CMS ACCESS Model, FDA-Cleared Data & Chronic Care at Home with Patrick Sheehan의 전체 녹취록

members have to drive three, four, five hours to get 51 00:05:06,506 --> 00:05:09,347 access to their entire care continuum. 52 00:05:09,475 --> 00:05:14,218 it shows you that again, technology has to be the answer to help scale it. 53 00:05:14,218 --> 00:05:25,364 that's why I founded Rainfall Health and that's why we're here talking to you and other amazing folks that we end up having on the program to discuss how those challenges that 54 00:05:25,364 --> 00:05:33,061 remain still you know rampant in healthcare and sometimes how do we make it just a little bit better? 55 00:05:33,061 --> 00:05:36,259 How do we make lives just a little bit better? 56 00:05:36,259 --> 00:05:38,951 And I you brought up the CMS 57 00:05:39,387 --> 00:05:42,339 ACCESS model, CMS has been doing a lot of innovation. 58 00:05:42,339 --> 00:05:52,064 I feel like not everybody realizes that a lot of innovation actually in healthcare and in other industries as well comes from government and government policy. 59 00:05:52,064 --> 00:05:56,316 I'd love to hear how do you all view 60 00:05:56,316 --> 00:06:08,963 the the model changes and the innovation coming out of CMS and also help us define the CMS ACCESS program through your lens and the Withings lens. 61 00:06:09,250 --> 00:06:10,581 Yep, absolutely. 62 00:06:10,581 --> 00:06:21,380 And I like the way you spoke about the continuity because one of the core principles that I've always had and Withings has always had and how CMS ACCESS is focused on is 63 00:06:21,464 --> 00:06:28,476 Delivering care in the home environment seems so common sense, but very difficult to execute on before the ACCESS model. 64 00:06:28,476 --> 00:06:41,330 So if you think about before the ACCESS model, integrating technology into healthcare was a difficult endeavor because of the way payments were structured, but also how 65 00:06:41,330 --> 00:06:43,701 distribution occurs in healthcare. 66 00:06:43,701 --> 00:06:48,323 Everything is being done through the patient-physician relationship. 67 00:06:48,323 --> 00:06:53,768 And payment has been done in the fee-for-service activity basis. 68 00:06:53,768 --> 00:06:58,171 Technology is not an activity based sector. 69 00:06:58,171 --> 00:07:01,594 It's really about outcomes and impact. 70 00:07:01,594 --> 00:07:11,282 So the ACCESS model, when it was first launched by CMMI, we were so excited and we remained very excited with our first cohort of patients coming in, is around this real 71 00:07:11,282 --> 00:07:16,887 idea that we can move away from activity based. 72 00:07:16,951 --> 00:07:21,673 reimbursement mechanisms towards an outcomes-based reimbursement mechanism. 73 00:07:21,833 --> 00:07:30,917 What's different here as well is unlike total cost of care or utilization-based contracts, this is focused on clinical outcomes. 74 00:07:30,918 --> 00:07:40,596 So improvement of blood pressure, getting weight loss metrics, improvement of cholesterol in a patient population. 75 00:07:40,596 --> 00:07:52,039 It's really exciting because it allows us to really focus on what matters most, which is those clinical outcomes and not necessarily, focus on the cost of care of that patient. 76 00:07:52,039 --> 00:07:55,550 We partner with organizations who focus on cost of care. 77 00:07:55,690 --> 00:08:07,733 The second piece is the distribution, and it's often overlooked, but one of the core pillars of ACCESS is making so that beneficiaries can get access to the technology and the 78 00:08:07,733 --> 00:08:09,442 care that they want. 79 00:08:09,442 --> 00:08:13,242 And that patient choice is a core pillar of everything here. 80 00:08:13,242 --> 00:08:23,302 So the ability for Withings to be able to reach out to beneficiaries directly and say, would you like to join this program? 81 00:08:23,302 --> 00:08:27,962 Because now Withings is a Part B medical group supplier. 82 00:08:28,122 --> 00:08:37,522 much like a cardiologist or a primary care physician billing Medicare, Withings is now billing Medicare in this outcomes-based payment model. 83 00:08:38,515 --> 00:08:48,469 ability to get paid on outcomes, the ability to engage and reach out and distribute our programs directly to beneficiaries are both, I think, transformational. 