Why Digital Health Pilots Fail: UC Davis Health's Dr. Reshma Gupta

Speaker A: Hi, uh, everyone, this is Eric Glazer. And welcome to Bright Spots in Healthcare, produced by Bright Spots Ventures. Our mission is simple. Find what actually works in healthcare and help you all scale it. Each episode, I sit down with senior leaders from health plans, health systems provider groups, and other innovators across the industry to cut through noise, share honest conversations, and unpack proven approaches, delivering real results in the real world. This isn't a show for spectators or buzzwords, really. It's for builders. It's for innovators. It's for the courageous, the ones not afraid to push progress inside their organization and beyond. If that's you hit subscribe, you're home. This is your Bright Spots in healthcare. Podc.

Speaker B: Eric Glaser recently hosted the ROI Centered Care Summit, produced by Bright Spots Ventures in partnership with Tidal Care and the American Telemedicine Association. This episode features a practical conversation with Dr. Reshma Gupta, uh, chief of population health and accountable care at UC Davis Health, on why so many digital health pilots fail and what it actually takes to redesign them into models that scale. Using UC Davis Health's Hypertension RPM program as an example, Reshma shares what broke, what changed, and what health systems need to think through before launching the next pilot. Here's Dr. Reshma Gupta. Well, great. Um, I'm always excited to talk with Eric. I feel like we've had a few conversations in the last year, and it's building on the conversation and digging deeper in. And so I'm actually really excited about this topic because I think it gets to the crux of the reality of implementation, which is great. Um, so advance to the next slide. I'm actually going to go over some of the challenges, um, that I think nationally I've seen not only through our local site, but also talking to folks all over the country that kind of do this work. And if it's implementing AI or it's implementing digital tools like remote, uh, monitoring, where does a rubber meet the road? And what are those challenges? And then I want to share a little bit about our specific story from one example, which was our remote monitoring program with hypertension and chronic disease, where it fell apart, how we had to revamp it, and then some of the lessons that we learned, and we'll open up for discussion. Advance forward. So this is kind of my summary right off the bat of a high level of, like, where a lot of cracks come in when it comes to why pilots flop or fail. Um, and I kind of organized it in two ways. One is some of the strategic limitations, and some are Purely more operational. So strategically I think there's so much hype around, um, AI and it is going to transform what we do, um, and these digital tools. But without having a really clear strategy that aligns where there's priorities and investments and resources and time to think about the best way it's going to implement, things will flop. So I always call that like the thousand flowers blooming. There's a million ideas, but without kind of picking those priority areas to really execute well in, a lot of them flop. Um, the second is I'm a chief of population health. I think about populations all the time. But defining our population of who's best going to benefit either from an AI tool or from a digital health strategy, um, can miss the ball if we don't do it well. So mismatching the intervention to the target population, um, is a huge deal. I'll share one quick example on this one because I think we'll get back to it with our local story. Um, there's a number of folks that are getting into remote monitoring with hypertension and the idea is that they're kind of creating the roi and that discussion with your finance teams in launching these programs that broad scale hypertension remote monitoring is going to reduce hospitalizations as the roi. In reality, if you define the specific high risk population that ends up in the hospital a lot, that's probably true. Advanced heart failure for example. But when you look at that population being low risk, stage one, even stage two hypertension, that's probably not going to be the roi. Those are not the patients that actually end up in the hospital. They might develop more advanced disease five or 10 years later. But to say that we're going to get that as part of the ROI makes it not sustainable because you may not get those outcomes. What it is really meant for is preventing advancement of blood pressure and chronic diseases over time. And that's really got to be part of the equation of what you're selling to the organization. So sometimes defining that population, um, is really, really important. The other is creating the roi. Just like we said, sometimes it's a collaboration between practices, delivery systems, um, if it's not a company who's delivering it themselves directly to consumers, but it's kind of interacting in traditional healthcare, um, the staff cost between what it means to hire up your own staff to do parts of the workflow versus using the vendor staff and services, um, really vary depending on what's negotiated in the contract. And sometimes not getting that on point can actually make your costs higher. Throw your ROI off Um, there's also billing considerations or alternative payment models that we have to look at. The bank, the payment reimbursement rates. Because right now, a lot of time in some of these areas, the ROI is actually not there unless there's kind of changes to the payment and billing, um, processes, which I'll share our example in a little bit. Um, and then the third is the right partnerships from a strategic perspective. We've got to have leadership, engagement, prioritization, all of those components to make sure that if we're bringing in a technology, bringing in a tool, it's not sitting on a shelf, but there's a team that's really ready to implement it and build it in a way that's going to be sustainable. So this slide really talks about the strategic limitations, but if we move forward, we'll get to some of the operational ones. And some of this focuses a little bit more on those digital tools. But, um, one is role definition. Right, so what is the internal team doing? What is the vendor doing? What is the vendor allowed to do to even interact back with your own health system or practice from just an IT