CGM in Type 2 Diabetes: Invaluable or Too Expensive?

Akshay B. Jain, MD


July 21, 2020

The use of continuous glucose monitoring (CGM) in patients with type 2 diabetes (T2D) is one of the myriad practical conundrums faced by those of us in clinical care. The 80th Scientific Sessions of the American Diabetes Association featured a debate to address this topic. Let's first review the arguments made by each side, and then I will share my own thoughts and suggestions.

'CGM Should Be Made Widely Available'

Athena Philis-Tsimikas, MD, was the first to speak, taking the position that CGM should be made widely available to everyone with T2D. CGMs are currently approved by Medicare and multiple private insurers only for patients on insulin therapy using multiple daily injections or insulin pumps.

Philis-Tsimikas began her presentation by showing the grim cost of healthcare resources spent in the management of diabetes, which is currently stated to be several hundred billion dollars worldwide, with a bulk of the costs going toward the management of poorly controlled diabetes and the complications that ensue. This picture is only set to get worse, with the latest data from such places as California showing that 55% of the adult population now has prediabetes or overt diabetes.

She described studies that showed that patients with T2D not on prandial insulin had a reduction in A1c by 1% when they used real-time CGM intermittently — even many months later, when there was no CGM on board. She also discussed smaller studies that showed improvement in physical activity levels and exercise adherence when CGM was used, even in patients with diet-controlled/non–insulin-requiring T2D, as well as those with prediabetes or early diabetes.

In a more real-world example, she described how one of her own patients with T2D and neuropathy started using CGM. He used the data to make significant lifestyle modifications, lost more than 50 pounds of body weight, and successfully came off almost all of his diabetes medications. Another patient challenged the CGM by eating a box of candies. The resulting hyperglycemia that he was able to visualize on the device was so impactful that he made lasting changes to his diet, leading to better glycemic control.

Her final point was how CGMs were being integrated in diabetes education classes for patients with T2D, with a small study suggesting decline in estimated glycemic variability and improvement in time in range, postulating that these can be helpful tools in newly diagnosed patients with T2D.

'CGM in T2D Is Not Cost-Effective'

Switching gears, Elbert Huang, MD, MPH, started out by lucidly explaining the concept of cost-effectiveness analysis. This is the measure that determines how to best spend the last dollar available for healthcare by identifying alternatives that maximize the health of a person within budget limits. He mentioned that unfortunately, the healthcare share of gross domestic product (GDP) in the United States outstrips GDP growth, and given the increasing number of patients being diagnosed with T2D, the projected direct Medicare spending on patients with T2D in the past decade alone has grown exponentially.

Cost-effectiveness analysis takes short-term studies and puts them into simulation models to understand long-term costs and projected outcomes. Using the DIAMOND trial as an example, Huang showed how the quality-adjusted life-years (QALYs) would improve in the CGM group by about half a year (which is quite significant), but the incremental cost-effectiveness ratio (ICER) would be $98,000 higher in the CGM group. If the cost of CGMs was to decrease, however, or the life of the sensor extended or use of a separate reading device eliminated, the ICER could be significantly reduced to $31,000-$42,000.

Using similar models, two studies in T2D showed that an annual CGM cost of $4808 was estimated to produce an ICER of $198,453 per QALY, whereas an annual CGM cost of about $630 would drastically reduce the ICER to $13,030 per QALY , making it significantly more cost-effective.

In these times, with the nation battling a pandemic, it becomes difficult to justify spending limited healthcare dollars on CGM over other measures, such as increasing access to healthy foods or use of medications like angiotensin-converting enzyme inhibitors and angiotensin receptor blockers, which are immensely more cost-saving on the basis of achieved outcomes.

My Take

In an ideal world, every patient with T2D would benefit at least a little with the information provided by CGM. There is no arguing that these devices have changed the way we manage diabetes and have quickly become an important tool in our armamentarium.

However, as with any new development, there is significant inertia between the commercial availability of these devices and their widespread accessibility, with the journey between the two often being long and frustrating for both healthcare providers and patients. At these times, it is important to look at the value these new developments would bring and whether they should be used for only selected populations or the masses.

For chronic conditions such as diabetes, there is power in data. When patients are given the opportunity to analyze data and understand how their body responds to different foods, different behaviors, and medications, they can take appropriate steps to optimize their diabetes. Despite the current literature suggesting that patients on CGM show significant improvement in glycemic control, the current reality of the situation is that these devices are not covered in the early stage of diabetes but rather much later, when patients are already on multiple daily injections of insulin or on insulin pumps.

At the same conference, a study was presented showing that use of the Freestyle Libre CGM (Abbott; Abbott Park, Illinois) led to a reduction of A1c by 0.8% at 6 months and 0.6% at 12 months (with the greatest decrease in A1c of 0.9% and 0.7%, respectively, in the non–insulin-requiring group). Other studies have shown decreased rates of hospitalizations and diabetic ketoacidosis in patients with T2D, including in those not on prandial insulin.

So What Can We Do?

The United States has the highest per capita spending on healthcare in the world, and this is only set to get worse, given the ramifications of the ongoing pandemic. Rational use of resources is extremely important. Are there ways to decrease the end-user cost of CGM devices so that the cost-effectiveness becomes more favorable and payers are more amenable to approving CGMs for varied populations?

On the basis of the new data we have seen, one option could be to identify the T2D population most at risk for glycemia-related events—such as worsening A1c, recurrent hospitalizations, and episodes of severe hypoglycemia—and utilize CGM for those patients first, regardless of whether they are on multiple daily insulin or continuous insulin infusion.

Another option might be a results-oriented coverage system, wherein patients are given initial coverage for 6 months of CGM use, and ongoing coverage is authorized only if they show a meaningful decrease in A1c levels by incorporating behavioral modification. Further coverage can be authorized as long as they continue to maintain good glycemic control, taking into account the increase in A1c associated with longer duration of diabetes.

Of course, healthcare professionals need to be educated on how to productively utilize CGM data to make meaningful changes to medications and counsel patients. This can be time-consuming and may dissuade the busy provider, but with regular practice, interpretation of ambulatory glucose profiles becomes quite simple and can be effectively streamlined.

Perfection of artificial intelligence that can sift through the information collected by CGMs (eg, glycemic trends based on time of the day, weekdays vs weekends, response to alcohol, exercise) and come up with reliable summaries and suggestions for patients would increase the patient's self-awareness and decrease the need for frequent clinic visits, thus reducing healthcare costs in the long run.

Finally, one of the simplest yet cost-effective methods would be utilization of CGMs on an intermittent basis. For instance, patients can use a sensor for about 14 days before their clinic appointment so that we have a better idea of glycemic excursions, and then use another sensor shortly after the visit to see whether the medication and/or behavioral changes are making a meaningful impact. This way, patients use one or two sensors every 3-6 months, and both the clinicians and patients can gain deeper insight into their diabetes control.

We are only in the nascent stage of what could be limitless benefits derived from CGMs. Rational and well-thought-out approaches at this stage are essential to ensure that the use of and access to these devices continue to expand, so that they can benefit everyone with dysglycemia.

Akshay B. Jain, MD, is a clinical endocrinologist who has practiced in three countries, focusing on mitigating the complications of diabetes and obesity. He is fluent in six languages and has spoken at more than 500 programs internationally.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.