84 00:08:48,469 --> 00:08:55,011 And they both enabled us to integrate AI and technology much more efficiently because of that. 85 00:08:55,011 --> 00:09:02,054 So it's been really a refreshing new twist to how we deliver value-based care. 86 00:09:02,054 --> 00:09:07,367 And in my perspective, the biggest bet on value-based care since the creation of 87 00:09:07,367 --> 00:09:10,166 of MSSP 10 plus years ago. 88 00:09:10,166 --> 00:09:12,442 So we're incredibly excited about it. 89 00:09:13,843 --> 00:09:21,878 I I you know, before we even started recording, we had this conversation around there's very few people who find CMS models as exciting as I do. 90 00:09:21,878 --> 00:09:32,486 So I'm I'm glad that you know, we stay on the same wavelength about the excitement of what this means and not just for us, our team, our companies, but actually most 91 00:09:32,486 --> 00:09:36,628 importantly for the patients and the U.S. healthcare system. 92 00:09:36,669 --> 00:09:42,635 One of the things that you mentioned that I think is so exciting is the other entities that you work alongside. 93 00:09:42,635 --> 00:09:48,148 other care providers to actually deliver those outcomes that we all care about. 94 00:09:48,148 --> 00:09:53,401 So tell us a little bit about how who those are and how you all fit together. 95 00:09:53,401 --> 00:09:57,443 And what's been on my mind, especially in 96 00:09:57,443 --> 00:10:07,608 the last couple of weeks, is there's this great article that was written by a clinician that mentioned the, you know, had a provocative headline that said, the healthcare 97 00:10:07,608 --> 00:10:09,145 system is now broken. 98 00:10:09,145 --> 00:10:12,651

It's doing exactly what it's intended to do, which is 99 00:10:13,267 --> 00:10:18,010 separate pieces delivering their own care in their own silos. 100 00:10:18,010 --> 00:10:18,860 It's never designed 101 00:10:18,860 --> 00:10:19,881 as a system. 102 00:10:19,881 --> 00:10:22,623 And I think that's what these models are allowing us to do. 103 00:10:22,623 --> 00:10:33,659 So shed some light on how you all work with other care providers, other care delivery and technology companies to make that impact. 104 00:10:35,051 --> 00:10:37,619 Yeah, I like the way you framed that. 105 00:10:37,619 --> 00:10:44,402 The system is designed to be somewhat fragmented, especially with the fee for service system. 106 00:10:45,378 --> 00:10:53,063 So we approach this, we have a history of working with health systems, ACOs, initially around remote monitoring. 107 00:10:53,063 --> 00:10:58,556 So we have dozens of health systems are working with us today with our technology. 108 00:10:58,556 --> 00:11:11,503 Now in the ACCESS model, which is Withings delivering care for hypertension, prediabetes, high cholesterol, obesity, ASCVD, the CKM spectrum. 109 00:11:11,995 --> 00:11:19,671 The way we're approaching this is integration back to the traditional healthcare system is not a nice to have, it's a must have. 110 00:11:19,831 --> 00:11:28,938 Now we're leveraging those EHR integrations that we have with health systems on remote monitoring, but we're doing it now for care coordination. 111 00:11:29,078 --> 00:11:35,063 And how do we define in scope of what Withings is delivering care on? 112 00:11:35,063 --> 00:11:41,722 So, you know, those are things like getting labs and managing medications, all with the purpose of 113 00:11:41,722 --> 00:11:44,643 let's have this patient be on guideline-directed care. 114 00:11:44,643 --> 00:11:49,494 There's nothing new under the sun here when it comes to managing the spectrum of conditions. 115 00:11:49,494 --> 00:11:59,736 What is new is how technology can enable it to touch more patients and how clinical variability when it comes to getting patients on guideline-directed care can be more 116 00:11:59,736 --> 00:12:00,796 seamless. 117 00:12:00,797 --> 00:12:11,354 So we're integrating back with PCPs, we're, as CMS would call it, co-managing patients, and it's been a really interesting new layer of care. 118 00:12:11,354 --> 00:12:19,167 because we were going through credentialing for another payer and it was, are you primary care, specialty care, urgent care? 119 00:12:19,167 --> 00:12:20,928 We're like, we're none of the above. 