perspective, functionality, if we don't map these out up front, you can get redundancy and then we're actually spending money on things that we don't even need. Um, and that throws your ROI off. Um, as well. In the business model, uh, there's various workflow and protocol things that are really needed to think about. So having standard documentations to coordinate between various teams and integrating this technologies. Um, having EHR tool, uh, optimization, good communication with patients so they're not receiving confusing information from various folks reaching out to them across, you know, a practice, a health system or, um, a company that's supporting this work. And then obviously the patient fatigue that can go with that. Um, the other is, and this is something that caught us by surprise is the duration of the program. So how many patients do we need to see a month? How does the billing work with that? Um, and over how much time are patients cycling through something like, for example, remote monitoring to make the ROI make sense? Um, and then the last part is actually the devices making sure to set up a situation. If we're thinking about use, uh, of devices and remote monitoring that situation where it's not going to be burdensome, um, where now we need staff to track down devices and, you know, pick them up from folks when they're done with them, um, that can really throw off your model. So all of these things have to be taken consideration. And when they're not. This is often why pilots fail. So let's go to the next slide. So give you a little background around us and also kind of the importance around picking that population. So in Population Health, this is kind of a standard diagram that we use a lot of the time. But if you look at our patient populations, we have a small group of patients with very severe disease, multiple medical conditions that often account for 40 to 50% of cost and utilization. There's another large chunk of patients that have many chronic conditions but maybe are a little bit more stable and have an opportunity to kind intervene with a lot of these tools and technologies. And then there's a number of patients at the bottom of the risk triangle, um, that maybe are at risk, but it hasn't happened yet, of poor health or generally are healthy. And our goal is really to keep patients, um, uh, from moving up this pyramid by creating a good, uh, medical model that can integrate these technologies, if it's AI or digital health, to ultimately keep people healthy at home with the right care teams and seen in the right settings. Next slide. So a lot of what we do is around chronic condition management and population health, you know, improving the outcomes that we've talked about, the access, the experience of the care team and the patients. But now is the time where, you know, we're going to be able to integrate, if it's care at home with technology, AI, remote monitoring into these at, uh, kind of rapid scale. So if you skip forward, I want to get in a little bit into kind of our local story. So UC Davis Health, for those of you that don't know, actually is the second largest, um, UC when it comes into patient population and geographic area. We see patients all the way from the border of Oregon to Nevada, Central California to the Bay Area of California. Um, so we initially, just as one example, um, wanted to start a remote monitoring program at UC Davis, starting in primary care and focused on hypertension. So across the clinics in this broad area, I gave a population breakdown. We see about half a million patients at UC Davis in general. Um, but we have about 100 primary care physicians that see, um, about 150,000 patients in primary care and then again about 30,000 pediatric population. Um, we were targeting initially for phase one blood pressures that were above 140 over 90. We've kind of moved to even newer guidelines since then. But from our estimate when we started, we said there's a few hundred patients, probably about 400 patients that could benefit from remote monitoring. And just talking to various folks, we, um, know that not all patients um, agree to participate in programs. And so we still thought we had a few hundred patients, um, accounting for that, uh, to target over time. Next slide. So this program initially was actually set up by another team and kind of landed to us a little bit later after it fell apart. And the initial pilot was smaller. And like most um, health systems or practices, we have a culture piloting, which means smaller population, slower ramp up, learn the workflows and then grow. Well that really didn't work for this and it's kind of going off of the lessons I mentioned. Um, so first of all we defined our workflow. So patients were going to get blood pressure cuffs are going to monitor their blood pressure, they're going to get coaching and education around it with a care team member for three months. Then we would be recycling their um, devices and then providing it to the next patient in these three month cycles. And that was what the original pilot had been planned for. That was not a model that was actually going to work out is what we realized. We also started with about 20 patients a month, so I think it was 5, 10, 20 and then continued on to kind of ramp up over the year. So again a smaller program pilot and as I mentioned the, you know, returning the monitors. There's also a lot of learning in that first pilot around billing capture and billing audits and making sure that even if we're doing that care that we're really accruing the revenue on the back end to make the program more financially sustainable or self sufficient as well. Um, and then the recommendation was initially before all the workflows had been worked out and really talked about was that we had signed up for, um, the company we were working with to provide a nurse, an RN to help with some of the workflows. And then after the fact and after signing the contract, we realized that many of our clinicians locally and care team members were not comfortable with some of the escalations going to an external RN than the own care team that has continuity with the patient if they're more high risk patients. Um, and so that was something that then became duplicative. We already paid for it with the contract, but then we ended up needing to support it in house. And then the second was they didn't have the ability to gain access to our schedule internally. And so