here.

The Medtronic System

I’ve personally worn the Medtronic System since December 2008, as part of an overall personal effort to reduce my A1C and as an ongoing effort to reduce the variability in my blood glucose swings. The key principals of my effort involved: training on carbohydrate counting and better diet, more accurate pre-meal insulin dosing, and efforts around controlling the “dawn effect.”

The Medtronic Real Time System comes in two flavors, as an add-on to existing Paradigm Insulin Pump users, and as a stand-alone Guardian device (for non-pump patients.)

What you get:

The kit.

This is a typical sensor change. The sensor is enclosed in plastic and comes in boxes of 3 or 10. Each sensor costs US $35 and are FDA approved for 3 days of use, however, most patients have discovered that you can easily use them longer, typically they are worn for 6 days (on day 7 the battery needs recharging.)

The tiny white battery/transmitter attaches to the blue charger and takes about 10 minutes to fully charge the transmitter, which will go as I mentioned seven days before needing a re-fill. The system comes with a sensor inserter device, which was designed to insert the sensor at a 45 degree angle. It has been since learned that the sensors need to go in at a steeper angle – about 60 degrees, so you must eyeball a deeper angle to get the sensor in.

Once in, the sensor can be attached to the battery/transmitter and will pretty much start working in about 3 hours – at which point you will be asked for it’s first calibration.

Oh my, that’s a big needle!

Yes, it’s a big needle. And it took a bit of “manning up” before I could get the first one in. But don’t let the size put you off – it’s fairly painless going in, and you do pull the needle part out – leaving only a tiny wire behind. The wire sits in your interstital fluid and goes after a reading there – which presents a few challenges to the system.

Big needle, small wire.

(Non-scientific Drew explanation) First, interstital fluid is the juice that sits between your fat cells. Typically, a glucose change will occur throughout the hemodynamic system first, eventually reaching fingers and far later reaching that delicious fat juice – typically it’s about 15 minutes or so behind.

So if you are dropping, know that the CGMS system is in fact 15 minutes behind where you actually may be.

Add onto this the calibration event. You have to be able to keep yourself in a steady state condition whenever doing a calibration – so the fat juice is what your finger stick is calibrating to. I have seen many users complain that these systems do not flat out work, but I suspect it may be the simple fat juice science, and inability to control blood sugars, that can be to blame.

The outcome, heading south.

So if the system is always 15 minutes behind, what good is it really?

There are several benefits I have found to wearing the system. First of all, it’s not so much being fussed about numbers, am I 3.5(63)? We’re kind of used to that mentality – making decisions about our diabetes management based on finger sticks and last actions. I just shot up and ate a meal, 45 minutes later I’m 3.5(63) – what do I do?

With a fingerstick – it’s one data point. 3.5(63.) But what was it 5 minutes before then? With the Real Time System (mis-named, should be time-delayed system) it’s less about the numbers and more about the trends. It’s easier for me to manage when I know the trend line – am I fallling and currently at 3.5(63), or am I going up and currently at 3.5(63)? See, now you have more information – take action or not.

Additionally, there are alarms. You can set a low and high alarm point. When the system sees you pass that point, a at first gentle reminder happens – woops – you are falling, just passed 3.2 and are heading south! This feature has already paid dividends at night – when we often go too low. And I can report the alarm starts of gentle – but comes to the point that it will wake the wife – which when you are real low is a very good thing.

If you live with a wife, children, or flat mates who act like children, the system also can be invaluable to them. Personally, I become very quiet and insist that I AM FINE despite clear signs that something is dropping on the good ship lollypop. Convincing a low diabetic to test their blood glucose can be difficult, or uncomfortable for those around us.

However, with a CGMS system, anyone can take a look at the display and see the whole story.

The final key to the puzzle is not in the system itself – but in the data you get out of it.

For me, I always felt a dis-trust to looking at finger stick averages. Personally, I never really tested when I was fine. It was always when I was questioning if I were high or low. So the averages to me were of a very random nature.

Because the CGMS takes readings (in this case, every 5 minutes) a significantly more detailed picture can emerge – data I find more actionable. Medtronic provide a Carelink web portal where all of the CGMS data can be uploaded and reports generated.

Sensor Trace.

