Use of CGM metrics more common in diabetes care, but still limited
Click Here to Manage Email Alerts
With uptake of diabetes devices continuing to expand, more providers are shifting away from looking solely at HbA1c and toward analyzing continuous glucose monitoring data.
People with diabetes using wearable CGM are able to share device-generated ambulatory glucose profiles with physicians and other care providers. These 14-day reports calculate the amount of time a user spends within a glucose target range of 70 mg/dL to 180 mg/dL. The report also includes the time spent in hypoglycemia or hyperglycemia, a user’s mean glucose level, glucose variability and a glucose management indicator, which shows the expected mean HbA1c during that period.
Photo by Anna Klonoff. Printed with permission.
“The advantage is that you can get a sense of what’s recently happening. The CGM applications have an ability to give the patient weekly updates on what their time in range is,” Anne L. Peters, MD, an Endocrine Today Editorial Board Member and professor of clinical medicine at Keck School of Medicine, University of Southern California, told Endocrine Today. “My patients who have the numeracy skills to understand it love to see their time in range. With proper education, patients can use time in range as a success tool.”
Many providers said they believe that time in range can paint a more accurate picture of glycemic management for a person with diabetes, but uptake among some physicians is low. Survey data reported by the Time in Range Coalition at the Association of Diabetes Care & Education Specialists 2022 annual meeting revealed that fewer than half of primary care physicians use time in range in discussions with their patients with diabetes, compared with more than 90% of endocrinologists. One reason for the limited uptake is likely the lack of CGM use in primary care clinics and among people with type 2 diabetes.
“We still have a long way to go,” Diana Isaacs, PharmD, BCPS, BCACP, BC-ADM, CDCES, FADCES, an Endocrine Today Editorial Board Member, clinical pharmacy specialist and CGM program coordinator in the department of endocrinology, diabetes and metabolism at the Cleveland Clinic Diabetes Center, told Endocrine Today. “CGM use is increasing, it’s becoming more accessible and more affordable. But still, the majority of people with type 2 diabetes are not using CGM. Time in range is most useful with CGM data because people doing finger sticks don’t have seven times that they are checking per day, usually it’s one or two. Time in range isn’t going to be that useful with one or two readings per day.”
Some experts said they believe that analyzing a CGM’s ambulatory glucose profile is too cumbersome and time-consuming for providers. A new metric called the Glycemia Risk Index (GRI) was developed earlier in 2022 with the idea of incorporating all the metrics included in a CGM report into one number that can provide a more accurate representation of a person’s glycemic control than time in range alone.
“I had always felt that when we looked at time in range, we were leaving something out,” David C. Klonoff, MD, medical director of the Diabetes Research Institute at Mills-Peninsula Medical Center in San Mateo, California, told Endocrine Today. “It’s ironic that you’ve had this rule that says, ‘Go beyond A1c,’ which makes sense because HbA1c has two problems: It’s got hemoglobin, which is a molecule that has nothing to do with glucose, and it only covers midrange. Some have replaced it with time in range, what many people say is all that is needed for describing the quality of a CGM tracing but still has the problem of only looking at the midrange.
“The first mantra said, ‘Go beyond A1c,’” Klonoff said. “I think the next one will be, ‘Go beyond time in range.’”
HbA1c vs. time in range
Even with the advent of CGM and new diabetes technology, HbA1c remains the most common metric used in diabetes care. HbA1c became the primary metric for analyzing glucose control almost 30 years ago, with the publication of findings from the Diabetes Control and Complications Trial (DCCT). In the trial, adults with type 1 diabetes were randomly assigned to conventional treatment or intensive therapy. During a mean 6.5 years, those receiving intensive therapy achieved a median HbA1c of about 7%, whereas the conventional treatment group had a median HbA1c of about 9%. Compared with conventional therapy, those receiving intensive therapy had a reduced risk for developing microvascular complications of 35% to 76%.
“Up until that point, there was a debate as to whether or not complications of diabetes were related to glucose levels or if there were some major genetic etiologic factors in why certain people got complications and others did not,” Robert Vigersky, MD, chief medical officer at Medtronic Diabetes, told Endocrine Today. “The observation was we know there are people who have high blood sugars all the time who didn’t get any complications. So how could it be glucose? The DCCT study proved once and for all that glycemic control matters, and the HbA1c was used as the metric.”
Subsequent studies built on the findings in DCCT correlating HbA1c with the risk for diabetes-related complications. However, HbA1c is not perfect.
“HbA1c has problems because some people make more hemoglobin than others,” Klonoff said. “Time in range is better than HbA1c, but it’s because of the hemoglobin problem inherent in depending on HbA1c; it’s not because it’s telling you something different about the patient. Time in range gets rid of the hemoglobin problem, which is an important issue.”
Perhaps the biggest shortcoming of HbA1c is its inability to provide up-to-date data on a person’s glycemic control. HbA1c reflects mean glucose level over 2 to 3 months, a long period during which a person with diabetes may experience a lot of glucose variability.