we ended up having to absorb that work anyways. And so we're kind of paying for something that we weren't really able to use or leverage in the end of the day. So these were kind of lessons learned that we Needed to revamp. Next slide. Oh, so thank you. So it was really vital to think through these impacts of, you know, before signing the contract, you know, that we have those lessons learned. And I think the big things we took away was that uh, longer program was needed for the patients, quicker tracking of those billing practices. Having the vendor, um, MA instead of the RN was probably all that we needed to show the ROI that we were looking for. And it really was in the best interest to see, spend the extra time to map these models out and where the pain points were. Because in the end, both from the company's perspective and from our perspective, we want this to sustain, we want the business to sustain. And what we don't want is a failed pilot where then there's a lot of hesitation to go back into this thinking that we're not going to be able to hit their ROI or kind of get to the goals that we needed. Next slide. So this is the program that we relaunched. So In January of 2025 we rolled out um, RPM to all of our clinics. Um, we did have tablets, ah, and blood pressure cuffs and kind of a vendor interface that they used. We have local team members, an ma, an rn, um, a pharmacist and a clinical part time clinician to help with some of the billing and education onboarding for patients in the program. And then um, our company also provided an MA for a lot of that, uh, extra hours, oversight and kind of escalation protocols to bring back to our team and then tailored education and resources. Next slide through that program. It was a six month journey. We realized we needed to get patients in the program for longer to actually hit outcomes. But also from a financial ROI perspective. Um, and so patients would start and every month they had kind of a clinical focus of education, how to check blood pressures, medication reconciliation, adherence, education, diet, exercise, all of those things focused in education as well. Each month they had a nurse and pharmacy visit each month that could, um, and if the patients wanted to graduate at six months they could if they reached their goal or they have the option to extend to 12 month program. Next slide. And ultimately they received the blood pressure cuffs as I mentioned, got the treatment and kind of agreement consent for the program and education. And there's various workflows between those team members to escalate out of range blood pressure readings, efficiency of patient onboarding reports. Um, there's various vendor protocols working with those escalation protocols that I mentioned. Video visits were used uh, for this program and then a lot of attention around the billing capture. So all of that was really revamped from the initial pilot. Next slide. And we were measuring outcomes a variety of things here from operational, from patient experience, um, looking at uh, actual blood pressure control and lifestyle changes and also revenue generation annualization as well. We built this model um, based off of the UM, AMA's framework about UM remote monitoring. Next slide. And ultimately by changing the program and it not being a pilot but being full scale, higher numbers, longer um tracking period for patients, um, we looked at our patients and actually said that patients that kind of enroll and are in this program complete the program by six months. We're averaging blood pressures that were above goal around 146 over 81 and by six months getting them into goal range which is less than 130 over 80 um, with an average about 126 over 73. And so this was uh, actually awarded um, a 50 greatest innovation um award ah by Popular Science and others. And so it's been a big big success I think within the organization learning from those lessons. Next slide. So big lessons learned. If we're thinking about AI, one thing we realize is um, or even from digital health that care teams want to have a voice and transparency and kind of these programs are rolled out to make sure that the right kind of trusted processes and input from clinical care teams are included to avoid bias. Um, especially for digital tools they're game changers from patient experience about engaging in their own care and getting to outcome. But I think the biggest lessons learned about the pilots failing was health systems. And our companies really want to ensure that fiscal responsibility from both sides. So we need to actually spend the time to think about efficiency and cost effectiveness um that these programs we have to also think about how they sustain programmatically but financially as well. So they're more seamlessly integrated into care models um to optimize that roi. And then it's really important to match the resources appropriately to the patient population and their needs of these are high risk patients or kind of lower risk patients, um, and how those match with kind of the staffing models that we build billing optimization that I mentioned. Um, and then the one thing is, you know the feel still remains really really nice novice in chronic condition management primarily in the ability to sustain that roi. So we have to be really thoughtful about how to create models. And what we realized is there's very few models out there that are doing it. And a lot of companies that are going into chronic condition management uh around these things are closing or kind of closing that line of business. Um but these are the lessons we learned that we needed more time, um, a longer program and more patients, not a pilot to make it successful. So I'll kind of hand it back to you a little bit for discussion. But I just want to end with this last slide around just some key points here. Is trying to, for us not jump on the shiny object effect. Really being able to understand what our needs are to kind of guide our strategy, um, going forward. The next I think I talked about was transparency. If you move forward with the slide, um, and making sure that both in measurement of outcomes as well as including the care teams and the company, to really be honest about what the needs are and where there's concerns is um, really vitally important. And then the third one moving forward is that engagement in a similar way. So I'll pass it back to you Eric, for a little bit discussion.

