Diana Issacs, PharmD, BCPS, BCACP, BC-ADM, CDCES

There are many challenges to the self-monitoring of blood glucose. One day, it’s going to be archaic, like how people used to have to test their urine for glucose. We couldn’t imagine that people would be peeing in a cup now and using that to test their glucose. Self-monitoring of blood glucose requires blood, which isn’t always fun. You have to use a lancing device, which is sharp and can feel painful. You have to draw a drop of blood. You have to carry supplies: test strips, a glucose meter and a lancing device. It’s pretty inconvenient.

It provides only one number at one point in time, so you don’t know the direction that the glucose is going. You could get a reading of 110 mg/dL, and someone could feel confident that their glucose is in the target range, but what if it’s dropping quickly? Unfortunately, that’s something that a blood glucose monitor isn’t able to tell somebody. However, a CGM can provide information about what the number is at that time along with the direction that it’s going in and … predict impending high or low glucose values.

CGM has many benefits and can improve disease management in both Type 1 (diabetes) and Type 2 diabetes. One thing CGM does very well is provide instant feedback on lifestyle choices. Someone could eat a cup of oatmeal and immediately see how that impacts their glucose value. I frequently have patients who tell me: “I noticed that when I eat oatmeal, it spikes a lot. I thought oatmeal was healthy compared with eggs. Eggs don’t raise me up very much.” Things like that are very helpful along with the instant feedback about the benefits of exercise, such as walking after a meal.

There’s also the safety component. A lot of people are scared about going low — rightfully so. It’s very common that people might eat an extra bedtime snack because they’re worried about dipping low overnight. A CGM has a low glucose alert feature, so they can have the confidence that it’s OK for them to go to bed with their glucose in the target range. If it drops, they’ll get alerted and will be able to take action on that.

In terms of the health care team, the data are invaluable. If all you have is an A1C (glycated hemoglobin) reading and a few fingerstick readings, it’s like flying blind. You don’t know exactly what to do to the treatment regimen. If A1C is above target, you know you need to adjust something but you don’t know (whether) the person is having lows. With a CGM, you have all this information. You can determine exactly when someone isn’t in target range and then have a more targeted intervention to help them increase time in range and obtain their A1C target.

There are a lot of data that show the benefits of a CGM in terms of improving glycemic management. We have (several) randomized controlled trials with a real-time CGM, thinking back to the old DIAMOND trials in Type 1 (diabetes) and Type 2 diabetes. We also have (several) randomized control trials with intermittently scanned CGM. I’m not going to go through all the data but the bottom line is that these trials showed that there were decreases in A1C as well as increases in that time in target range.

Time in range is becoming an important metric. That’s the time spent between 70 and 180 mg/dL. The reason this is becoming so important is because we know we can bring down A1C. A1C is an average, so we can bring it down by causing tons of hypoglycemia. Obviously, that isn’t good clinical management and won’t lead to good clinical outcomes. By increasing time in range we can still achieve the A1C target, but in general we’re going to have safer outcomes and people are going to feel better because they’re not going to be on this roller coaster of high and low or having tons of lows.

Other important things we’ve seen from the clinical trial data are improvements in quality of life and less diabetes distress, which is so important. There have been (several) cost-benefit analyses showing that through the use of CGMs and improvements in quality of life — as well as other important end points like reducing hospital admissions for severe hyperglycemia and reducing emergency department admissions for hypoglycemia or hyperglycemia — there’s a huge cost benefit to using these devices.

When we think about the burden on health care professionals, we know that we are being asked to see more patients. It can be challenging to take care of people with diabetes, especially if you have a 20-minute slot with their visit. CGMs have helped people reach their A1C targets and other metrics. The reduction in hospital and emergency department admissions is important and reduces the burden for health care professionals.

A CGM could increase the burden if it isn’t implemented in a systematic way. That’s not a great thing to say. I’m a big advocate of the use of CGMs, but the reality is that you’re going to get more data, so your team has to have a process of who will look at the data, how you’re going to download the data and what you’re going to do with it. Sometimes that involves a bit of training up front for the team, or at least figuring out who on your team is going to be the technology champion. Fortunately, there are a lot of people who want to be technology champions. When you think about the diabetes care and education specialist or the pharmacist, dietitian, nurses or your MAs (medical assistants), who on your team can you utilize who’s interested in this space and wants to develop more expertise that can help you? That way you can implement good practices so that it doesn’t feel like an extra burden and is helping your practice provide the best care for your patients.

When we think about patient factors to consider for CGM use, the patients’ preferences are always my top concern. I want patients to have choices and realize what their options are. We have four options on the market and I want them to understand some of the differences, like alarm capabilities. Some have predictive alerts, so you can predict hypoglycemia in the next 20 minutes. Others don’t and have simpler alerts for highs and lows. Some have the option to turn all alerts off. Others don’t have that option. Some are implantable; some aren’t. The point is that people need to understand the differences so they can make an informed choice about what’s going to be the best option for them. Another example is calibration. Some require fingersticks and some don’t. You want people to know and understand those things and see what will work best for them.

Is there a subset of patients for whom we wouldn’t recommend CGM? It’s an interesting question. I come back to choice. I had a patient earlier this week who has Type 1 diabetes and has been checking with fingersticks four times a day and is doing great. He’s reaching his A1C targets. He’s doing well. He said he doesn’t want to wear a CGM. He’s concerned that he might feel overwhelmed by the data. He feels that what he’s doing is working very well for him. I respect that choice. In the future, if he’s not meeting his treatment targets, I may try to … encourage him to try it, but he’s doing well now. It’s not something we have to force everyone to do.

