With Rise of Diabetes Technology Comes Value-Based Payment
Wth seemingly everyone having a smartphone and carrying it everywhere, it was only a matter of time before apps took center stage in efforts to prevent and manage diabetes.
But figuring out how to connect health plans with this technology is proving complicated. Medicare’s efforts to launch the Diabetes Prevention Program (DPP) by early 2018 have slowed as it works to include digital providers, which are needed to scale DPP to all who will need the program.1
Even paying for more conventional devices isn’t always straightforward. As diabetes technology evolves, patients want more choices, whereas payers want to hold down costs.
To delve into these issues, The American Journal of Managed Care® (AJMC®) hosted the discussion “Technology in Diabetes Care: From Prevention to Disease Management” during its April Diabetes Stakeholders Summit, a Peer Exchange™.
Moderator Dennis P. Scanlon, PhD, professor of health policy and administration and director for the Center for Health Care Policy and Research in the College of Health and Human Development at Pennsylvania State University, in University Park, Pennsylvania, led the discussion. Joining him were Robert A. Gabbay, MD, PhD, FACP, senior vice president and chief medical officer, Joslin Diabetes Center, Boston, Massachusetts; Mary Ann Hodorowicz, RDN, MBA, CDE, CEC, a Chicago, Illinois–based consultant, dietitian, and trainer; Kenneth Snow, MD, MBA, medical director, Aetna; and Neal Kaufman, MD, MPH, founder and CMO of Canary Health.
The April discussion foreshadowed an important development when Snow said that “without a doubt” the diabetes device world would be moving toward value-based payment models. On June 26, 2017, as this special issue of Evidence-Based Diabetes Management™ went to press, Aetna and Medtronic announced a risk-sharing agreement for patients transitioning from multiple daily insulin injections to Medtronic’s insulin pumps, including the new Medtronic MiniMed 670G. The agreement covers patients with types 1 and type 2 diabetes (T1D and T2D). Medtronic’s reimbursement will be partly based on outcomes-based measures for patient experience, clinical outcomes, and total cost of care.2
Technology and the Diabetes Prevention Program
As the discussion began, Gabbay explained that results from a landmark National Institutes of Health study show that a lifestyle intervention under “very controlled, rigorous conditions” could produce better results than medication (metformin) for preventing diabetes.3 “The challenge of that study [is that] it was done in a very resource-intensive way to ensure that people adhered to their lifestyle,” Gabbay said. “Now, how do you apply that to the broader population?”
That was the early concern for payers, Snow said, even though they were excited about the DPP. Kaufman, who was very involved in the early years of the DPP, weighed in, pointing out that a single trainer could help 30 to 40 people a year because the program required 16 weekly sessions at the start, followed by monthly in-person sessions. Training was done one-on-one.
“The CDC recognized that that was not a scalable or sustainable model and began looking at how to provide it in other ways,” Kaufman said, and then described 2 “threads” of approaches. The first, most common approach involves training to 10 to 15 people at once, with highly trained educators leading the groups. The best example is the YMCA pilot study funded by the Center for Medicare & Medicaid Innovation (CMMI), which provided the evidence to scale the program across Medicare.4
The second approach is to use technology. The question here, Kaufman said, is whether a program offered in person would translate in a digital format, using a cell phone. “I was very much involved back in 2006 to create a digital version of that program,” he said. “At that time, we didn’t know if it was going to work. We didn’t know if people would accept it, but we recognized that people need choice …
“And so, we, and now a number of other companies, have been able to demonstrate that you can take an in-person program, use great design and great approaches to make technology work, and have individuals use it effectively.”
Scanlon asked the panelists to elaborate more on the YMCA study, because it formed the basis for Medicare’s decision to fund DPP, a decision that will bring diabetes prevention to seniors on a widespread basis in 2018. Many DPP providers believe that once commercial payers that administer Medicare Advantage see the value of the program, DPP will become even more embedded in health plans.5 Hodorowicz said that the dynamic of self-monitoring—tracking weight and exercise and keeping a food diary—and having to report to a lifestyle coach increases motivation more than taking these steps alone does.
Scanlon asked if the number of sessions patients must attend helps, as well. Kaufman replied that based on the YMCA study results, it does. The study contributed to the regulations CDC uses for program recognition: 16 weekly sessions, followed by less frequent attendance for a total of 26 sessions.
What about evidence? “This was a big decision for CMS to decide they were going to go ahead and pay for this,” Scanlon said, adding that it has not been without controversy. “Will other payers follow?”
Snow noted the enormous impact: “Any decision by CMS, whatever cost there is has to be multiplied by millions.” But one difference with Medicare, he said, is that beneficiaries don’t leave, so Medicare reaps the rewards of its investment. For other payers, a preventive service may or may not pay dividends if enrollees switch insurers. Still, “some payers have already made decisions regarding continuing coverage similar to CMS,” and Aetna is among them, he said. “Some will be looking at the type of data that comes out in the real world and real-life experience to see if the results that are hoped for are actually achieved.”