120 00:12:20,928 --> 00:12:21,898 We're something new. 121 00:12:21,898 --> 00:12:30,361 We're a technology-based care provider that is there to integrate back to the health system, integrate back with the PCP. 122 00:12:30,361 --> 00:12:31,872 It's a different layer of care. 123 00:12:31,872 --> 00:12:37,074 I'm curious, Eddie, because you see so much on the other side, right, on the health system side. 124 00:12:37,074 --> 00:12:39,971 You know, when you think about these expansive 125 00:12:39,971 --> 00:12:48,814 CMMI models like ACCESS, which are different than the more condition specific ones, how do you kind of think about that from your lens? 126 00:12:49,867 --> 00:12:50,678 Yeah. 127 00:12:50,678 --> 00:12:57,403 I think for for us we're kind of building the ship as we sail it with these models. 128 00:12:57,403 --> 00:13:09,613 And I think that's the most important thing to understand is we actually get to architect a way to work together and set up these systems from the ground up. 129 00:13:09,613 --> 00:13:17,059 I think one of the hardest things in healthcare is ripping out existing infrastructure to input new ones. 130 00:13:17,059 --> 00:13:20,041 I think this is the chance and I think that's what 131 00:13:20,061 --> 00:13:25,546 CMS and CMMI have been doing that's so brilliant is starting afresh. 132 00:13:25,546 --> 00:13:26,526 So we can actually 133 00:13:26,526 --> 00:13:32,551 architect the system of our dreams as if you will, you know, as we move forward. 134 00:13:32,711 --> 00:13:36,174 And that means that we get to define these relationships from the ground up. 135 00:13:36,174 --> 00:13:41,358 where do we find that we want data to flow? 136 00:13:41,358 --> 00:13:47,724 Where do we find we want actual care delivery and services to go? 137 00:13:47,724 --> 00:13:49,973 And how do we get closer and closer? 138 00:13:49,973 --> 00:13:51,874 Where the patient already is. 139 00:13:51,874 --> 00:13:54,316 So from our lens, we think 140 00:13:54,316 --> 00:14:03,282 it's mostly about how can you create a unique experience for every patient. 141 00:14:03,342 --> 00:14:05,774 And it's almost less about everyone else. 142 00:14:05,774 --> 00:14:18,672 we spend a lot of time with health systems, and I think that you know the best thing to deliver patient satisfaction, quality of outcomes, and cost management is actually to 143 00:14:18,672 --> 00:14:19,805 interact with the patient. 144 00:14:19,805 --> 00:14:21,826 only as much as you need to. 145 00:14:22,226 --> 00:14:23,467 That's so interesting, right? 146 00:14:23,467 --> 00:14:23,667 you'd 147 00:14:23,667 --> 00:14:29,529 think that if you wanted to be really responsive, you spend much more time with the patient. 148 00:14:29,529 --> 00:14:33,545 Actually, patients want you to only be there for as much as they need you. 149 00:14:33,545 --> 00:14:43,806 if they're doing fine and they can manage it with, some of the devices and stuff, for example, you all provide, I actually think that's the win is you can actually manage the 150 00:14:43,806 --> 00:14:47,487 outcomes that the patient cares about, the clinicians care about. 151 00:14:47,681 --> 00:15:00,929 and what Medicare as the payer cares about by actually engaging a little bit less, but at the right moments, at the right time with the right data, so that

you're not recreating 152 00:15:00,929 --> 00:15:04,511 and you're only engaging at the points that create that value. 153 00:15:04,511 --> 00:15:14,549 So it really is about with the evolution of AI and where it is today, I think we can actually manage this sort of 154 00:15:14,549 --> 00:15:20,353 as you alluded to a long-promised dream of value-based care, that we can actually personalize it. 155 00:15:20,494 --> 00:15:26,918 And that leads me to my next question for you, which is how are you all leveraging AI? 156 00:15:27,939 --> 00:15:36,107 and how do you think it drives the patient outcomes and experience that are so core to clearly what you all provide? 157 00:15:37,810 --> 00:15:38,551 Yeah, 158 00:15:38,551 --> 00:15:47,175 and it segues, it's a good segue from what you mentioned earlier about the experience and engaging the patient at the right time. 159 00:15:47,175 --> 00:15:53,354 In my past, one of the things I've both loved about value-based care, but 160 00:15:53,354 --> 00:16:08,658 have seen friction is the clinical care models that are developed by organizations, whether in health systems or technology-based organizations, are in a vacuum very 161 00:16:08,658 --> 00:16:10,959 effective. 