UnitedHealth Group launches Level2 type 2 diabetes remission program following successful pilot – July 15, 2020

Offers Dexcom G6 CGMs, Fitbit trackers, AI-driven feedback, and virtual coaching through an app to 230,000 type 2 members in 27 states and Washington DC at no additional cost

Expanded August 25, 2020 with fascinating interview with CEO Amy Meister and Chief Growth Officer Will Ferguson.

Two weeks after announcing coverage of Tandem’s t:slim x2 pump with Control-IQ, reinforcing its commitment to value-based healthcare and data-driven approaches, UnitedHealth Group again made a splash today announcing its launch of a new type 2 diabetes remission program called Level2. The intervention provides eligible members a Dexcom G6 continuous glucose monitor, a Fitbit activity tracker, smartphone app-based alerts, personalized clinical coaching, and virtual specialist consultations. The program website notes that the program is geared to promote behavioral change and health lifestyle decisions by giving participants real-time access to biometric data (e.g., activity, stress, responses to food, glucose levels, sleep, heart rate) while also limiting financial risk to employers. In a testament to the focus on data-driven behavioral change and personalized therapy, UHC characterizes CGM as an “essential” component of the program, requiring all participants to wear them, and further encourages participants to wear the Fitbit, while not mandated, to gain a “more complete picture” of their lifestyle choices as a “supplement” to CGM data. Dexcom G6 is featured prominently in the screenshot below, described as “small and sticky like a Band-Aid…only a lot smarter.” Impressively, an estimated 230,000 employer-sponsored, fully-insured UnitedHealthcare members with type 2 diabetes in 27 states and Washington, DC will gain access to the program at no cost. Wow! To our knowledge, this would be, by far, the largest program to offer CGM to type 2s. Though not direct comparators, for context, Livongo for Diabetes (no CGM offered) has “over 328,000” members while Omada has over 275,000 members across all its programs (mostly diabetes prevention, which again, does not offer CGM). This announcement from UnitedHealth is incredibly exciting as there is so much opportunity to explore new business models for CGM wear in type 2 diabetes; offer value beyond standalone CGM by combining it with Fitbit data, behavior change programming, education, and more; and improve outcomes for people with type 2 diabetes.

The program will provide gift card and cash incentives to encourage positive lifestyle choices such as walking goals, wearing the CGM, meeting with a coach, reporting behaviors in the Level2 app, or simply enrolling, although the stipulations will vary between states and plans. Interestingly, the app leverages machine and AI technology to deliver feedback while participants can also secure coaching with a diverse array of experts including dietitians, social workers, nurses, and physicians through the app, text, and video. UHG also noted that Level2 will be made available to “select employers with self-funded plans” later in 2020 and that deployment among members with other chronic conditions in addition to integrations with other CGM systems and wearable trackers are in the works. In a real win for patient choice and flexibility, members can use their own Fitbit if they already have one, end participation at any time, and can take part even if their blood sugar is deemed “under control” or if they participate in other diabetes management programs. Level2 even takes care of orders/refills of CGM sensors and transmitters along with lost/damaged item replacements at no cost, thereby limiting common logistical hassles participants may face.

The broad rollout of the program was based on a successful pilot study among UnitedHealthcare members (n=790) and supporting type 2 individuals. UHG’s press announcement notes that in the pilot, “certain” participants achieved a clinically meaningful reduction in A1c while those with baseline A1c values above 8% saw on average >1% reductions. Though not quantified, “some Level2 participants” achieved type 2 diabetes “remissions” (A1c <7% and no longer required medication). In the pilot, Level2 eliminated the need for more than 450 prescriptions (~0.6 medications/participant). In another interesting note, the announcement noted that pilot data has shown sudden changes in CGM readings in type 2s may suggest potential COVID-19 infections.

• As mentioned above, we believe this is the largest type 2 diabetes program offering CGM to-date. Abbott and Omada announced a partnership last year to provide FreeStyle Libre as part of Omada’s type 2 diabetes program, but we haven’t heard many updates since then. At ADA last month, we heard data demonstrating Onduo members using Dexcom CGM saw greatly improved outcomes compared to the Onduo users using connected BGM only. Newly appointed Virta Chief Medical Officer Dr. Robert Ratner shared in an interview several months ago that the company in its type 2 remission efforts is also introducing CGMs given its educational and positive behavioral reinforcement value. Interestingly, while the ADA 2020 Standards of Care includes Time in Range and CGM metric goals, it also specifies that there is no “one-size-fits-all” approach with the technology given vast differences in reimbursement, literacy, and patient interest. We’re hoping that movement from both payers and companies through efforts such as Level2 can bring more light to evidence showcasing value of decision-support and coaching tools to patient populations that technology has yet to reach.

