Background: a continuous glucose monitor is a device that is worn on the body, and has a sensor inserted by the patient beneath their skin. It converts a measurement of interstitial fluid into a blood glucose reading. That data is transmitted to a device (either an insulin pump or a separate reader carried by the patient) and alerts them to high or low blood glucose. Perhaps more useful is it can alert to a rising or falling blood glucose reading, letting the patient take appropriate action before the situation becomes unpleasant. While it is helpful at spotting trends and treating them appropriately, it is critical to some patients at night. When sleeping, some might not ever know if their blood glucose is trending downward to dangerous, if not fatally low levels. A CGM spots this and puts out an alarm that lets the patient treat it accordingly. It has, and does save lives. Worth noting is that the FDA publicized a recall of the Dexcom G5 system because the alarm function did not work. In fact, the recall stated, "The FDA has identified this as a Class I recall, the most serious type of recall. Relying on this device may cause serious injuries or death"
There are several things that trouble me about using a CGM to give the data for making insulin dosing decisions. A dosing decision is a complex proposition - doing it safely requires accurate blood glucose data. Before a patient injects insulin before a meal, they calculate part of the insulin dose to counter the quantity of carbohydrates they will eat. Another part of the dose accounts for how high or how low their blood glucose is above the desired range. That requires a blood glucose test, then the patient calculates how much insulin must be added or subtracted to the amount for the carbohydrates. The sum is how much insulin they will take from a shot or an insulin pump. With this proposal, the CGM will provide that blood glucose number, not a blood glucose testing meter.There are a number of reasons why this is a bad idea.
- Current Dexcom G5 Mobile safety literature advises patients not to make treatment decisions based on the CGM.
- The Dexcom G5 must be calibrated 2 times a day using a blood glucose meter. Calibration is used to ensure that the algorithm in the system accurately converts the glucose found in interstitial fluid to a reasonably accurate blood glucose value. I calibrate my CGM before meals when glucose levels are most likely to be stable. Doing one more finger stick for the third meal is a small price to pay for greater dosing safety and accuracy.
- The purpose for supporting this expansion of use is so that patients relying on Medicare will be (perhaps) be able to use the Dexcom device. One online advocacy group (diaTribe) also states that approving this for Dexcom will (perhaps) speed approval for Abbott’s factory calibrated FreeStyle Libre. There is a lot of supposition underlying their cause to action.
Here is where I get really annoyed.
Back in May, United Healthcare (a large health insurer) announced that Medtronic would be their "preferred provider" of insulin pumps and Continuous Glucose Monitors (CGMs). They justified this by saying they "are working together to make a widely used and effective product available to our members while pursuing new ways to lower the overall cost of living with diabetes. Patient safety, service and cost were key considerations in our decision-making process to work with Medtronic, who will be designated the preferred provider of insulin pumps for UnitedHealthcare Commercial and UnitedHealthcare Community Plan (Medicaid) members starting on July 1." (You can read more about announcement here.)
Social media leaders in the diabetes community, bloggers, advocates and patients were up in arms. Their objections were twofold. One was that this decision kept patients from having a choice. No longer could they choose an insulin pump and/or CGM that they wanted. The other was that denying this choice was an impediment to access of devices. If United Health's selection of Medtronic devices didn't meet their needs, the alternative was no insulin pump or CGM, or pay for it and the expensive supplies out of pocket. They fought, they yelled, they rocked social media and regardless of the comparative merits of their position, but they have yet to prevail. Underlying the fight was the notion of the little guy versus big insurance and big device manufacturing conglomerate.
Just two months later, many of the same advocates, and some heavy hitters in the diabetes industry are now waging a different battle. This one isn't designed to promote 'choice,' but seemingly to get a government endorsement of an unproven treatment technique and to have it available only to users of one device. In other words, they are supporting a government endorsement of one device other others with a similar level of accuracy (Medtronic Enlite). Really? Where is 'patient choice' in that? As for being scientifically based, as opposed to supporting 'greater access,' I've hunted far and wide to find any double blind studies substantiating the safety/efficacy of Dexcom dosing? Dexcom itself says not to do that. The FDA says not to do that. But the diabetes online community thinks it is a good idea. Why? It makes some feel like they made a difference in bringing CGMs available to people who might otherwise be unable to afford one. (I support that, but not at the expense of safety.)
What is left unsaid in the discussion is the money at stake. If Dexcom gets approval for this, they stand to make literally billions of dollars. Their competitors won't. Doctors will be able to bill more to support patients using the device. Certified Diabetes Educators will also clean up as they train and counsel an explosive growth of new patients.
While this is good for doctors, nurses, CDEs, and patients, I'm not sold. At all. My experience with CGMs (from the very first Dexcom, to their second generation model, to the Medtronic SofSensor and the new Enlite) runs over 7 years. I think I've got a handle on this. Why? My experience indicates that:
- CGMs can, and often do vary significantly from the blood glucose they show the user versus what the blood glucose meter values.
- For some of us, particularly those of us with volatile swings in blood glucose levels, the 20 minute lag between CGM estimates versus real time metered blood glucose readings is important.
- CGMs can also be erratic, both in the blood glucose values and trends they display, as well as being substantially different from metered blood glucose levels.
Please write the FDA and tell them that this is an
idea whose time has not yet come.
You can email firstname.lastname@example.org and put
"Public Comment: Advisory Panel on Dexcom CGM use for insulin dosing."