Speaker A: Yeah, there's so many places I, I want to go here. Um, so one of the frameworks I, I laid out in the beginning of today's summit, uh, was to look. Yeah. And I, and I gave guidance to everyone watching and listening today was to talk about like, where is friction? Listen to where is friction being removed? And so in this case as we're looking at like, okay, what are the friction points as it relates to causing failures of these pilots? And if I can, because I'd like you to expand upon what I wrote down because I don't think I caught nearly enough of these. Uh, one of them is you are documenting the workflows. But there was definitely a gap in understanding uh, the workflow and the integration points specifically. So uh, I wrote down documentation of workflows and coordination of CARE and the protocols. I don't think you used the word protocols, but you insinuated it a few times and then, and those integration points. And the other important point was once you get that workflow and system in writing, you need to get a buy in from the providers so they could see all the different points where maybe they could get uncomfortable versus retrospectively. So as a one friction point, and I know there's a few of them, uh, the lack of alignment with system wide models was the other one I wanted to ask you about before. How would you like, how would you help expand upon that for everyone listening?

Speaker B: Yeah. And I think even the system wide, I think they integrate to each other like the things that you were going to ask about. So I think the biggest thing is that whoever is the implementation group. Right. Whoever that is, if that's a practice or if that's a whole health system, whatever group that is, is that there's a lot of concern just, just having run a number of these programs now about what happens for patients. Like these are the quetza, who are the patients that are not going to receive this service and, and can we build a program that's going to allow it? So who's being left out? Right. The second concern that they have around transparency is what about the high risk patients that need a lot of high tech when we're automating things, they get really concerned about the quality of care and the safety around those things. Those are the two number one and two questions. So if you're building these programs in the planning and the strategic planning about even building the contract out with a company to join efforts on this, that's the friction point that they feel and that's where I think things fall flat, um, at least from an engagement perspective, the transparency perspective. And the best way to do it, to get around it is to say that this is really part of the clinical care model. So who in the institution is actually one going to implement it, but who owns the clinical care model? So for hypertension maybe it's primary care, cardiology, nephrologists, you know, for diabetes it's probably endocrinologists and primary care usually, but there's other groups as well, pediatrics, obgyn, that often could benefit from these types of programs. And so it's bringing those folks together and just as a leader knowing I've got to bring the right people to the table. So even if I'm working with our uh, company, I'm using some of the ways they've organized and adapting it using my common sense. I never deploy anything without making sure that I've had multiple meetings with those key care team model owners is what I call them, because they're going to guide them to ask the hard questions and they're going to. I would much rather find the cracks in the system before I actually try to launch it and have a plop. I think the part around the ROI is a little different though because I think those teams don't necessarily see that either. Um, and that's really coming down to good business modeling. That's a different issue. But this point around transparency, engagement, clinical caremal, I think they're all integrated, they're one in the same.