• Level2 will be made available to select participants in: Arizona, Arkansas, Colorado, Connecticut, Florida, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Michigan, Mississippi, Missouri, Nebraska, North Carolina, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Carolina, Tennessee, Texas, Virginia, Washington, D.C., West Virginia, and Wisconsin.

Close Concerns’ Questions

• Is Level2 required to have prescribers in every state the program is offered in?

  • Level2: Correct. Level2 has licensed providers in all 50 states and the District of Columbia.

• Are coaches and prescribers direct employees of Level2 or contracted?

  • Level2: Level2 coaches and prescribers are direct employees of Level2.

• Can you share the number of coaches to number of participants ratio? How about number of physicians to number of participants?

  • Level2: We use both physicians and advanced practitioners. Supervision is guided by each state’s rules regarding physician oversight of advanced practitioners.  Our patient-to-clinical care team ratio varies and is dictated by our proprietary algorithms.

• How quickly can the program scale to the 230,000 potential participants outlined in the initial press announcement?

  • Level2: We have set the business up with the intent and ability to rapidly scale.

• Can you share any engagement statistics? How many interactions are happening for each participant each day? What percentage of those are automated vs. from a coach?

  • Level2: All outreach is done by human touch. The AI flow is from our AI analytics engine to our care teams and not directly to participants.

• Are there plans to publish the data from the pilot study (n=790)?

  • Level2: Yes. We are working with academic partners on several publication opportunities.

• What factors informed UHG’s decision to build out its own program (i.e., Level2) rather than contracting out to other virtual clinic programs (e.g., Onduo, Omada)?

  • Level2: With our data and technology, coaching and clinical capabilities, nobody in the marketplace is better positioned than UnitedHealth Group to offer consumers an integrated health plan and care solution like Level2. With Level2, we focused on the research that supports the idea that new approach to T2D is needed to attack prevalence and cost. Despite a number of good point solutions in the marketplace, none have been able to decrease prevalence – or cost of care – for people living with type 2 diabetes.  With Level2, we realize the value of truly integrating our coaching and clinical care offering with a comprehensive health plan.

Interview with CEO Amy Meister and Chief Growth Officer Will Ferguson

Albert: Thank you so much for spending time with us. If we could, we’d love to start with asking you, what is the mission of Level2?

Dr. Amy Meister, CEO: We wanted to bring hope and we know that if we can attack the root cause of diabetes and actually get people to a state of remission, then we have an amazing opportunity to change the way diabetes is cared for. Additionally, one of my biggest pain points has been as a care provider, trying to be able to give to my patients things like continuous glucose monitors and the right medications, because there's always hoops you have to jump through with prior authorization.

I like to call healthcare the sausage making factory, because right now the traditional approach is you have to see one doctor who's going to refer you to an endocrinologist who, if you have benefits, is going to refer you to a nutritionist or perhaps an exercise regimen. What we've done in Level2 is remove all those barriers so my providers are actually able to just provide care.

We have a comprehensive virtual ecosystem where, yes, we're focused on diabetes. But, it would be a huge misstep for me to say I'm treating a disease instead of a person. So, we don't look at the individual as just having to deal with diabetes as a clinical diagnosis. But we look at them more holistically and how that impacts other diseases and on the psychosocial front as well.

We have a really unique opportunity at UnitedHealth Group. We have something called the individual health record, and it has over 57 million health records in it. On top of that, we have medical claims, pharmacy claims. Things like lab values that are put into our data link and now we’re combining it with the wearable signals, so things from activity trackers, things from the continuous glucose monitors, all that stuff coming into our engine. We run it through some proprietary algorithms to help our care delivery team figure out two things. One, who's the right first care provider for this person? Should it be a dietician? Should it be a registered nurse? Should it be a physician? Who needs to see this person? And then on top of that, what do we tell them? So is it intermittent fasting? Is it Mediterranean diet or do we just need to get you off some of the toxic medicines that are causing some of these problems or a combination of all of those things?

Our data is so rich that we're able to do that. And now what we've done in the self-funded marketplace, we've actually anchored this to a benefit design. And the power of that allows us to have people that select our health plan to a zero out-of-pocket costs for diabetic medications, for physician visits on our extended healthcare ecosystem under the Level2 umbrella.