162 00:16:11,179 --> 00:16:14,600 they have, there's no mission without margin. 163 00:16:14,600 --> 00:16:15,960 In a vacuum, they work. 164 00:16:15,960 --> 00:16:21,041 What doesn't always work is does a patient want to be a part of this? 165 00:16:21,322 --> 00:16:22,678 And 166 00:16:22,678 --> 00:16:34,643 How do you handle these multiple taping of pieces together, especially hardware pieces that are required for CKM conditions? 167 00:16:35,384 --> 00:16:49,850 What's happening here is the patient is always a consumer, and there's this huge interest going on in healthcare around the use of wearables, the use of at-home devices for... 168 00:16:49,974 --> 00:16:56,816 self-management, understanding oneself better, and that's where we're meeting the patient where they are. 169 00:16:56,816 --> 00:16:59,637 So, Withings has this long consumer history. 170 00:16:59,637 --> 00:17:11,370 We have over 15 million users around the world, 5 million users in the US, almost a million are 65 plus, and we're starting the relationship with the patient with, want, the 171 00:17:11,370 --> 00:17:14,661 patient wants, the consumer wants to know more about themselves. 172 00:17:14,661 --> 00:17:16,182 Let's start with that entry point. 173 00:17:16,182 --> 00:17:17,850 Here is a... 174 00:17:17,850 --> 00:17:21,653 a scale that can measure your body composition that can do a lot more. 175 00:17:21,653 --> 00:17:24,455 Here's a blood pressure monitor that can do a lot more. 176 00:17:24,455 --> 00:17:29,458 And we start with that relationship and we can deliver medical care on top of that. 177 00:17:29,458 --> 00:17:42,337 Where we're integrating AI is how do we automate all the administrative tasks, and I'm sure you're well familiar with all this, that increase the cost without improving the 178 00:17:42,337 --> 00:17:44,328 outcomes. 179 00:17:44,328 --> 00:17:47,070 you know, getting a patient onboarded, 180 00:17:47,098 --> 00:17:56,022 is a set of administrative tasks that don't improve the outcomes, but are requirements for us to just deliver care. 181 00:17:56,022 --> 00:17:58,183 So we've applied a lot of AI there. 182 00:17:58,183 --> 00:18:08,277 And now what we're looking at is how do we start applying AI so that that promise we make to that patient, right, of you're going to understand more about yourself with these 183 00:18:08,277 --> 00:18:11,048 devices and all of our markers are FDA cleared. 184 00:18:11,048 --> 00:18:16,710 We're applying AI to help surface information, help drive insight. 185 00:18:16,790 --> 00:18:18,030 to the patient. 186 00:18:18,611 --> 00:18:22,531 We're developing AI right now, and it's been very exciting. 187 00:18:23,312 --> 00:18:35,115 How do we help our clinicians and our clinical team, our nurse practitioners, clinical pharmacists, also supervised by medical directors, how do we help our clinicians be able to 188 00:18:35,115 --> 00:18:39,296 digest and make sense of all this data that's coming in? 189 00:18:39,296 --> 00:18:42,117 So patients have two, three devices. 190 00:18:42,277 --> 00:18:44,708 And know when is the right time? 191 00:18:44,708 --> 00:18:46,259 When is the opportune time? 192 00:18:46,259 --> 00:18:48,879 to make a medical decision. 193 00:18:49,179 --> 00:18:54,539 So we're very big right now on AI as enabling better decisions. 194 00:18:54,539 --> 00:18:58,539 We're big believers in how this technology is going to grow and be more powerful. 195 00:18:58,539 --> 00:19:02,039 But every decision at the end of the day is going to be made by a clinician. 196 00:19:02,039 --> 00:19:03,999 Every decision is going to be made by a patient. 197 00:19:03,999 --> 00:19:10,819 We're just using AI to make that process more seamless and also surface the right information at the right time. 198 00:19:10,819 --> 00:19:13,379 And that's been so effective already. 199 00:19:13,379 --> 00:19:16,082 We had testimonial from one patient who was at the 200 00:19:16,082 --> 00:19:26,305 ACCESS launch event, who defined it as, this is a new way of, this is a new healthcare system that is more proactive and responsive to my needs. 201 00:19:26,305 --> 00:19:35,908 And that really stuck with me, because if we can do this in the home environment, and we integrate it back into the broader healthcare setting, which we're doing, that means the 202 00:19:35,908 --> 00:19:39,409 whole healthcare system becomes more proactive and responsive to the needs. 203 00:19:39,409 --> 00:19:42,598 So we're taking this step by step, 204 00:19:42,598 --> 00:19:45,920 But I'm curious because you are dealing with AI all the time as well. 205 00:19:45,920 --> 00:19:53,314 I'm curious about how you're thinking about