Coupled with the insulin pump data, meal information and insulin doses are combined with readings. This provides an very naked view of a diabetic’s life (there is no place to hide!)

Link to larger view:

For me, the data has been very accurate. I only every calibrate my sensor when in steady state, and the reward has been very accurate sensor data. At $35 each for the sensor, and knowing they get 6 days a pop, you are looking at 1,408 pounds per year assuming daily use and a sensor change every six days. I have run the sensor 10 days without issue – you simply pop off the battery/transmitter after 6, recharge – pop back on and are good for another 6 days.

Because it’s just a wire, the sensor sites tend to completely heal very quickly – even my 10 day site healed fast!


* Expense.
* Difficulty using (tribal knowledge vs. documentation)
* Feels like 1.0 technology, requires comfort with experimentation
* The sensor injector needle can be scary
* 15 minutes behind
* Does not work for everyone (due to the above, and maybe other unknown factors like insufficent body fat?)
* Do not get this if you think you can avoid finger sticks, you still need to test before meals and whenever you calibrate.


* Like getting a finger test every 5 minutes
* Results can be observed on wireless device, and uploaded for rich reports
* Alarms if you go too high or low
* People you live with can have visibility
* Seeing trends help you make corrections better

Combining all of the things; better diet, carb awareness, pump, and sensor – I was able to achieve my A1C goal of 5.5 (22 years unable to get below 6.0) The sensor data has helped me identify problem areas, like I start my dawn effect at 3am – which has lead to somewhat tighter control.

If I had to order which have had the biggest impact, I would say:

* Diet and Carb awareness have had the largest impact.
* The Sensor Data has helped identify problem areas, and it’s kind of really cool to be able to visually see where you are whenever you want (without having to show blood for it.)
* The pump is a convenience that helped with dawn phenomena

I hope this provides you with a real-world example of someone who is using the system.

Best regards,

ps. I have a cheat sheet I wrote that provides tips to new sensor users.

MM Sensor Cheat Sheet:

* Only calibrate when your numbers are not moving. If you just ate, or just injected, then that’s a bad time to calibrate. I try to do it when I first get up and just before meals – when I know I am not moving. Why do you need to do this? Well – the sensor is reading interstational fluid (the juice between the cells in your body) rather than the blood flowing through your capilary system – they have figured out that interstational fluid is roughly 15 minutes behind the blood – so when you calibrate it is important to be steady state.

* Make sure you put the sensor in at a 60 degree angle.

* Many of us have discovered that it’s a good idea to put the new sensor in and let it “get wet” a few hours before attaching the transmitter and starting it up. So check your old sensor timer and maybe a few hours before the old one expires pop the new one in.

* If the sensor reports itself “lost” did the pump get too far? Sometimes in bed the pillows/pump/blankets impair the signal.

* I too suspected that calibrating high and low would give better results – I have not found this to be true. The sensor – when well calibrated – seems very accurate between ~60 – ~190 – above or below that it seems to get a little off. Not sure what the other folks have found – I’d love to hear their experiences.

* If you put in a new sensor and the site bleeds or there is trauma, you may want to wait an hour or so before attaching and starting it up. When there is blood, the sensor will sometimes report “Lost Sensor” or “Sensor Bad” – I have never given up on a sensor and waiting it out for a few hours usually gets them recovered (I think the blood causes the sensor initial confusion – but when the site clears up the sensor seems to do really well.) Your mileage may vary on this.

* Before going to bed, it’s a good idea to check when the next calibration is due – so you’re not woken up in the middle of the night for a calibration – do a calibration before bed will often avoid a wake up call.

* The sensor relies on the last 4 calibrations – so if you mess one up (your numbers were moving when you calibrated) the sensor will be off until you do four more calibrations – I have found that re-starting the sensor (tell it there is a new sensor) wipes out the bad data – others will do a calibration every 15 minutes for an (glucose steady) hour to force the bad calibration out.

* When the sensor is spot-on, and it beeps because you crossed the low threashold (mine is set to 68) remember that it’s 15 minutes behind and you are probably actually LOWER. Test and take care of your low sugar based on the finger stick reading. Don’t be suprised if it takes the sensor time to catch on to the correction (it will see it, but takes 15 minutes or so – so you may get another low alarm.)