“HbA1c is really easy to use in practice, but that does not mean it’s the right thing to use,” Isaacs said. “It doesn’t tell you anything about hyperglycemia, hypoglycemia or glucose variability. It’s easy, you can look at your patient and they have an HbA1c of 6.7%, and you can say everything is going great, we’ll just continue present management. But what that doesn’t tell is if there is any hypoglycemia. Many times, I’ll put CGM on these patients and see they are going low overnight, or they are going low in-between meals, and that’s not safe.”
CGM metrics and diabetes-related outcomes
Recently, researchers have correlated time in range with HbA1c. In a study published in Diabetes Technology & Therapeutics in 2019, Vigersky and Chantal McMahon, PhD, MBA, paired contemporaneously obtained HbA1c and time in range from 18 studies of people with type 1 and type 2 diabetes. They found an excellent correlation between HbA1c and time in range, with every 10% change of time in range associated with a 0.8% change in HbA1c.
HbA1c remains the primary outcome metric for diabetes clinical trials, but numerous studies have explored whether time in range can also be linked to diabetes-related outcomes. A study published in Diabetes Care in 2019 examined DCCT data from seven finger-stick samples collected 1 day every 3 months from 1,440 participants. The study, conducted by Roy W. Beck, MD, PhD, president and medical director of the Jaeb Center for Health Research Foundation, and colleagues, found that the risk for the development of diabetic retinopathy progression increased 64% and the risk for the development of microalbuminuria increased 40% with each 10% decrease in time in range.
In another study published in Diabetes Care in 2018, researchers analyzed CGM metrics from 3,262 people with type 2 diabetes in China. The study cohort was divided into quartiles based on time in range. Participants with less time in range had more advanced diabetic retinopathy, and time in range was significantly correlated with all stages of diabetic retinopathy.
A more recent study, published in The Journal of Clinical Endocrinology & Metabolism in 2021, examined associations between glycemic variability and time in range with diabetes-related complications. The researchers found lower time in range was associated with a higher risk for microvascular complications and hospitalization for hypoglycemia or diabetic ketoacidosis.
“This study was just following these patients for 24 months, and this to me was sort of the home run in terms of hopefully convincing, at least the scientific world, that we can use CGM metrics instead of HbA1c,” Irl B. Hirsch, MD, professor of medicine at the University of Washington School of Medicine in Seattle, told Endocrine Today.
Few studies have examined time in range and associations with patient satisfaction, but providers said they believe that there is a strong link between the two — patients are motivated to improve their glycemic control when they can view their time in range in real time.
“CGM teaches and reinforces healthy behaviors,” Isaacs said. “Also, in terms of medication-taking behaviors, if someone forgets to take their medication, they see they’ll be more out of range.”
“I see the evolution in my patients,” Hirsch said. “Whether you’re using Dexcom with Clarity or the FreeStyle Libre with LibreView, the patients can actually see the time in range on their phones. They can look at their time in range and glucose management indicator whenever they want to.”
People with diabetes using CGM have more data at their disposal than ever before, but it is not necessarily easy for them to understand.
“I’ve had patients who I put on a CGM that overdosed their insulin,” Peters said. “I’ve had patients who ended up in the emergency department because they overreacted to the data. I’ve had patients who saw the data and stopped giving insulin for reasons unclear to me. I’ve had all sorts of weird reactions to seeing data. You can’t do any of this without educating the patients and educating the providers.”
New metric simplifies CGM data
Understanding the data is a concern for many researchers. Klonoff said it may be difficult for clinicians to look at all of the metrics provided by an ambulatory glucose profile and distill the information quickly. Although time in range has its benefits, Klonoff said, looking at that metric alone and ignoring other CGM data painted an incomplete picture of a patient’s glycemic profile. Specifically, time in range lacks specific data on time spent in severe hypoglycemia and severe hyperglycemia as well as glucose variability.
“There’s so much information, it can be difficult to put it all together and answer the two questions: How is the patient doing, and, if they’re not doing well, what needs to be done,” Klonoff said. “Those to me are the most important questions that the physician or the clinician asks. When you have seven different numbers and ranges presented in an [ambulatory glucose profile] report, it can be difficult to put it all together.”
To combat this problem, Klonoff and dozens of other diabetes experts came together to create the GRI. The new metric was unveiled in an article published in the Journal of Diabetes Science and Technology in April. GRI incorporates seven metrics present in a 14-day ambulatory glucose profile. The formula incorporates time in very low range, in low range, in high range and in very high range, as well as mean glucose and glycemic variability, to calculate hypoglycemia and hyperglycemia components. Those two components are added together to create the GRI, which is a number ranging from 0 to 100. A higher number indicates a worse glycemic profile. Klonoff said providers can look at a GRI and quickly decide whether they need to take action regarding therapy.
Peters, who was a co-author on the paper, said GRI could have multiple positive implications in clinical practice.
“It’s helpful to me because it speaks to the point that a patient can have a good time in range, but still have too much time below range,” Peters said. “This tries to pull out the people who need the most attention.”