this technology and its applicability in value-based care. 206 00:19:54,693 --> 00:19:58,104 well, thanks for inviting me into the AI discussion. 207 00:19:58,104 --> 00:19:59,635 for us, 208 00:20:00,215 --> 00:20:11,026 it's, we have a privilege of working alongside brilliant folks like yourself who are on the front lines in the same way, right? 209 00:20:11,026 --> 00:20:15,922 Like collecting the data and being there with the patient. 210 00:20:15,922 --> 00:20:23,275 And we get to see it across different specialized verticals, across several service lines. 211 00:20:23,625 --> 00:20:31,530 And we also then get to see that information and how it needs to get used within for Rainfall Health. 212 00:20:31,530 --> 00:20:34,221 We do it as episode care management. 213 00:20:34,221 --> 00:20:43,087 So that means it's always time bound in areas and windows that are most impactful to the patient journey, right? 214 00:20:43,087 --> 00:20:44,568 30 days post surgery, 215 00:20:44,568 --> 00:20:45,889 90 days post-surgery. 216 00:20:45,889 --> 00:20:50,441 There's a lot of data that shows that if we can manage patients in that time period. 217 00:20:50,475 --> 00:20:56,670 We actually really improve their long-term viability and long-term care and outcomes. 218 00:20:56,971 --> 00:21:06,638 And so for us, it's all about taking data that's coming from all these different sources and actually giving it a standard. 219 00:21:07,039 --> 00:21:18,989 And and by that, it's not dissimilar from your approach of taking data streams, but the way we define standards is what it means in terms of quality. 220 00:21:19,143 --> 00:21:25,118 for the clinicians across different geographies, across different timelines, across different episodes. 221 00:21:25,279 --> 00:21:37,339 And the AI piece for us is trying to both translate some of that stuff to the clinician so that they know the action 222 00:21:37,339 --> 00:21:38,550 steps that are needed. 223 00:21:38,550 --> 00:21:46,857 And again, we're big proponents of always keeping the patient and the clinician at the center of everything that's happening. 224 00:21:47,077 --> 00:21:57,043 And from the patient perspective, being able to meet them in terms of where they already are and translating what's happening to them. 225 00:21:57,083 --> 00:22:03,637 And by that I mean we have our AI voice caller that speaks 76 different languages that can help the 226 00:22:03,637 --> 00:22:07,309 patient just navigate what's happening to them. 227 00:22:07,309 --> 00:22:12,331 So that so they know all the possibilities of what to do. 228 00:22:12,412 --> 00:22:14,749 And that usually means that. 229 00:22:14,749 --> 00:22:22,533 Any patient might see as many as a dozen different care providers in that 30 day episode. 230 00:22:22,533 --> 00:22:31,097 But from the patient perspective, there's sort of like one advocate, one touch point for them that can help navigate through that entire piece. 231 00:22:31,097 --> 00:22:38,730 And for Rainfall, it's really about enabling other folks to shine, right? 232 00:22:38,730 --> 00:22:40,461 Whether it's hardware. 233 00:22:40,743 --> 00:22:49,446 combined with software like Withings or it's, you know, somebody just on a pure services side or even brick and mortar like skilled nursing 234 00:22:49,446 --> 00:22:50,056 facilities. 235 00:22:50,056 --> 00:23:02,299 It's really about the patient flowing through the system and making sure that they are in a great place overall and tracking it in terms of those components that I mentioned, 236 00:23:02,299 --> 00:23:07,563 which is patient satisfaction, which is just as important as quality of care. 237 00:23:07,563 --> 00:23:16,880 which is just as important as outcomes and cost management, because you can't really have all three working, you have to have all three working in unison. 238 00:23:16,880 --> 00:23:19,962 You can't really be taking one or the other out of the equation. 239 00:23:19,962 --> 00:23:23,507 so that's at least how we tend to view it. 240 00:23:24,052 --> 00:23:25,573 That's really, that's great. 