thinking like a pancreas September 30, 2011

The Great CGM Debate: Which System Reigns Supreme

by gary scheiner, mS, cDE

Tweet

Gary Scheiner

Earlier this month, Abbott finally announced the official withdrawal of the FreeStyle Navigator continuous glucose monitoring (CGM) system from the US market. Despite receiving rave reviews for its accuracy, Navigator was plagued with manufacturing issues, high costs, and reimbursement problems and was not being marketed or supported for the past year.

Navigator’s demise leaves two companies in contention for CGM supremacy: Medtronic and Dexcom. I’ve worn all of the systems for extended periods, and have trained dozens of patients on each, and having no formal allegiance to any particular company, I thought I would give you my opinions.

To add to the fun, a few weeks ago I managed to get my hands on some of Medtronic’s new Enlite sensors and an inserter (anyone can use these in Europe who has a prescription). Medtronic says that it has designed its new sensors to be better than its traditional Sof-Sensors, so I figured it was time to do some side-by-side-by-side, three-at-a-time comparisons (Medtronic-Sof Sensor, Medtronic Enlite Sensor, Dexcom Seven Plus). Keep in mind that I am what they call in scientific circles an “n of 1.” That means that this experiment reflects observations from just one person and does not necessarily prove anything – so take this for what it’s worth.

So let’s get to it. Here’s my impression of the Dexcom Seven Plus, Medtronic REAL-Time (using the Sof-Sensor), and Medtronic REAL-Time (using the Enlite sensor).

Dexcom Seven Plus has a leg-up in a number of areas. For me personally, the sensor is very comfortable to insert. I usually feel nothing when inserting and withdrawing the introducer needle. (Though some patients do comment on pain associated with the DexCom sensor, it is widely perceived as more comfortable than Medtronic’s Sof-Sensor). The Dexcom sensor itself lasts a long time: I can make it last 12-14 days (though bear in mind the Seven Plus is FDA-approved for only seven days of use). The built-in adhesive holds the sensor in place quite well for a week or more. The sensor and transmitter are flat and tapered. The calibrations can be performed at any time (even when your blood sugar is in a state of flux). The transmitter never requires charging. The display on the receiver is large and bright. And the alerts, both audible and vibratory, are strong enough to be noticed by most people even while they are sleeping. The receiver can be set to vibrate initially, and then only beep if the user fails to confirm the vibratory alert.

Another pleasant aspect of the Dexcom is its simplicity. There are very few steps in the start-up process, and only a handful of menus on the receiver/programmer. It is intuitive enough that most people can self-train using the company’s online directions. Error messages are few and far between.

Dexcom’s accuracy is generally quite good. I find that the sensor matches the fingerstick calibrations fairly well by the end of the first day of sensor use, and matches most closely from days 3-12, at which point things start to deteriorate. Again, individual use may vary, but my experience seems to match that of most of my clients. From initial calibration to sensor failure, I experience an average discrepancy of 10-12% between the Dexcom sensor and simultaneous fingerstick calibrations.

On the downside, Dexcom sensors are more expensive than Medtronic sensors ($80/sensor vs. $35/sensor), but the difference isn’t much when you consider how long the sensors last on a “daily” basis. If used according to prescription (seven day use for Dexcom sensors and three day use for Medtronic Sof-Sensors), both sensors would cost around $11/day. Assuming on average that Dexcom users wear sensors for 12 days and Medtronic users wear sensors for 7 days, the daily cost is $7/day vs $5/day. Unlike the Medtronic transmitter, Dexcom’s does not store any information for times when you are out of range of the receiver. Having to charge up the Dexcom receiver every three to five days is a bit of a hassle, especially for people like me who travel a great deal. Additionally, the charger connects to sensitive filaments in the transmitter, meaning that when the filaments bend out of shape or the charger cable goes bad, the system may have problems charging. In terms of data analysis, the Dexcom receiver does not generate any statistics for the user. Downloading the receiver to a PC (no Mac compatibility) is rather slow. In addition, the Dexcom receiver does not let the user know how long the sensor has been in use, how long until a calibration is due, or when the current sensor expires. One more thing: taking medications containing acetaminophen (such as Tylenol products) will trigger a very high false glucose measurement by the sensor for several hours.

But the major drawback to the Dexcom system is that the display is not yet integrated into an insulin pump. With so many CGM users also on insulin pumps, this is a significant shortcoming. Having to wear a pump and carry around a CGM receiver isn’t just inconvenient, it can also be impractical. This is particularly a problem for young children and anyone who is very active. Since it is not tethered to me the way my pump is, I’ve misplaced my Dexcom receiver more times than I’d care to count. Sometimes I have to bribe my kids to track it down for me: “First one to find Dad’s Dexcom gets five bucks!”