Speaker A: Okay, okay. And as far as matching, so you've got this clinical care model matching the resources appropriate to patient populations, I think is what you said. And I was curious, like what you really mean by that. And then also, and this is the most important question, how do listeners apply that thinking at their organization?

Speaker B: Yeah, good question. So I'll be really concrete about it. So think about the same example, like remote monitoring for hypertension, but um, you know, kind of stage one, lower risk hypertension. They mainly have hypertension, maybe diabetes, but not too much else. There's a large patient population like that mainly seen in primary care. Then you have the situation of someone who needs remote monitoring for blood pressure, but they've advanced heart failure with four hospitalizations in the last year on multiple medications where the blood pressure goes high, goes low. Um, and so there's a lot of like, not a lot of space to make error. Right. And there's a lot of maybe side effects they've had to medications where if now we're changing the medications based on the blood pressure monitoring, if that cardiologist is not involved, the patient could get really sick and end up in the hospital unintentionally just because they're so sensitive with how, you know, their disease is. So when we think about remote monitoring, um, a lot of the work that happens especially to help health systems, the benefit is that 20, ah, four, seven coverage and monitoring. But if things go wrong and the blood pressure is too high or too low or they're symptomatic, whatever that might be, um, you know, if you got patients that are not that sick and we're just trying to optimize their blood pressure, we probably only need like a medical assistant and then they can, if the blood pressure gets really high, we don't need to respond in five minutes, we can respond the next day. It's literally no big deal to have them call in and then use our teams. Right. And kind of traditional triage processes or they, you know, they can do it too and have an ma. Um, if you write up really, really standard protocols of like what to do in this situation and what medication, you know, protocols to use. But when you get that other patient that's had lots of side effects, really sick, right. Been in the hospital a whole bunch, um, that's really complicated. And so you're going to need to have an RN that can really like triage and navigate that situation on the phone with a patient when those situations come up up in escalation protocols and that person we have to really think about is that best person to sit within your own healthcare team that sits right next to your cardiologist to make joint decisions, or is it better to outsource that, you know, with the company doing that. Um, and what we did is we went with the RN right in the initial pilot, but they couldn't even schedule. So even once they triaged, they had to come back to us and our staff had to actually do all the work anyways. And so it didn't save anybody time. It became duplicative and costly.

Speaker A: Does that get back to, like, the workflows that get documented? Even though that was intended, you guys were on that. That the workflows failed to map out the integration points necessary to make the system work, or, like, where was the failure there? Because it sounds like you were trying to do all the right things.