And what's great about this is really bringing concierge medicine. And we are able to take full risk on that population because we believe that our tools and resources in this new approach is that powerful. We are getting amazing results to date. And I would say a couple of things that I want to highlight. One, that we've learned is by eliminating the traditional model of care, requiring bricks and mortar and having the chain of -- we’ll just say the gating chain of command to get the care that you need. We are finding that we are accelerating results that would take three years in the traditional system to a month to about three months on average. And to put some numbers behind that, we, on average, when you look at people that have the poorest controls, so A1c over 8%, we're seeing in one month that we're able to drop their hemoglobin A1c by 1.5%. For the FDA to approve a new drug, six months, half a percentage point [A1c drop]. So, we're considering -- we really want Level2 to be seen as a digital therapy, continuous care ecosystem rather than a point in time solution.

Kelly Close: I echo Albert’s thanks for spending time with us. That’s amazing to hear and music to patients’ ears, a digital therapy. What kind of Time in Range differences do you see? What do you see in terms of time above range and time below range?

Dr. Meister: Yeah, we do. Absolutely. Let me cite one particular population we looked at, just a variety of folks, 750 people, over one month. The first month we had over 77% of people move from out of range to in range, which is great. I attribute a lot of that to honestly, the CGM, because for the first time you have people don't have to go to their bag, dust off their glucometer and poke their finger. They're getting that real time data. They're getting 288 readings a day. It just brings a whole new line of sight and self-motivation when you truly understand how things even outside of diet, stress, sleep activity, the time of day that you eat, the order in which you eat your food, all those things that we don't learn as providers in the healthcare system impact your glycemic control. The time you take your medicine, all that stuff. Now people are able to see that real time, which is just amazing.

Kelly: Wow! The power of over three quarters of 750 people all MOVE from time OUT OF RANGE to in range – wow. That’s so great for them. Can you say anything else about the CGM? This is really interesting to us, and was the baseline Time in Range pretty low? [Editor’s note – for the answer to the last part of the question, see several questions down.]

Dr. Meister: Yes, and we've done a couple things algorithmically. We're doing a couple of things. We really focused heavily on Time in Range, and what we look at are people that have higher variability. So, they may have a good A1c, but they're spending too much time too high and too much time too low – that's why they look good. So those people, that’s what we call high-risk.

The other group of people that we put in that high-risk bucket too are probably a little more obvious, which would be people who have low ranges and then people who have high ranges. I actually put the low ranges since we're dealing with many people that are highest-risk, even for type 2 as one of the most dangerous groups, because, as you know, low blood sugar can kill you if it's not treated acutely. Where usually high blood sugar we have a little bit of time to deal with.

So, those are all people that we ingest that data that we outreach immediately to figure out not just what happened now, but what can we do to prevent that? What do we need to do? And, because I have doctors on my team that outreach, we're able to adjust their medicines right then and there. There's no waiting for somebody else to step up to be able to do that for the patient. So, we eliminate some of the burden of playing tag you’re it in the healthcare system, which is which is really nice.

We take a look at those populations and then we're also looking at the polypharmacy populations. This is done through our really our AI engine. We take a look at the medicine folks are on. We have a nurse outreach that does medication reconciliation because a good number of medicines, particularly the sulfonylureas and things that are less expensive, often people choose to pay out of pocket. So, we don't have that information and claims data. We’re entering it into the system, teaching our machine to even be smarter about that kind of stuff. [Editor’s sidenote – at the time of our interview, we didn’t catch this fully, but UHG is figuring out using TIR even when someone is taking a damaging medicine like an SFU off label and it’s not even in the health record. Whoa.]

We're optimizing the medicine. So, if they're on a antihypertensive, that could blunt the effects of low blood sugars. If they're on an antihypertensive, that's going to drive obesity. We change it. So, we really focus on that medication optimization period and then whatever our science generates, what type of diet would be best for them to help really get them in a better state of health.

Incidentally, we also just hot off the press today, took a look at people really since October of this past year to kind of give you an idea. And we've had over 13,000 people that we've gotten data on that have participated. And on average, most people are losing five percent of their body weight in three months. And that's just incidental finding.

Kelly: Wow. Is there anything about, like cholesterol or blood pressure, because that’s a really massive change on glucose and variability. What was the final finding?

Dr. Meister: Between 70 and 180 mg/dl, we had 77% were there.

Kelly: Wow. That’s incredible. What was it coming into the study?

Dr. Meister: About 33%.

