In the last year, Jai Smith has cycled through 13 primary care doctors. Ever since being diagnosed with type 2 diabetes in 1995, she’s tried her best to manage a disease that has devastated her family: Her grandmother and four uncles died from its complications.
But she’s struggled to find a doctor in her hometown of Little Rock, Ark., who will give her what she wants to manage the condition: a continuous glucose monitor. Like many patients with diabetes, Smith uses fingerprick glucose tests to help dose her medications. The 44-year-old was immediately interested when she heard about CGM, which uses an embedded sensor to collect a proxy for blood glucose around the clock.
“I’m already having to prick my finger three to four times a day,” she said. “Immediately I’m like, ‘OK, I should be a shoo-in, let me ask my doctor.’”
But that doctor didn’t think it was a fit. Neither did the next one, or the one after that. Most, she said, gave her the same materials about nutrition and exercise, and occasionally tweaked her medications before sending her out the door with a follow-up appointment. Some told her that CGMs weren’t appropriate for people with type 2 diabetes; others told her there was no way the device would be covered by her Blue Cross Blue Shield insurance.
Smith’s frustration is becoming more common as CGM technology is exposed to patients with type 2 diabetes, the product of growing advertising budgets and a slowly growing evidence base. The job of fielding those requests falls mostly to primary care physicians, who lead care for patients with type 2 diabetes. But as demand grows, many of these doctors remain unprepared to interpret the evidence needed to prescribe them or navigate the minefield of securing insurance coverage.
That’s because up until now, the devices have primarily been a tool of endocrinologists, specialists in hormone regulation who first used them to protect against dangerous dips in blood glucose in type 1 patients. Most patients with type 2 diabetes will never see an endocrinologist, though. In Smith’s long run of doctors, she’s only seen a specialist once.
“It’s just a numbers game,” said Steven Edelman, an endocrinologist at the University of California, San Diego. There are only 8,000 or so working endocrinologists in the United States, and 34 million Americans with diabetes — about 30 million of whom have type 2 diabetes. Like Smith, about 90% of them receive their care from a PCP. As the evidence and demand for CGMs in type 2 diabetes grows, those physicians will need the resources and tools to get sensors into the right arms and bellies.
It’s only in the last few years that doctors have even considered CGM a viable option for patients with type 2 diabetes.
The tides began to turn in 2017, with the publication of the Dexcom-funded DIAMOND randomized, controlled trial. It looked at the effect of real-time CGM use for people with type 2 diabetes who were on an intensive insulin regimen: daily background injections, plus extra doses at meal times. Their A1C levels, a measure of average blood glucose, dropped by 0.3% more than those using standard blood glucose strips over six months — a small but statistically significant change.
Around the same time, devices from Dexcom and Abbott were getting smaller, cheaper, and more convenient to use. Patients with type 1 diabetes, who are especially at risk of dangerous low bouts of blood sugar, had been willing to put up with the finicky technology because it could save their lives. But the easier the devices got to use, the more it made sense to test their use in type 2 diabetes.
Medicare started covering CGMs for patients on intensive insulin in 2017, and by now, many endocrinologists will prescribe CGMs for type 2 patients who take multiple daily injections of insulin or use an insulin pump. The devices seem to help these patients keep their A1Cs in a healthier range — and prevent acute complications, just like with type 1. Recent claims-based research funded by Abbott has shown that six months of the company’s FreeStyle Libre, a “flash” CGM that collects readings with a swipe over the sensor, can reduce emergency visits and hospitalizations for type 2 patients using insulin.
“The people with insulin are the low-hanging fruit,” said Irl Hirsch, an endocrinologist at the University of Washington.
Research is also beginning to show that CGM could help type 2 patients on longer-acting basal insulin, who take their medication just once or twice a day. Earlier this month, researchers reported results from the MOBILE study also funded by Dexcom, a randomized, controlled trial of CGMs for people with type 2 diabetes on basal insulin. Unlike past studies which looked at patients being treated by endocrinologists, this study recruited patients from primary care doctors, who were then advised by diabetes specialists.
“They drew the population from primary care, because that’s really where these people are being managed in the real world in America,” said Thomas Martens, medical director at the International Diabetes Center and lead author of the study, which saw a statistically significant decrease in A1C over eight months with CGM compared to traditional fingerstick monitoring — a 1.1% drop instead of 0.6%.
Another study this month looked at real-world outcomes and found that CGM use among patients with type 1 and type 2 diabetes who used insulin was associated with lower A1C levels and lower rates of emergency department visits and hospitalizations for hypoglycemia.
“Perhaps most telling is that people with type 2 had greater improvement than those with type 1,” said Revital Nimri, an endocrinologist at Schneider Children’s Medical Center of Israel, during a session focused on primary care CGM use at the annual meeting of the American Diabetes Association. “Nevertheless, only 1% of the type 2 population got access to CGM, compared to 61% of those with type 1.”
But their effect in the real world will depend on whether — and how — primary care physicians are able to deploy them.