Speaker B: Yeah, we were trying to do the right things, but I think we didn't realize, um, I think ahead of time for the team that was initially piloting that. But what does that workflow really going to look like? So after the escalation, what's going to happen after they triage the patient? They're going to need them to get back into care to make that final decision around medication. Because so many times it was complicated with those patients that they weren't sure exactly what to do being external from that care team. And so it ended up coming back to the care team to have to reassess the whole thing. And we just kind of thought, well, they would just figure it out, they would just do it. We get to our outcome. And that wasn't the case. They needed to consult, actually the patient's continuity provider. So there was no benefit to outsourcing that piece of it. Um, and then the second part was scheduling. They ultimately needed to get in the schedule, right, to have those conversations, and they couldn't schedule within our system. So what we realized is that for that really, really kind of top of the risk triangle group, as I'm saying, that that patient population needed the vendor monitoring, we needed the devices, we needed the integration with the ehr. Like we still need that program, but the person to actually respond to it actually had to be embedded. And then we actually have on call night service, weekend service, to deal with those situations that we've created internally to our team. Um, so that staff might, from the vendor might call the patient and assess the symptoms and at least get them to the right site of care in case there is an emergency or something. But to really change the medications and the plan, they had to integrate with us. It's very different at the patient's low risk because what they need to do is not as complicated. And we really could outsource a lot more of that to make it cost effective and see more patients.

Speaker A: So what's the uh. I'm curious if you had to do it. Um, I want to try to think about the right framing. I guess I'll frame it as you're going to go into um, similar, you know, going to go to a hospital system on the east coast. I won't name one, but let's say we're in somewhere in the Carolinas. Pick one teaching hospital, kind of university based, whatever. And so uh, if you were going to come in as a consultant, what, what would be your advice on how to better plan a. Protocols and process for a pilot? I mean it sounds, it sounds almost like uh, it sounds almost like you need to do a couple dry runs with different patient pilot, you know, patient types. Uh, not a real, almost like you know, pretend dry runs. I mean I, I just got uh, done listening to a long podcast on Walmart and you know, and Sam Waltman, Sam Waldman in the early days, besides going into his competitor stores, any new store that opened up in the beginning of Walmart, he would be there for uh, going through everything, right. And the whole, all the workflow of like the customer and everything that happened. So I'm wondering if that kind of detail of going through three or four dry runs of different uh. You tell me. M. You're the consultant.

Speaker B: Yeah, no, I, it's, it's what you're saying. And I think there's a responsibility there though that I. So there's some. I would give you know, kind of benefit of the doubt of, you know, for a lot of folks that are doing this, both companies, you know, they don't have a lot of models where they've seen it fail. So therefore they already know what to ask for and where things fall apart. And same with that, uh, the system side. Right. They haven't maybe done it as many times before and so they don't have the vision of like automatically what to look for. So I do think the field will get better, but just everyone having more experience over time when they roll out new things. But it's exactly right. If you are rolling it out that first time and you don't know, then it needs extra attention to just what you're saying. It's implementation design. Right? That's the name of it. Right. Or it's basically just getting out there on the ground and saying what we do now is we have clinical scenarios and we say, okay, if this patient is, you know, this part of that wrist triangle, they're low risk, they're that Medium risk or that high risk patient. And I actually create like a clinical case and make them complicated. And then I have the teams actually walk through it like with those clinical, clinical team members, those clinical m. Um partners and the vendors and say, so what happens here? And if this happens, then what happens? And then how are we dealing with this? Who is going to own that? How do we document that? How do we get back to the team? And it takes, I mean it takes a little bit of time, but it's not a lot of time. But if you don't do that and the pilot fails, that's much more expensive than a lot more wasted time than probably a handful of hours together. Right. Kind of mapping this out. And I would say that that's a very practical way, say low risk, high risk role, responsibility, like very basic stuff of implementing almost anything. But I just don't think it's often done. I think there's, right now there's a chase of that shiny tool. Get it out there. You know, we're, that we paid for this thing, let's get it rolling, let's get it launched in a month. And those, they just said like, we'll figure it out in parallel. But to be honest, on the ROI side, you gotta figure some of that out before you sign the contract.