Thanks to the YMCA pilot, Gabbay said, it’s no longer theoretical that the DPP saves money. “That, in essence, was exactly what CMMI was set up to do—to be able to do pilot studies that show cost savings or high value, and then spread that more broadly,” he said.
When asked by Scanlon to address the role of diabetes educators in offering the DPP, Hodorowicz said that for years, patients with prediabetes were referred to educators but did not quality for coverage, even though a certified diabetes educator was perhaps the best person to help that patient avoid progressing to diabetes. But now, with some commercial payers offering coverage and » CMS poised to do so, the American Association of Diabetes Educators is training DPP coaches. “It’s an easy fit to include prevention with an existing self-management program run by diabetes educators,” Hodorowicz said. “It’s a perfect marriage.”
Kaufman agreed. If a person is at risk for diabetes, he said, “we need to help them improve their life’s trajectory, to help them so that they don’t add a new chronic condition every 3 to 5 years, as many people do.” That starts with lifestyle intervention. Focusing on the glucocentric requirements of eligibility for the DPP that will help people lose weight and become active is too restrictive, he said.
Once payers have the data, Snow said, they will look beyond the issue of diabetes: Does helping at-risk patients lose weight also help avoid issues such as joint damage or back problems? “[These] are clearly not glucocentric but still related to the same population,” he said.
“The real secret sauce is, how do you get people to sign up and show up?” Kaufman asked. “Once you can get them to the program, we can almost predict for 1000 people or 10,000 people what’s going to happen. But how do you get the right person to the right program at the right time to engage them, to activate them, to get them to see that there is a benefit?”
Gabbay agreed that for some patients, in-person programs such as the YMCA will be best, but for others, digital solutions will work well. Hodorowicz said she believes that for the younger generation, digital solutions will be more popular, and Snow noted that Medicare enrollees are more accepting of technology than people realize.
It’s important to distinguish between a digital tool, such as a text, which acts as a reminder, Kaufman said, and an intervention, which is a full program of health improvement. Gabbay said he regularly uses the smartphone, such as to show patients that they are walking less than they think. By contrast, he said, “if you want to do something to prevent diabetes, you need a whole intervention, based not only on counting steps but also [on] dietary changes and other behavior change approaches.”
The Clinical Rationale for Continuous Glucose Monitoring
Scanlon moved the discussion to the importance of continuous glucose monitoring (CGM), now considered the standard of care for those with T1D and becoming more common for those with advanced T2D. A 2011 study6 in AJMC® was the first to quantify the cost of a hypoglycemic admission to the hospital and was cited by JDRF and others in their successful effort to convince the FDA and Medicare to change policies that will ultimately allow beneficiaries to have coverage for the Dexcom G5, although there have been some implementation issues.7,8
There’s evidence that—particularly for patients with T1D willing to use CGM technology—it shows a benefit, Snow said, both in terms of improving glycemic control and lowering hypoglycemia risk. “Once you get outside of that, causation data becomes very thin,” he said, “so you see observational data, where there’s clearly a relationship between hypoglycemia and…if you can give them the message that prevents the hypoglycemia event, well, then that is optimal.”
The challenge is to make sure that patients have access to both the CGM and the right education, which Hodorowicz said is key, because when patients know how to use the data the device provides, educators can more easily instruct them on how to “embed lifestyle changes.” “The good news is that Medicare is starting to cover [CGM],” despite strict criteria, she said.
“It’s fantastic news,” Gabbay agreed. “At Joslin, we have a large type 1 population, and for them, they reach Medicare age and they have to go off their continuous glucose monitor, which is a big problem,” he said. “But I think you’ll see, in the not-too-distant future, this spreading to more use in the type 2 [population] based on the kind of evidence that people who are on multidose insulin can clearly benefit.”
From a payer perspective, Snow said, the fact that the FDA approved the Dexcom G5 for dosing was reassuring—this was a key step in Medicare’s reversal of its longstanding refusal to pay for CGM. “Once you have that FDA stamp of approval, there’s significant advantage, not the least of which is it usually means that there’s legitimate scientific evidence,” he said. “And that scientific evidence is what supports the use.”
CGM data will be more reliable than patient logs, which could be helpful with older patients. “That helps diabetes educators and physicians with medication adjustments and lifestyle changes,” Hodorowicz said. “With the pediatric population, it’s wonderful because parents are so involved with their children’s control, especially of type 1.”
Audible alerts can help patients or caregivers avert an approaching hypoglycemic event. “Payers are starting to recognize that,” Hodorowicz said, “and it’s not just for the type 1s, it’s also for the type 2s on multiple daily insulin doses.”
Devices help, but it’s essential that patients already be engaged in their own care, Kaufman said. He pointed to a program developed by Stanford University, the Chronic Disease Management Program, which is decades old and has been highly successful with patients with T2D.9 “If your A1C was above 9%, it went down by 0.93% at 6 months, and 1.27% at 12 months with a 6-week intervention,” he said. “How does it work a year later? Because people change their lives, change their emotions; they change their sense of well-being, and, therefore, they were able to follow doctor’s orders better, follow nurse’s orders better.”