The development of GRI is the latest attempt to develop a metric that improves on HbA1c and other currently available CGM metrics. In 2017, Vigersky was the lead author of a paper published in the Journal of Diabetes Science and Technology that proposed a “glucose pentagon,” a five-pronged metric that incorporated time in range, mean glucose level, hyperglycemia intensity, hypoglycemia intensity and glucose variability. Vigersky said GRI could be a useful metric but that it is too early to tell whether it will be accepted by both providers and patients.
“It remains to be seen whether or not that actually will have traction,” Vigersky said. “Can people understand it? How does what you do in modifying someone’s therapy change that GRI, and can you start thinking in terms of GRI instead of HbA1c or time in range? It’s an evolving area, a very interesting area, but we’re not there yet.”
Moving forward with CGM metrics
Many experts said they see the value in CGM metrics when compared with HbA1c, but several barriers still remain to their wide use. One is the lack of CGM uptake overall. Although CGM is being adopted more broadly each year, there are still barriers to access, particularly for people with type 2 diabetes.
“The [American Diabetes Association’s] new guidelines that came out this year based on the MOBILE study said that we now have evidence for using CGM in patients with type 2 diabetes with basal insulin only,” Hirsch said. “Many of the payers have not used that yet, and then many of the patients don’t have access to it, especially if they’re not taking insulin.”
Even if CGM access improves, persuading patients and providers to use the information it generates is another discussion. In data from the Time in Range Coalition, only 46% of PCPs reported they use time in range in clinical practice compared with 88% of endocrinologists. Additionally, only 56% of PCPs said they were even aware of time in range.
Klonoff said GRI could give PCPs an easy single number metric to encourage them to use CGM data in the future. Researchers are exploring whether a GRI could be calculated without the use of CGM.
“We’re now working with a group that has been doing seven-point glucose measurements on patients as opposed to CGM, and they’re now in the process of looking at the GRI. We are also working with a university clinic that has found that, compared to time in range, GRI correlates better with their clinicians’ impression of which patients need extra attention,” Klonoff said.
Isaacs and Vigersky said working with physicians to educate them and their staffs could allow CGM metrics to be adopted more broadly at the primary care level.
“There are many non-endocrinologists who don’t understand the technology and don’t have the time to look at and review the data, and unfortunately, there aren’t enough endocrinologists in the world, so not enough people are getting access to CGM,” Vigersky said. “Primary care still treats the vast majority of people with diabetes, and they just don’t have the knowledge and the time and expertise to understand CGM and the implications of CGM. They’re still stuck in the HbA1c world.”
“Each practice needs to develop a workflow,” Isaacs said. “I don’t think it’s realistic or fair for primary care providers to take on all of the new responsibility of doing this. Utilizing team members like nurses, pharmacists, dietitians and the diabetes care and education specialists is what is going to make it penetrate more.”
Awareness about time in range is also lacking at the patient level. According to the Time in Range Coalition, 44% of people with diabetes are aware of time in range, and 33% are actively engaged with the metric. Awareness and engagement with time in range is higher among people with type 1 diabetes.
Peters said education at the patient level should improve awareness about time in range and possibly encourage more providers to use CGM metrics in their clinics.
“Not every patient needs the same instructions. Some can just look at a website or YouTube video, and others need an educator,” Peters said. “But if you teach the patients about wearing these and the patients start feeling empowered, that’s where we can move the provider’s understanding of this.”
- References:
- Beck RW, et al. Diabetes Care. 2019;doi:10.2337/dc18-1444.
- El Malahi A, et al. J Clin Endocrinol Metab. 2021;doi:10.1210/clinem/dgab688.
- Heverly JK, et al. F30. Presented at: ADCES22; Aug. 12-15, 2022; Baltimore.
- Klonoff DC, et al. J Diabetes Sci Technol. 2022;doi:10.1177/19322968221085273.
- Lu J, et al. Diabetes Care. 2018;doi:10.2337/dc18-1131.
- Nathan DM, et al. Diabetes Care. 2014;doi:10.2337/dc13-2112.
- Vigersky RA, et al. J Diabetes Sci Technol. 2018;doi:10.1177/1932296817718561.
- Vigersky RA, McMahon C. Diabetes Technol Ther. 2019;doi:10.1089/dia.2018.0310.
- For more information:
- Irl B. Hirsch, MD, can be reached at ihirsch@uw.edu.
- Diana Isaacs, PharmD, BCPS, BCACP, BC-ADM, CDCES, FADCES, can be reached at isaacsd@ccf.org; Twitter: @DianaMIsaacs
- David C. Klonoff, MD, can be reached at dklonoff@diabetestechnology.org. Twitter: @DavidKlonoff
- Anne L. Peters, MD, can be reached at annepete@med.usc.edu.
- Robert Vigersky, MD, can be reached at robert.a.vigersky@medtronic.com.
Collapse
Read more about
Click Here to Manage Email Alerts