Speaker A: So let me, let me ask you this because it sounds like what you're saying, Rashmont, is this, this you need to spend um, some time and energy. And energy is money in almost dry running some of these pilots first and really working out the points. I want to talk about the vendor's obligation to that too because I think the vendor has the most experience. But we'll get into that in a second. I'm going to put a pin in that. So that's number one. So organizations listening, make sure that you're running some dress rehearsals. Then the pilot actually needs to be more aggressive than it is now because you said a couple of times the patient population is not enough. So spend some human ah, resources in time and money. Time is money in doing some dry runs or dress rehearsals and workflow. Then do a bigger pilot to really make sure that you're giving it a fighter's chance to work and you can see the real deal data. So if I was going to, putting, laying that out almost as a, let's call it a, I want to package this up as a, a pilot process. Okay. We're going to call it the, the Gupta pilot process. Gpp. Okay. How do we make the gpp, uh,

Speaker B: scalable and repeatable yeah, yeah, I think so. I'm just going to take what, what you said back. So I think doing dry wands that are not casual, that are thoughtful, that are case based, that go through all the scenarios has to be done automatically before you even set up the contract to even know what do we, can we not do and what do they need to do to get the contract correct. The second is when you start working with the vendor being able to um, and you start really kind of launching the program, probably having a month with a lot of reflection about the workflows working and all of that. And that's what a pilot technically is probably appropriately used. But then really when you launch the program, getting past that, that month, you've got to go to bigger scale to get the ROI correct. If you sit in that pilot mode for six months or a year, that's what a lot of people do, then you're never going to hit the numbers and that pilot failed because you got stuck in almost the preview um, instead of actually like what the true pilot. So you should never evaluate um, what you're doing in the first month for the roi. Right. That's just getting people used to the workflows. The true pilot has got to be longer where then you're going to evaluate and ensure the roi and that's got to be at a bigger scale. All you know.

Speaker A: Got it. Okay, so, so let me ask you this about the vendor. So I as some people know spent 12 and a half years, a little more 12 and a half years at Teladocal. Teladocal bought Livongo, big CCM provider. Okay, so no, uh, longer work there? Um, probably, but a little bit of bias here. But you said that to, to the audience, I'm curious. Like the, the car cardiometabolic sort of chronic care management companies are still, it's a nascent industry and so there's things that need to mature and we have to be aware of that. Tell me more about what you mean about that. So uh, both providers and maybe the solution providers they can learn from and the second thing I'd like to ask you to talk about and then we're going to move on because we're running over now. Uh, is isn't it a differentiator for the provider to come in and say here's all the places that there could be screw ups and shouldn't they be coming in with the workflow? They've been doing it in different provider organizations so they should know about the billing problems. They should know all about that.

Speaker B: Um, yes, correct. I'm going to answer that one first because you just said it. Um, uh, yes. I think there's a responsibility from companies who once they get more experience, I think the first time they're not going to have the ability to even know, but as they get more experienced, be able to say, hey, these are like common pitfalls. Our other, you know, prior people we've done business with have fallen. And let's just talk through these before we go ahead and get started because you may not realize that we should talk through these. And then I think a lot of it is on the system side, um, because they're the ones that do understand the workflows, right. And they have those clinical care models. But to, to be realistic, I think it rarely happens in reality. And a lot of times, and I think this is where pilots flop a lot, is we glom onto the shiny object and then we try to build the implementation of the IT in the first time we're building the workflow for the care model and it takes forever. And it's super frustrating for the companies as well because there is no already set care model and they're trying to do it all at the same time. And what I often say is when you bring in technology, it amplifies like what's already there. So if you don't have a care model, it's going to amplify dysfunction and discordination. And so this health system does have to take that responsibility. So I'm going to answer that one first. Um, repeat your first question again then.

Speaker A: So the f. The first question was about. Geez, I already forgot it. So we talk. Oh, cardio, metabolic and the complexities. God, I can't remember it now.

Speaker B: We'll come back, I'm sure someone.

Speaker A: Yeah, we'll come back to it.

Speaker B: Let.

