Med Student with Diabetes Cautions Against Unrealistic Expectations for CGMs

Dear Editor, I am a medical student in the M.D. program at Oregon Health and Sciences University and a type 1 diabetic of almost 10 years. I use a Medtronic pump and I also use their continuous glucose monitoring system (Paradigm Real-Time).

I am writing to caution readers against developing unrealistic expectations for the current generation of continuous glucose systems (Dexcom and Paradigm Real-Time). Both of these systems are useful to establish blood sugar trend data, but they are inherently limited in several ways. I feel these drawbacks are underemphasized by companies wishing to sell extremely expensive products, and thought readers should know about a few of the not-so-desirable features before investing in these systems.

First, their readings are substantially delayed in time: i.e. they often lag as much as 20 minutes behind actual blood sugar concentration. This is because they do not sample capillary blood. Standard fingerstick monitors sample blood from capillaries, which is representative of the current blood glucose level. The sensors are detecting glucose in interstitial fluid, and the glucose concentration in this fluid may take up to 20 minutes to change in the same direction as blood glucose. Thus they cannot reliably protect against hypoglycemia, as 20 minutes of delay is entirely too long for hypoglycemia to go untreated. Exacerbating this problem is the increased inaccuracy of the sensors at both the low and high blood glucose values: Just like conventional finger-stick glucose monitors, they are less accurate at the very high and very low ranges.

Second, neither of the systems substitutes for fingerstick testing. A diabetic using these monitors must calibrate them using the same blood tests that they currently perform. The calibration must be done multiple times daily. And both companies’ literature explicitly warns against using anything but a fingerstick value to make decisions about insulin doses. Diabetics thus will not be reducing the number of fingerstick tests they need to perform.

Third, the devices are very expensive to purchase and use. The initial cost of the Medtronic system is $1500 and each box of 10 sensors (which are approved for three days though many people use them for up to 7) costs $350. Even more breathtaking is that the Medtronic system is only designed to last for 12 months until the transmitter ceases to hold a charge; after 1 year of use the diabetic must pay Medtronic another $600 for a new transmitter! Thus the first year of use costs, at minimum, $1500 plus 335 days of sensors, a grand total of $5,408. To this cost, add the replacement cost of the transmitter and you reach about $6000 per year, or $16.43 per day.

This cost is not covered by most insurance plans! Only a few insurers are reimbursing this expense. All of this money is thus in addition to the regular expenses associated with having diabetes, which is costly even with adequate insurance.

Fourth, pump users will require a second subcutaneous insertion. For those with lean bodies this can be problematic, as the abdominal fat they most likely use for the pump site can become scarce when two, rather than one, insertion sites must be found. The sensor probe that resides under the skin is sensitive to being bent (unlike the insulin pump infusion set) and thus any area that is continuously flexed or that bears weight is not a great choice for the sensor probe. That means the top of the buttocks and the legs are not optimal choices for locating the sensor.

It seems to me that Diabetes Health is covering many of the positive aspects emphasized in Dexcom and Medtronic promotional information without giving the drawbacks equal time. I believe this will lead to unrealistic hopes among diabetics eagerly searching for more effective ways to manage their disease. I urge any prospective buyers of these systems to look into these issues, and many more, prior to using one of these devices (sometimes a trial can be arranged with a Diabetes Educator). While the trend data they provide is valuable, and those diabetics in need of this data will surely be able to improve their daily management, the limitations are numerous and need to be disclosed.


Zachariah Kramer
OHSU M.D. Class of 2010

– – – – –

Editor’s note: You can click on links to Medtronic’s and Dexcom’s reponses below. We appreciate the balanced discussion of this important topic.

Read DexCom’s Response

Read Medtronic’s Response 

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.