“And that becomes crucial in anything that has to do with self-management,” Kaufman added. Scanlon noted that this was work pioneered by Katie Lorig, DrPH, and remains foundational in chronic disease management.
Scanlon and Kaufman discussed the need for peer-to-peer support, either in person or in digital formats. Gabbay noted that vast potential for the latter is “just beginning to be tapped.” “Continuous glucose monitoring certainly helps, and even the simple ability of downloading blood glucose meter data and looking at that with the patient is really helpful,” he said. Providers don’t want too much data, however, and decision support tools help manage it all, which he called the next revolution.
“The next revolution after that,” Kaufman said, “is being able to assess it in the moment and give feedback to the individual in the moment.” The challenge, he noted, is that managing diabetes requires patients to constantly make micro decisions. No single decision is hugely important; it’s the sum of them that determines the outcome. The perfect device would require no interaction.
It’s hard to balance the desire for tools without overloading patients to the point of “alarm fatigue,” when they tune out the efforts to help them, Snow said, “because it just gets in the way of living.”
Payer Coverage, Patient Choice in Insulin Pumps
In this rapidly evolving area, Gabbay said, the FDA appears to be open to approving devices more quickly that it had been. “The floodgates are about to open because there are a whole series of iterative changes that can be made to push forward semiautomated insulin delivery,” he said.
As Scanlon noted, however, the issue is figuring out for whom is the device appropriate. “And how do we determine appropriateness?” he asked.
“That’s one of the great challenges, because there’s not a lot of good data that identifies who makes a good pump user,” Snow said. “We can extrapolate [whom] we believe that might be, but there’s not really a lot of data that [are] going to predict for the person sitting in front of you, whether they are likely to succeed or fail.”
Gabbay said that this is changing with the movement toward semiautomatic delivery; the bar is no longer as high for patients’ carb-counting and insulin-adjustment skills. “You could take patients with poor glycemic control and maybe not terribly adherent and [put them] on semiautomated insulin delivery, and that would at least improve their blood sugars overnight, and they would benefit,” Gabbay said.
CMS approved external insulin pumps for T1D and T2D Medicare patients who had been on multiple daily injections (and who have less control as they learn to use the pump in the first year), Hodorowicz noted. “The cost benefit has to be there for CMS to approve coverage for such an expensive item,” she said.
When asked about the differences between models and patient preferences, Hodorowicz said that a waterproof device is important to many patients. Some devices are disposable; some offer wireless infusion versus manual control. Different types of alarms matter to patients, as does ease of calibration. “I think the ease factor is critical,” she said.
CGM devices are on the verge of becoming much smaller and cheaper, Gabbay said, although pumps may remain relatively expensive and complex. Snow noted a possible disconnect between what makes sense from an individual’s point of view versus the population perspective. Giving an expensive device to a patient with a high glycated hemoglobin who seems unmotivated and unengaged may not make sense, but if a person’s A1C of 10% drops by 1.5%, then the population has become healthier overall.
A person with T2D with poor control could benefit from a continuous glucose monitor, Kaufman said, by identifying personal patterns and seeing the connection between certain foods and spikes in blood glucose. “The question becomes, can it become a behavior-change incentive to allow you to become even better at managing your own condition?” he said.
The panel discussed an agreement that will require most adult UnitedHealth patients with diabetes to transition to Medtronic pumps and CGM devices as warranties expire.10 “Sadly,” Gabbay said, “it’s somewhat inevitable because there will be competition around price, the same kind of things that happen in the pharmacy world—I don’t see why it won’t happen in the device world.”
Snow said that historically, if 2 drugs have equal efficacy but 1 is cheaper, the savings are passed down the line and ultimately result in lower premiums. “Clearly, there can be a benefit without a detriment of care,” he said.
Scanlon asked if payment models would change in this area, too. “Without a doubt,” Snow said. “It is the direction of the payer industry.” Many pharmaceutical companies have entered these agreements, “and the device companies are looking that way, as well,” he said.
“The idea of just paying for your treatment is one that was there in the past, but more and more the question is: is there a value from your drug, from your device? Regardless of what it is, is it adding value in terms of either less expensive care or better care? And it doesn’t have to be both.”
Gabbay added that A1C is no longer the only measure of success. “If you reduce hypoglycemic episodes, well, that’s probably a good thing, and that should be a metric, as well.”
The field of diabetes technology has entered an exciting era, Kaufman said. “A paradigm shift is happening,” he said. “We’ve got consumerism, where individuals are taking more control and responsibility for many things, including their health and health care. We’ve got technology being able to deliver something in the moment. You have it in your pocket at all times.
“We’ve got the concept of population health, the concept of value-based care coming to a health system that wasn’t set up that way. If the patient is at the center, we always think of them as what’s most important and their needs, that we, as healthcare providers and payers and employers and health systems, will make it so that we’ll help patients have those outcomes that they need.”