In early May, a wing of Ohio State University’s Wexner Medical Center was eerily empty. The space had been cleared of patients as the pandemic raged. But it wasn’t going to waste.
Inside, a group of nurse practitioners were playing a game of digital tag. Lauren Chrzanowski, who has type 1 diabetes, was wearing a continuous glucose monitor, an implant with a transmitter that sends real-time glucose measurements to a dedicated receiver — when they’re close enough, anyway. On that day, she was trying to figure out just how far away her CGM could be from its receiver and still work.
Hospitals around the country were racing to answer the same question. Normally, they monitor patients’ blood glucose with a blood test, taken several times a day. But as Covid-19 began to spread aggressively in the U.S., “we all started to look at each other and think, ‘How are we going to in good conscience put patients on IV insulin when the nurse will have to go in every hour?’” said Eileen Faulds, an endocrinology nurse practitioner who works with Chrzanowski at OSU.
The answer, in part, was remote monitoring. In April, the Food and Drug Administration made a call to temporarily allow hospitals to monitor blood sugar using CGMs — which are only approved for home use — until the public health emergency passed. By May, several hospitals around the U.S., including OSU, had received hundreds of free and low-cost CGM units from manufacturers including Abbott and Dexcom.
To use them with patients, they had to figure out how to adapt a device designed for at-home use — which is why the OSU team was playing tag.
While Chrzanowski stayed in a windowed patient’s room, Faulds zig-zagged around the floor, carrying the CGM’s receiver. From the nurse’s charting station outside the door, they learned, they could both check the receiver — about 10 feet away from the patient’s bed — and inject the needed insulin through 20 feet of tubing. Other hospitals simply taped the PalmPilot-like receivers up to the patients’ door frames.
Early data suggests those jury-rigged systems have improved care and outcomes for hospitalized patients. And some care providers think CGMs have the potential to improve inpatient outcomes if adopted more broadly.
While the FDA hasn’t approved CGMs for that purpose, data being collected on their use during the pandemic could help establish in-hospital remote glucose monitoring as the standard of care, ushering in a whole new market for manufacturers of those devices.
For individuals with type 1 diabetes (and some with type 2), the at-home advantages of CGMs are clear. An embedded sensor means no more finger pricks; instead, a small needle sits just under the skin, constantly measuring a proxy for blood glucose in the interstitial fluid around the body’s cells.
Thanks to that stream of data, many people who use CGMs have displayed improved control over their blood sugar, which comes down to two numbers that appear on their receivers: average blood sugar, and how frequently that value falls within certain targets, called time in range. Both can be improved by changes in diet or insulin delivery.
And for individuals with diabetes, it seems, “the same things that allow you to improve time in range in the outpatient setting translates to the inpatient setting,” said Christina Astley, a pediatric endocrinologist at Boston Children’s Hospital.
Evidence of that was emerging before the pandemic. Standard practice is for patients to remove their CGMs when they enter the hospital, but some doctors such as Irl Hirsch had their patients keep them on.
“I have been known, for patients wearing Dexcom, to go ahead and share their data with me or the fellow on the service so we could make sure that they weren’t getting into any trouble,” said Hirsch, an endocrinologist at the University of Washington Diabetes Institute. “Though the fellows weren’t crazy about their phone beeping at three in the morning.”
Learning how to respond to those alarms, which go off when a patient’s glucose gets dangerously high or low, is one barrier to systematically implementing CGMs in hospitals. It takes time and money to get hospital staff trained and comfortable with using a new technology, though Faulds emphasized that providers have a vested interest in using technologies that reduce their workload and improve safety.
Then there are the technical issues — like those the OSU team encountered in May. Some CGMs have the ability to transmit results to a phone app, not just a receiver, and “the intention was that we would also use the phones,” said Faulds. But they came loaded with SIM cards and data plans that could jeopardize the privacy of patient data. The team got stripped-down phones, but the majority of them still wouldn’t penetrate the hospital’s firewall. In the end, the only option was to use traditional receivers.
Despite these hurdles, many care providers and device manufacturers were advocating for inpatient CGM use long before the pandemic. What had kept them from moving forward is an even higher regulatory hurdle.
“There was a little of a chicken and egg problem,” said Matthew Taylor, a medtech device analyst at UBS. “Dexcom’s talked about this in their past conference calls,” he said. In conversations with the company before the pandemic, “FDA basically said, ‘We think this is a good idea in terms of value for patients, but you need data looking at use of CGM versus not.’”
But because CGM has reliably improved outcomes for patients with diabetes, denying the tech to a subset of patients in a randomized trial could be unethical. Audra Harrison, an FDA spokesperson, said the agency couldn’t comment on ethical issues involving future data submissions, but said that sponsors “can choose the appropriate control group for their studies and endpoints.”
So far, though, few studies of CGM in hospitals have studied patient outcomes. “We’ve got 50-plus studies, even pre-Covid, for using CGM in the ICU, but they were really focused on accuracy,” said Faulds. CGM’s interstitial fluid measurements often require calibration against actual blood glucose, and patients encounter stressors and drugs in the hospital that could interfere with their accuracy. Still, said Faulds, “I don’t think accuracy is going to be the biggest issue; I think it’s just proving how that ties to patient outcomes. We need outcomes data.”
In a silver lining, the pandemic is providing some of it.
“The pandemic has given [manufacturers] kind of a backdoor to be able to generate data,” said Taylor. Along with partners at Mount Sinai, Emory, Stonybrook, and NewYork-Presbyterian hospitals, OSU is pooling data about inpatient outcomes during the crisis.
“It’s not a randomized controlled trial, you know, we don’t have controls here, which is going to be the issue,” said Faulds. “But if we can get all of these institutions to publish our data, I’m hoping that they continue to allow us to use the devices.”
While robust results will take more time, small studies based on in-hospital CGM use during the pandemic have already pointed to better outcomes. At Scripps Whittier Diabetes Institute in San Diego, 110 non-ICU patients with type 2 diabetes on average saw reduced blood glucose levels and improved time in range. In a similar patient group in Baltimore, a randomized controlled trial showed that CGM use reduced instances of hypoglycemia, or lower-than-normal blood sugar levels.
Meanwhile, Dexcom is running its own registry to track outcomes. “This is something that Dexcom is taking very seriously to develop pathways to improvement,” said Tomas Walker, the company’s U.S. medical director.
“We have a very close relationship with the FDA, and we’re always working with our regulatory bodies to maximize what we can do for patients,” he added.
“I think the most likely outcome is that [FDA] will ask for more data of some kind to make the approval permanent and official,” said Taylor. He added that it still isn’t clear what format that data might take, such as a registry or new studies. “We encourage the collection of data on the safety and effectiveness of these devices in the hospitalized population,” said Harrison. “Provided their use is safe, such data may be helpful in supporting clearances or approvals of these devices for this use in the future.”
To Faulds, a permanent FDA greenlight to use CGMs in the hospital would be a big step forward.
“I’ve done inpatient diabetes management for eight years,” said Faulds, “and I can’t even tell you how this would change nursing practice. It would change the way we manage patients in the hospital, it would be huge. I have no doubt that this is the future of glucose management.”