241 00:23:25,573 --> 00:23:30,066 You one of the use cases I'm very obsessed with right now that 242 00:23:30,112 --> 00:23:42,027 And again, we have the benefit of having millions of consumers working with us and tremendous amount of data is we've been able to develop and we had a publication in the 243 00:23:42,027 --> 00:23:52,922 Mayo Clinic recently about predicting GLP-1 outcomes based on early, early data points that we collect with our body composition scale. 244 00:23:52,942 --> 00:23:58,461 And that's been something that we've shown to our clinicians, like, is this something of clinical value? 245 00:23:58,461 --> 00:24:11,328 And it's been incredibly instructive, not just for clinicians to be able to assess titration schedules, but also the patient feels way more at the center of what they're 246 00:24:11,328 --> 00:24:13,959 doing, which is kind of just on what Rainfall, right? 247 00:24:13,959 --> 00:24:25,165 The patient is a consumer and we've, historically, we haven't done a great job of bringing that standard experience that a consumer would expect to healthcare. 248 00:24:26,602 --> 00:24:34,602 So what we're seeing here is that now the patients are able to see, how is my weight loss going to project? 249 00:24:34,602 --> 00:24:36,722 And they're making better decisions. 250 00:24:36,722 --> 00:24:39,102 They feel more informed about their own health. 251 00:24:39,102 --> 00:24:41,942 And it's been helping us achieve better outcomes. 252 00:24:42,862 --> 00:24:44,622 So there's so much that can be done. 253 00:24:44,622 --> 00:24:49,442 And it really starts with, is this going to be part of a patient-centered goal? 254 00:24:49,862 --> 00:24:51,402 We have clinical outcomes targets. 255 00:24:51,402 --> 00:24:53,042 We have patient-centered goal targets. 256 00:24:53,042 --> 00:24:55,042 Is this

part of a patient-centered goal target? 257 00:24:55,794 --> 00:25:10,440 is something that we can safely use, which is another big part you spoke on, too you know, the, you're, dealing with, and in Rainfall, you're working with well-approved, 258 00:25:10,440 --> 00:25:14,875 well-respected medical providers and devices. 259 00:25:14,875 --> 00:25:19,453 Withings, we take a similar approach here, but on the device side, we're big on FDA clearance. 260 00:25:19,453 --> 00:25:24,509 So any data that is coming back to our clinicians, they can safely act on it. 261 00:25:24,509 --> 00:25:33,195 And I think that's another piece that we're looking at a lot right now, which is consumers are very interested in what's going on devices. 262 00:25:33,196 --> 00:25:37,000 Most of them are not FDA cleared devices. 263 00:25:37,000 --> 00:25:38,231 They're wellness devices. 264 00:25:38,231 --> 00:25:47,728 And that's a huge difference because a wellness device means, hey, and consumer or patient, this is a wellness device. 265 00:25:48,029 --> 00:25:53,169 You can do with this what you wish, but there are risk associated and this is not. 266 00:25:53,169 --> 00:25:55,289 reviewed or approved by the FDA. 267 00:25:55,570 --> 00:25:58,710 And the patient has a very limited toolbox of what they can do. 268 00:25:58,710 --> 00:25:59,861 They can sleep better. 269 00:25:59,861 --> 00:26:02,991 They can maybe change their activity. 270 00:26:03,091 --> 00:26:07,152 When an FDA clearance comes in, now the toolbox expands. 271 00:26:07,352 --> 00:26:10,593 And it's no longer, hey, patient, here's what you can do. 272 00:26:10,593 --> 00:26:13,574 Now it's, hey, patient, here's what we can do together. 273 00:26:13,574 --> 00:26:17,295 We can put you on a second line therapy for your hypertension. 274 00:26:17,295 --> 00:26:20,385 We can titrate that medication based on this data point. 275 00:26:20,385 --> 00:26:22,854 So we're thinking a lot about this too. 276 00:26:22,854 --> 00:26:33,410 which is the AI is fantastic, what data is going into the AI so that when the actions are occurring, we have the biggest toolbox possible in terms of actions taken. 277 00:26:34,304 --> 00:26:45,117 Wow, that's such a differentiated approach in many ways where it you know, from where we are standing and we spend a lot of time on these standards and you know, folks that want 278 00:26:45,117 --> 00:26:50,108 to raise their hand and say, We're gonna go above and beyond the minimum requirements. 