Dexcom is trying to work through the integration problem with the FDA. The company currently has two CGM integrated pumps in development, one with the Animas Vibe and one with the Insulet OmniPod. In Europe, the combination Animas Vibe Animas/Dexcom pump and CGM has been approved since June. However, approval of the Vibe as well as the Insulet/Dexcom system appear to be further off in the US. If all goes well, both products may be available in the US by late 2012. (Editors note: for another user's experience with the Dexcom Seven Plus, see our Test Drive in diaTribe #15).

The Medtronic REAL-Time CGM (using the Sof-Sensor) includes both the stand-alone Guardian system and sensor-augmented pumps (Paradigm 522, 523, 722 and 723). The major advantage with the sensor-augmented pumps is the all-in-one concept. No need to carry around a receiver for the CGM since everything displays right on the pump screen. The transmitter stores up to 40 minutes of data in case the receiver is temporarily out of range. Once back in range, the last 40 minutes of data go right into the pump or Guardian. The latest Medtronic REAL-Time systems allow the user to customize the various alert settings by time of day – a nice feature for those who want to be warned of highs/lows more aggressively at some times of day compared to others. Like Dexcom, Medtronic offers alerts for high/low glucose levels and rise/fall rate of change. But unlike Dexcom, Medtronic also offers predictive alerts – if the system predicts a high or low threshold will be crossed, it informs the user. Although this feature can generate many “false positive” alarms, it can be helpful to those who must be extremely vigilant about preventing hypoglycemia.

When viewing trend graphs on-screen, the user has the option of scrolling back in time to view specific data points. Advanced statistics, such as average, standard deviation, and “time in zone”/“area under the curve” can be generated right on the pump/receiver, a feature absent on the Dexcom. Data from Medtronic REAL-Time downloads through the Internet to a website called “CareLink,” which produces some excellent graphs, charts, and statistical analyses. The download requires a simple USB plug and play device that works on Macs as well as PCs. When using a sensor-augmented pump, the CGM data are combined with pump data (basal and bolus delivery, carb entries, fingerstick readings) to create more comprehensive reports.

However, there are several downsides to the Medtronic REAL-Time system, beginning with complexity. Starting a new sensor is a multi-phase process that requires almost as much luck as technical skill. Error messages often occur during the early stages of sensor usage. Calibrations should only be entered when glucose levels are in a steady state and must be performed on a certain schedule, or the system stops generating data.

Medtronic’s Sof-Sensors can be a little more uncomfortable to insert, using a long, heavy-gauge introducer needle. Extra tape is needed to hold the sensor and transmitter in place from the get-go. The sensor longevity falls far short of Dexcom’s, rarely lasting more than a week (the Sof-Sensor is FDA approved for three days). Regardless of how long the sensors continue working, they tend to be reasonably accurate only during days two to four or two to five. Each time I have used a Sof-Sensor beyond five days, the accuracy took a major nosedive despite proper calibrations. In my experience, the average difference between Medtronic Sof-Sensors and my simultaneous fingerstick calibrations has been about 18%.

In terms of system features, the alerts and alarms on the REAL-Time systems are not strong enough to be noticed by many users. This includes both the vibration and the beeps. The display screen is smaller than Dexcom’s, and the contrast is not very good. And the transmitter requires re-charging at least once every 6-7 days, or it stops working.

Medtronic’s new Enlite sensors, which are pending FDA approval in the US, have a shorter and much thinner introducer needle, and insert at a 90-degree angle to the skin. There is also extra tape on the sensor to help keep the transmitter (the same MiniLink transmitter used with the Sof-Sensor) firmly attached. Using the new Enlite insertion device, the new sensors are considerably easier to insert and more comfortable than the Sof-Sensors. I would rank the Enlite sensors equal to Dexcom’s in terms of comfort and insertability.

However, for me, the accuracy of the Enlite sensors was not satisfying enough. Over a three-week period, my Enlite sensors averaged a 19% variance from the fingerstick calibrations (Editors note: in clinical trials, the Enlite has averaged a 14% variance from finger stick calibrations, and even better accuracy when used in conjunction with the Veo insulin pump). Unlike Sof-Sensors, Enlite tended to err on the low side rather than the high side. While this may be beneficial for those trying to prevent hypoglycemia, it can produce enough false low alarms to discourage the average user from fully trusting the device. (Editors note: for another user’s experience with the Enlite, see our Test Drive in diaTribe issue #32).

So there you have it: My “n of 1” experience with the currently available CGM systems. As you can see, there are pros and cons to everything. That’s where your personal preferences come into play. What’s really important to you? Accuracy? Convenience? Cost? Comfort? Loudness? Whether you’re looking to use a CGM for the first time or are ready to upgrade from your current system, there are many options to consider. As for me? I’ll just stick with my*