Kelly: Amazing. This population went from 33% TIR to 77% TIR in three months – that is incredible to hear, and particularly about the intervention of CGM combined with advice on food and likely exercise, etc as well?

Dr. Meister: It's really amazing when you give people tools. It’s really also given me a lot of courage and hope in healthcare. Again, to realize that people do care, you just have to give them the right tools and resources to help them.

Kelly: This is a dream. What did duration of diabetes look like? Was that mostly under 10 years or under seven years or something like that?

Dr. Meister: No, there's no consistency there. We have everybody from newly diagnosed to those that have really had diabetes for a very long time.

Kelly: That’s remarkable. Can you talk about what else you observed?

Dr. Meister: We optimize their therapy … for some folks, let's say somebody has coronary disease, perhaps not even related to diabetes, coronary disease that was pre-existing before they were diagnosed with diabetes. We know now that the SGLT-2 class inhibitor really is cardio protective and there's a lot of studies that are looking at that medicine outside of diabetes in general. So, in that particular case, I would de-escalate their medications, but I would not remove that medicine because just the other research showing that it's going to be beneficial to that person. But I would still monitor them very closely and get them off as much as I could. And even many of their antihypertensive as you reference anti-lipidemic agents, antihypertensive agents, all those things we would look at getting people off.

Albert: Amy, you mentioned earlier that the program gave you hope for healthcare because it shows that people care about their health. Toward that end, we wanted to ask about with Level2, what your approach to data is. If I remember correctly, you all are drawing data from the CGM. A lot of times, I think you're encouraging a Fitbit to be involved. There's meal data, mood, food, sleep, is that right?

With all of that data, it can, for some people lead to being overwhelmed or burnout or even, if the numbers aren't great, it can be discouraging. What’s your approach there?

Dr. Meister: Now, that's a great question. So, this is the way we've approached it. As the Level2 team is really ingesting the data, we are seeing some people as what I call the worried well, for the most part, the folks that are, “Oh my goodness, my blood sugar going up from 114 to 116 mg/dl, what do I do?” So, a lot of that you have a plan really designed for those people to try to help them to understand.

I’m going to break it down in buckets. When we’ve got somebody to the point clinically to where they're pretty much optimized and in that worried well bucket. They may be on one medication, they may be on two meds. They may not be on anything, but they're still hyper vigilant. What we try to do is retrain the brain to become their own CGM to where they associate the way they feel with certain foods. In other words, if they have a 2:00 p.m. afternoon slump. Is it because at lunch they had a cupcake? And when they're able to start really understanding whether it’s activity, whether it's food, whether it's stress and how that makes them feel. So perceived glucose, if you will, and connecting that. That's a really powerful tool and helps us wean people off of being completely wigged out by data.

The other thing that we've done intentionally initially, and we've gotten some mixed reviews on this, but we've really pulled in most of the data internally. So, if someone's looking at their phone, they can obviously see what their glucose is. If they're looking at their activity tracker, they can see their stats or check it in the app. But a lot of the data and what we focus on to them is that, let us put this in the recipe. It's like when you're baking a cake, you're not going to eat one ingredient. And it would be really overwhelming if you had to eat each individual ingredient. Let's bake it up and then let's have a slice of that cake. Maybe that wasn't the best reference because of diabetes. But nonetheless, the point is, is that we bring all of that data in, mix it up and then we turn it -- part of our secret science is that we have algorithms that review all that, filter out the noise and really get to the point so we can have meaningful conversations.

If you've ever seen a CGM graph in the wild, it looks like a plate of spaghetti. When it comes through our algorithms, it has meaningful peaks and it has a point. Then, what I do and my team, we have the ability to do both synchronous and asynchronous video visits. And this is super cool. We actually go into our EMR or our coaching portal and we pull up what the curves look like over a 10-day period of time, along with the other information that's layered in there. So the activity, foods, relevant things that are going to explain probably -- if there's something concerning or even -- and what we really, really like to do is call out the good times rather than the bad times, because we want people again to focus on, wow, one day. Fantastic. Let's talk about what you did on Wednesday, because we want to have every day like Wednesday. So that's a little bit of a unique approach.

And then what we're able to do is I actually walk through the screen. It's a split screen. So, you see me on half the screen and the other part of the screen you see your own data. And then my recommendation, I send that within our secure chat app as a video visit that the participants can, at their own convenience, watch that, ingest it. We see most people watch them anywhere from three to five times and then they can reach back out. We can either communicate by chat, we can have a live telehealth visit.