The first barrier is simply awareness. “I have some primary care colleagues who are very comfortable prescribing the FreeStyle Libre and telling patients the basic instructions to get it started, and I have some folks who have pretty much not even heard of a CGM or wouldn’t really know the first step,” said Tejaswi Kompala, an endocrinologist at UCSF and Livongo’s director of clinical products.
But the public is growing more familiar with CGMs, thanks in part to free device trials and ads like Dexcom’s multimillion-dollar Super Bowl spot featuring Nick Jonas.
“Even though this was geared toward patients, it was seen by doctors too,” Hirsch said, adding that he started getting emails from colleagues in primary care about prescribing CGMs; Kompala, too, began receiving more inquiries from patients before she could bring up CGMs herself in visits.
Companies are also ramping up direct outreach. Dexcom has targeted its education initiatives at primary care doctors who prescribe a lot of insulin to their patients with diabetes; more than half of patients on intensive insulin, type 1 or 2, are treated in a primary care setting. “[If] you look at the penetration of CGM in the endocrinologist community, it’s so much higher than the primary care community,” said Dexcom CEO Kevin Sayer. “We are a business, so we’ve got to look for new customers, and that’s where a lot of the patients reside.”
Primary care physicians will also need to be confident in how to best prescribe and manage their patients’ care on CGMs. “It isn’t that you just stick a device on someone and they get better,” said Anne Peters, an endocrinologist at University of Southern California’s Keck School of Medicine and co-author on the MOBILE study. “You put a device on someone, you work with them, and you improve outcomes.”
“The bad news is the primary care providers are not trained for this yet,” said Hirsch. “They don’t have a good infrastructure for looking at data. They don’t have a good infrastructure for dealing with all the hassles of getting the patients their devices.”
The problem isn’t so much that CGM data is inscrutable, but that primary care doctors aren’t often trained to read the standard format for the information, which is called an ambulatory glucose profile and shows several measures of a patient’s blood glucose activity. “If primary care doctors are educated on how to read a CGM, it’s amazing how much they can focus in on the major problem within 30 seconds,” said UCSC’s Edelman.
During the pandemic, Hirsch led a continuing medical education module on best practices for telecare in diabetes, including CGM use. After the session, 86% of the attendees said they intended to change elements of their practice — mostly by incorporating CGM data or advocating for coverage with insurance providers. But that kind of change requires an investment of time that PCPs rarely have.
“If you ask a primary care provider: Do you think that people with diabetes that you’re seeing in your clinic might benefit from CGM? They’d probably say, ‘Probably, but I don’t have the time,’” said Mahmood Kazemi, chief medical officer for Abbott’s diabetes care division.
They also often don’t have the technical setup to easily ingest CGM data. Patients can share their data with physicians through the manufacturers’ portals or aggregators like Tidepool and Glooko, but that still requires toggling between screens during a visit that’s already short on time. To record that data in EHRs, providers still have to manually type in glucose metrics, and some resort to screenshots or even printouts to scan back in.
“The truth is endocrinology clinics, they’re set up to get access to the data; they’re sort of primed for this,” said Martens. “Primary care, not so much.”
Some providers and device manufacturers see a solution in better interoperability. Martens’ International Diabetes Center just launched a pilot with Abbott that directly integrates its data into the practice’s EHR. “This is really key for adoption in a primary care setting,” said Kazemi.
There’s only so much that simple data porting can do to help, though. “The thing that’s missing in a lot of these platforms is suggested therapeutic interventions,” said Edelman, which could help primary care providers navigate the ever-growing list of diabetes medications and nuanced insulin dose adjustments. Experts said it could be valuable to feed CGM data directly into EHRs to inform clinical decision support for primary care physicians. “The real dream is artificial intelligence to help guide insulin-based therapy,” Martens said.
Before all that, though, primary care doctors must jump a high hurdle: getting CGMs covered.
Currently, CGMs are easily reimbursed for patients with type 1 diabetes, but Medicare coverage for type 2 diabetes requires meeting a long list of criteria. The Centers for Medicare and Medicaid Services recently stopped requiring patients to measure their blood glucose four times a day in order to qualify, but three-times-daily insulin injections are still a requirement. And primary care physicians have far less experience navigating the rules so their type 2 patients don’t get hit with hundreds of dollars in additional expenses each month.
Many patients, meanwhile, are relying on primary care providers to help jump through all those hoops and prescribe CGMs when appropriate, which they aren’t always willing to do.
“I know it’s probably a small chapter in the medical book about diabetes,” said Smith, who is still waiting for her monitor. “But if you were a doctor and you have 300 patients and 275 of them are diabetic, you don’t want to go a step further for your own patients? That’s the part that I just can’t understand.”
Six months ago, she finally found a doctor she’s willing to stick with, who took her off most of her medications, including insulin, to start from scratch. With most commercial insurance following Medicare’s lead, she won’t qualify for CGM coverage any time soon.