I find that with most people, even if they’re reluctant to start it — maybe they don’t want to wear something on their body or they don’t like this idea of all the data — once they try it out, they usually realize, “This was a lot smaller than I thought and I don’t have to look at the data all the time. I can look only when I want to.” Many people do very well with it.

In terms of providing education when recommending CGM, there are a lot of websites I utilize. I like to provide initial education about the different options in a very objective way and then I refer to websites like https://diabeteswise.org, which is a great site. It’s a nonindustry-funded site that objectively goes through the options and asks people questions about what’s important to them so that they can figure out what might be the best technology for them. I also recommend websites like https://diatribe.org, which is up to date with cutting-edge technology. If there’s a … CGM innovation, we know it will be the first to report on that. I like to send patients to those sites to do a little additional reading for follow-up before they (decide).

Utilizing CGMs for Diabetes Management in COVID-19

CGMs have been instrumental during the COVID-19 pandemic. We’ve all had to pivot to virtual care. We were fortunate. At Cleveland Clinic, we were doing virtual visits long before the pandemic but we do (them) much more frequently now than we ever did before. What’s nice about a CGM is that the data with most of these devices are Bluetooth connected, so it’s going into a portal that can be viewed … instantly to see how someone is doing. Even the data that aren’t automatically going into a portal have the capability to be plugged in and downloaded. In terms of virtual care, we’re able to view those data. It’s a lot more informative than if we only had an A1C value.

One of the unique things about a CGM is that we have all these different … key metrics. In addition to time in range, we have the glucose management indicator (GMI). That’s essentially an estimated A1C. I’m able to use that to track how someone is doing and see (whether) they’re meeting their treatment targets. During COVID-19, that’s been very helpful because a lot of people don’t want to come in. They don’t want to go to the laboratory. I have people who haven’t been to the laboratory in a long time, but I still have a very good objective assessment of how they’re doing. In fact, there’s a lot of movement to see (whether) GMI and time in range could overtake A1C or at least be supplemented with A1C so we’re not relying on this one metric every three months for evaluating diabetes care.

When we think about CGM accuracy, they’ve all shown that they’re pretty accurate. We have a term we use to describe the accuracy: mean absolute relative difference (MARD). The lower, the better. According to the Food and Drug Administration, when MARD is below 10%, the device is accurate enough to base treatment decisions on. That would include treating hypoglycemia or determining insulin doses based on the numbers. We feel pretty comfortable with their accuracy. It continues to improve.

One thing we have to be aware of is that there’s a slight lag time with CGM. The gap is closing, so the lag time isn’t what it was like 10 years ago. But because fingersticks are measuring the capillary and CGM is measuring interstitial fluid, there’s a slight delay. That means that if somebody recently ate something, the glucose on the fingerstick might be a little higher than it would be on the CGM. Similarly, if someone recently exercised or … gave an insulin injection, that number may be a little different. Educating people on those differences is very important.

Beyond that, calibration is sometimes a factor that goes into device selection. When you calibrate, you’re essentially doing a fingerstick and entering (the result) into the device to calibrate it. Any time you do that, you could be entering in some error, because we know fingersticks aren’t perfect. If you have anything sticky or sweet on your fingers or if you didn’t get a good enough sample, that can skew the reading. If you put a bad reading into the device, you’re probably going to skew the accuracy. That can be a determination. Devices that are factory calibrated are often preferred because you have less interference and it’s easier for the person because they don’t have to do additional fingersticks.

CGMs have evolved so much over the (past) several years. It has been amazing. Some big advancements are regarding calibrations. It used to be standard that all CGMs required fingersticks, and now we have factory-calibrated options that don’t require any fingersticks or have them as optional. There have also been impressive advancements with wear time. One of our devices goes up to 14 days. Another is implantable and goes up to 180 days. Wear time has evolved. It used to be about three days to start.

The other area where we’ve seen a lot of advancement is with integration. We have CGMs that integrate with insulin pumps to have automated insulin delivery, which is incredible. We also have CGM integration with connected insulin pens, which can also use those data and be able to view it in one report, which is remarkable. Several mobile apps can link in with these CGM data. Accuracy has improved a lot over the years, and we’re seeing that the sensor is getting smaller.

Despite all of the amazing advancements that we have had in the CGM space, there are still some unmet needs in research and clinical treatment. I do a lot of management in pregnancy, and we lack clinical studies in pregnant women with preexisting Type 2 diabetes or true gestational diabetes, where women develop diabetes during their pregnancy.

It would also be valuable to collect more information on the use of CGM and the outcomes in people with Type 2 diabetes who aren’t taking insulin. Last year, we had the MOBILE study, which studied Type 2 diabetes on only basal insulin and showed an impressive A1C reduction in that population. I’d like to see more on the complete noninsulin users. The other area where we need more research or best practices is the implementation of CGM. Many of us are sold on CGM being a great tool. However, how does a busy practice suddenly implement this tool? How do they get their team up to speed on reviewing these reports that look different from typical glucose meter readings? We need more of that so we can efficiently get this into practices across the country and world.

This article originally appeared on Medical Economics.

Articles

Continuous glucose monitoring has many benefits and can improve disease management in for both diabetes types. 