We can just simply have a telephone call, whatever is needed after they have time to review that. So that's really how we're managing the burdensome part of the data and really organizing it and putting it in a fashion that makes it ingestible for our participants and our team providing care.

Albert: I'm not sure if you all have gathered enough data for this, but I know you used a cake analogy earlier and it kind of sounds like, all those pieces from the CGM, from the meal logging from the Fitbit are kind of creating something that's bigger than the sum of its parts. I'm curious if you've done any sort of analysis to look at what might be causing maybe some of the biggest effects.

Dr. Meister: Yes, we actually are looking at the individual components as well as CGM alone. Additionally, what we're looking at, we have a few studies right now that are ongoing. A couple of things just to give you kind of a sneak peek. We are looking at prepared meal services, how that can have an impact. So, food delivery, a controlled diet, how does that impact? We've looked at CGM alone and estimated its impact on control. There was a nice study that we talked to. I think it was done by University of Colorado. And they looked and they estimated about a $5,000 a year savings just from CGM use alone, implementing it both in type 1 and type 2. So, that was an interesting study that that we used. We’ve been trying to duplicate some of the analysis to get those numbers. I don't have those numbers readily available at the top of my head, but we have done that and we can pull that together to share. We're also looking at a couple other factors, too. Everything from age, salary grade to other health activation indexes to really try to tease out what part of each therapy is contributing to each participant successes or failures in.

Albert: We know you don't have all the numbers yet, but are you seeing, even qualitatively, that different things are working for different groups? Or is it that certain combinations of things seem to be more powerful than others?

Dr. Meister: I think it's probably too early to tell exactly. What I can say overall is that one of the big areas that we've seen in comparison to other approaches is that our engagement numbers are significantly higher and we're doing a better job of engaging people that are demographically challenged by some social determinant of health than we've seen with other with other programs.

So, for instance, if you look at our own UnitedHealth Group employees, we see that we've been very successful. About 56% of those we offered in the lowest salary tier, under 30,000 dollars a year, are opting in that want to participate, which we think is fantastic. Because when I was in the onsite clinic business prior to this, also looking to try to capture those same people, to get them engaged with the healthcare system. Really didn't have near the luck we were expecting across our own employees that we've launched this. Our moonshot goal was a 10 percent adoption rate and two months in we’re at 30 percent and keep going up. We’re super excited about that.

Kelly: I’d love to continue to ask more about where Albert was taking us. What were you seeing in terms of what therapies they're using, what's happening in terms of stress, in terms of data on exercise? What’s happening with pedometer data?

Dr. Meister: We do. And actually, to answer your question about hyperglycemia, we actually have a plan of attack for that, too. And it's all personalized. And what we do is when we see a spike, when we see hyperglycemia as part of our algorithm. We work not just on the medication piece, and that's too much of the focus, but we work on -- okay, let's try walking 15 minutes at a gentle pace after you eat, even if it's in circles around your kitchen table. Let's see what happens there.

Let's try some meditation for five minutes so we have all those things that we're throwing in there as well. The last thing I think is important to say to that, I don't know if you've seen it out there or not, but we've developed kind of -- I call it this alarm system that goes along with this, particularly now with COVID-19 out there. We want to make sure that if we have someone that's controlled really well. It’s in that time and range and all of a sudden, they're starting to have unusual trends, that's going to trigger us to outreach to say what's going on.

And it may just simply be I'm on vacation and living it up this week, but it also could be I'm really not feeling well. I've got a sore throat and I just was in Florida or something along those lines. So that would prompt us to really get care and get some screening done regarding COVID-19. And then we had another guy really great success story that's been out there on the news for us, evangelizing that he had COVID-19 and we were really able to find who -- he’s a type 2 but he was on an insulin regimen. And his issue was he was so sick he didn't feel like getting out of bed and he didn't feel like eating. So, he was certainly at-risk. And he's one of those worried well people, for the most part, where he's going to take it because his doctor said so.

But in this case, that could be very detrimental. So, we were able, with the CGM to really help nurture him and make sure minute by minute that we knew what the right therapy was and we kept him in range the entire time. He didn't have any lows, didn't have any highs. We were able to do that by really micromanaging his care with our system in the background to be able to make sure that he was safe.

Kelly: It's our sense that patients can't just themselves, get Level2, right? It's all about the plan they're in and how they're chosen.

