Continuous Glucose Monitoring in Primary Care: Understanding and Supporting Clinicians’ Use to Enhance Diabetes Care

Thank you very much for your interest and your questions, which are excellent and very discerning.

We do not know why higher proportion of Medicare-covered patients was associated with greater likelihood of continuous glucose monitor (CGM) prescription in our study. We expected to see greater likelihood of prescribing CGM with higher proportion of private/commercial insurance, which generally has a lower burden of coverage criteria and documentation to support it, and were somewhat surprised to find this association with Medicare. Right now we can only speculate (higher proportion of Medicare likely also means lower proportion of Medicaid, with Medicaid coverage criteria varying widely between states but almost universally more limited coverage than Medicare?), but ultimately this question begs for more research to understand.

As you mention, Medicare changed its coverage criteria for personal CGM in 2021. The major change was eliminating the requirement of performing any self-monitoring of blood glucose (via capillary or “finger stick”); until this change, four or more daily finger sticks were required to qualify. Data collection for this study was completed before Medicare’s 2021 expansion of access. Since then, though, we have seen and heard (and have unpublished data from a follow up study ongoing right now) that eliminating the finger stick requirement has made many more people with Medicare successful in acquiring personal CGM. With any Medicare coverage criteria change, other insurers often follow, sometimes sooner, sometimes later. This affects a substantial proportion of patients nationwide and should increase access to personal CGM.

However, the major barrier remains obtaining actual authorization for coverage, even when the coverage criteria are met: significant documentation must be submitted to substantiate one’s assertion that a patient meets the coverage criteria. This includes several statements about the various criteria as well as providing a letter of medical necessity and chart notes. CGM and diabetes are certainly not unique here; there are similar requirements for advanced imaging studies, medications with less than preferred formulary status, supplemental oxygen, and numerous others. In the real-world environment of primary care practice, compiling and writing such additional documentation places a further burden on already stressed and busy primary care teams. Some providers and practices expend the effort to complete the daunting authorization process or to appeal a notice of denial, and some simply can’t or don’t, for a variety of reasons.

But we are hopeful that changes in the authorization process will come and that further advances in electronic health record functionality will decrease the effort required to produce the necessary documentation. And as the evidence continues to accumulate that CGM has shown benefit in more people with diabetes in more circumstances, we are hopeful that the coverage criteria will become fewer and that the authorization process will therefore become more streamlined.

In that vein, based on the increase in available evidence, the American Diabetes Association (ADA) Standards of Care have expanded CGM recommendations in each of the last several years, most recently in January 2023. Even before that, in late 2022, Medicare completed a public comment period for further expansion of personal CGM coverage, proposing to remove the requirement for multiple (3 or more) daily administrations of insulin or insulin pump use and to replace it just with any insulin treatment (which would include even once daily insulin) or a history of “problematic hypoglycemia.” This problematic hypoglycemia provision would open the door for Medicare coverage of personal CGM for at least some people not even using insulin. This was proposed on the basis of expanding evidence, including studies demonstrating reduction in HbA1c, increased time in the target glycemic range, and reduction in hypoglycemia with personal CGM compared to self-monitoring of blood glucose. There was no opposition to this expansion during the publicly available session, and we are hopeful that access will again expand significantly in the coming weeks or months, in line with the updated ADA Standards.

We have difficulty imagining a requirement for CGM education; some of our primary care colleagues do not treat diabetes, e.g., those with practices limited to sports medicine, substance use disorder, urgent care, or occupational health, among others. But we agree that CGM education should be more available and more accessible for those who are interested. As coverage expands and IF the burden of documentation reduces, education content should also become more streamlined, focusing more on the clinical use of CGM rather than the administrative issues of where to send a prescription and how to get it covered.

We are very encouraged to hear from you, in particular, as undergraduates who aspire to careers in health care, and to know that you are already reading studies like this in journals like Annals of Family Medicine. This gives us great hope for our collective future. And finally, we wholeheartedly agree with you that more research is needed to understand and reduce disparities in care, to produce more evidence to understand where CGM is (and where it is not) helpful or appropriate, to enhance diabetes care wherever people get that care, and to improve quality of life for all people with diabetes.