Speaker A: Let's let. Yeah. So there's a couple questions from the audience I want to take. Um, it strikes me though, one thing you just said, and I wonder your thoughts on this. One of the things when I was running solution strategy, um, I used to promote that we do is somewhere in the presentation talk about where pilots and programs fail and talk about some of our failures we've seen and why and what has been done to mitigate those. Never happening again. And uh, it one shows the experience. It's a teaching moment to the provider organization in the room or the payer, depending on the scenario. And it also builds credibility to the solution provider that not only it shows that one, they admit that they're not perfect, but two Their experience that they have. And I'm wondering your thoughts. Has anyone come in there and showed you, like, okay, this is where things mess up, so you don't mess up?

Speaker B: Yeah. We actually interviewed, uh, uh, probably like 20 places when we were kind of revamped. We were given the initial pilot and it was like, this is broken. And before we cancel the program, can we try something? And then we're able to kind of go around. But so we did that, we called up a bunch of folks that, you know, we know we're kind of in this space in general, you know, meeting some at a conference and other places, and we just asked them, how are you setting up? What are you facing? And they were just facing a lot of the same things. Um, and I'll, I'll bring this up because I remember your first question now, which had to do with why does chronic condition fail? Oh, yeah. And so this was one of the things that came up. We just kept realizing, like, they're not hitting their roi and then the programs have to cancel or the, you know, one of our partner institutions, like, we use the same company, and the company told them we're canceling this program because it's not making us money, essentially. Like, it's not making sense on the books for us. So then we worked with them at the same company, um, to say what didn't work and what is working now or like, what could work. And it was just, it was the money at the end of the day. So we. What we realized is in chronic condition, you're not going to get an effect in three months. You really have to think longer, just from a clinical perspective, probably at least

Speaker A: 12 to 18 months, because there's.

Speaker B: And we barely got there at six months. And then 12 months, we really saw that sustainment. And so we said, let's give them the option and let's ask our patients. And that's what they wanted. And a huge portion of our patients do extend from six to 12 months because it works. Um, so that was one huge learning. And I think that. And, um, so I don't actually think I got a lot of advice on solutions. And so that's why I think it was so novel for us, because we just had to use our common sense of what makes sense from a clinical model. Clinically, you're going to get those outcomes in a longer period of time. Don't expect to sell the ROI being that you're going to reduce readmissions and admissions for blood pressure alone. But maybe what it is is that we bill better and we get a high enough volume so that we're able to capture the billing. And then the billing has to equate to the cost of the staff and has to equate to the cost cost of the vendor. And then I'm self sustaining because I'm billing back. Right.

Speaker A: And that's the crux of the entire theme of today's session. So yeah, that's the key. A final question because I'm going to combine three guest questions into one. So we have three questions, I think, and not going to happen to one. Uh, regarding the individual patients that were part of the program and you experienced some challenges. Uh, uh, a, um, were there social drivers involved that were, went unaddressed? Was there like language or loneliness? And then did you see that significant drop off after the first few months in engagement?

Speaker B: Yeah, um, 100% we did see that. But that's. I think I would assume that because if you look at what's out there just demographically, like in certain communities, there's high food insecurity and many of these conditions, diabetes, hypertension, heart failure, ckd, the things that we're in for, remote monitoring, they're diet sensitive conditions. Right. We need low salt, we need low sugar. And so, um, the social needs were huge, but mainly around like food and diet I think was the biggest ones. Um, and so I think that came with the education we provided and the ongoing back and forth. Right. For a long period of time to do that behavior change. So we did see reductions in folks that uh, you know, participate. You know, we can't go out there and fix the community of resources and their access to the grocery stores that are walkable and the safety of getting to those grocery stores with fresh produce. So we can't change everything from just this program alone. But I think it does add to that bigger clinical care model right. Where we're kind of chipping away at that over time, 100%. We screened for those, we identified a lot of uh, food insecurity and social needs and then tried to link them to the other services in our system to address them the best that we could.

Speaker A: Great stuff, great presentation. Really appreciate it. Thank you.

Speaker B: Thank you all. Ra.