Dr. Meister: Yeah, I think the answer is, right now we're marketing direct to employers, but future state, we might have a direct to consumer offering. I think the key is having people inquire with their employers about, hey, can you offer Level2? I think this program would be great with me. I think getting that interest can get us a seat at the table to help expand what we're doing across the country.

Kelly: Is there any way that you can say the names of some of the employers? And how many people did you say are in it right now?

Dr. Meister: First question about the names of employers, we just have to look to see in terms of disclosures what we have and what I'm allowed to share with my employers, but I could probably give them an aggregate number.

Across our fully insured book of business at UnitedHealthcare, we offered the program and we're in the process of rolling it out. We've offered it to 230,000 participants. And we to date had 13,000-odd people and some change that we've taken care of. And regarding the UnitedHealth Group employees themselves who was first really self-funded client, our own parent company, if you will. We have about 2,000 members active right now who have just joined Level2.

And our goal for 2021 is to add another 10 to 15,000 active patients to the Level2 ecosystem from our various employers. We're really focusing a lot on the mid or smaller market employers because of a number of reasons. One, the benefit design cycle for the larger national accounts is about an 18-month timeframe. And two, look at the impact of COVID-19 and everything else going on, the mid-market seems to be the most adaptive and responsive and really asking for a solution.

Kelly: That’s so interesting to us. Why do you think that is?

Will Ferguson, Chief Growth Officer: This is Will. There’s just that tremendous amount of lift that's required for a Fortune 500 company to make benefit decisions and systems and platforms because they have potentially multiple offerings and committees involved. So it's just they possibly have procurement involved. They have HR involved, benefits involved. It’s the nature of the business about large space.

Kelly: It's kind of rare in my experience with the CEO so close to all of these numbers. What’s the longest duration where you would say a patient has actually kind of -- I don't know if the word is achieved, but has moved to like diabetes remission.

Dr. Meister: Yeah, sure. So, first, I want to be clear about our definition of remission, because there's several definitions of remission out there. But what we're using and I'll explain the why part -- is a hemoglobin A1c of less than 6.5% and managed on metformin or no medicine alone. And the reason we selected that was really by looking across our book of business, meaning UHC aggregately and we have over 4 million diabetic lives to really look at. We looked at the costs. And when do we start to really improve the curve with costs. And that seemed to be the area, when I get people into that prediabetic space, that's the biggest opportunity. Of course, I'd be thrilled to death if I got somebody to have less than a 5.7% and on no meds. And our goal is once we keep them there for at least two months, we put them in that bucket. But we don't kick people out of the program.

We treat this just like we would Hodgkin's lymphoma. You’ve had the disease. We need to be aware of it. We're going to always look in the rearview mirror. We're really going to remain the accountability partner. And that's where I think we tried to figure out because coaching has such a funky connotation. It either can make your stomach feel upset or you can feel happy about having a coach. It depends on what your personal insights have been.

And what we've decided is we're trying to position the coaches as your ally. They're going to be your advocate. They're going to be someone to really help you. So, we have a lot of people that around the holidays are reaching out and saying, hey, I've been off my CGM, my counts, everything’s great. I know I'm going to cheat and have a difficult time. Can you give me another one?

And the answer is absolutely yes. So, I think that type of long-term relationship or longitudinal care is also key in this area. But again, we relaunched the program about October this past year. I don't have a percentage to share offhand. We're just mining the data now to really understand how many people have already gotten to that sweet spot, that date of remission. But we're also quantifying the expense or cost of the number of medications we've discontinued or dosage reductions to really look at some financials around that.

The other area that we're also looking at quantifying is the service optimization. So truly understanding it's going to take some time on this because we want obviously year over year data to have it be more valid. But to understand with our new virtual care platform, how often are people going out of network for care? How often are they going to ER for unnecessary care, urgent care, et cetera? And one thing that I'm really excited about there is the fact that we actually own the data so we don't have to rely on trying to estimate or getting data from an external source.

Mr. Ferguson: And I'll just summarize to with that piece is that we're not offering any ROI guarantees. We're doing a whole health dollar for dollar, fully delegated, wholly owned risk guarantee. So once we identify the individuals, we will take risk on their whole person, as Amy mentioned before. So not just the diabetic span, not just things that are related to the condition. We're going to manage that entire individual for the entire year, take on all risk for that person.

Kelly: We can’t thank you enough for all you are doing. I’m truly overwhelmed by the intelligence and the density of goodness in what you are doing. Please thank your colleagues and team members for the remarkable changes they are enabling.

--by Ani Gururaj, Albert Cai, and Kelly Close