279 00:26:50,108 --> 00:26:50,758 And that's what 280 00:26:50,758 --> 00:26:59,451 I'm hearing that Withings is trying to do and is doing successfully, which is going to FDA when you don't necessarily need to, right? 281 00:26:59,451 --> 00:27:04,172 You'll still have the consumers, you'll still be able to get them 282 00:27:04,255 --> 00:27:08,247 even if it's a limited set of services, but you're going to FDA. 283 00:27:08,247 --> 00:27:20,004 You're raising your hand and saying, please judge us as a third party, do that audit, because we feel so comfortable that we're willing to put our devices and our technology 284 00:27:20,004 --> 00:27:22,125 for scrutiny. 285 00:27:22,266 --> 00:27:28,569 And that's such an amazing way to gain trust and also showcase why 286 00:27:28,569 --> 00:27:31,411 there's a differentiated space that you all can play in. 287 00:27:31,411 --> 00:27:33,160 And then that is 288 00:27:33,160 --> 00:27:36,803 truly what leads to being able to provide that better experience. 289 00:27:36,803 --> 00:27:41,276 Is that kind of how you all tend to also look at the strategy, which is like, hey, look, we think we 290 00:27:41,276 --> 00:27:45,174 are delivering a better product and we wanna be able to showcase it. 291 00:27:45,645 --> 00:27:46,446 Absolutely. 292 00:27:46,446 --> 00:27:57,152 We've been, there's been this gray line, right, between the consumer and the patient that's been, you know, as we've all been playing and understanding how do we make this a 293 00:27:57,152 --> 00:27:58,192 strong 294 00:27:58,192 --> 00:28:00,692 integration between the consumer and the patient. 295 00:28:00,912 --> 00:28:10,172 And our belief is one of the big things is the data that is used by a consumer needs to also be able to be used by a clinician. 296 00:28:10,172 --> 00:28:16,132 And the way to do that is through FDA approval, FDA review. 297 00:28:16,372 --> 00:28:18,752 That's a huge piece of what we think about. 298 00:28:18,752 --> 00:28:22,512 It also we have the benefit of thinking long term. 299 00:28:22,512 --> 00:28:26,905 So Withings is a large, but we're still founder-owned. 300 00:28:26,905 --> 00:28:37,332 company, we're a profitable company, we have a long term vision that we're trying to execute on and we don't get distracted by the short term possible wins. 301 00:28:37,332 --> 00:28:48,460 We're thinking long term about how does an individual be able to prevent and manage chronic conditions at home and our belief at the end of the day is enabling that data 302 00:28:48,460 --> 00:28:55,894 point to be used by the person at home and by their clinician with the biggest set of toolbox possible so they can 303 00:28:55,894 --> 00:28:57,115 manage their needs. 304 00:28:57,115 --> 00:28:58,716 we're always going to go above and beyond. 305 00:28:58,716 --> 00:29:01,959 We're going above and beyond by setting up our own medical group. 306 00:29:01,959 --> 00:29:13,718 We're going above and beyond because we think it's common sense that there should be a continuity of care that goes from the in-person to the virtual into your home, even into 307 00:29:13,718 --> 00:29:17,632 your bathroom scale that you stand on every morning. 308 00:29:17,632 --> 00:29:21,054 That's our core belief, and we're going to be focusing on that. 309 00:29:21,685 --> 00:29:22,756 That's so amazing. 310 00:29:22,756 --> 00:29:30,542 And, again, a differentiator approach,

which does again come back to the core foundings and how you all are set up. 311 00:29:30,542 --> 00:29:39,547 Now I'm going to switch gears just a little bit away from the patient because even though that's such an important tenet of all care that we provide, to the clinician, who's 312 00:29:39,547 --> 00:29:40,227 another, 313 00:29:40,227 --> 00:29:42,569 huge component of this. 314 00:29:42,569 --> 00:29:47,452 You mentioned you're working with multiple health systems, including MedStar. 315 00:29:47,753 --> 00:29:51,755 Who do you engage with at the system level? 316 00:29:51,775 --> 00:30:03,453 how because again, a lot of our audience are health system executives, and who are the right folks that you would engage with that you get to showcase both the power of what 317 00:30:03,453 --> 00:30:11,615 you're building and providing to the patient and also how you can probably make their lives easier and deliver better care, right? 318 00:30:11,615 --> 00:30:21,119 Like that's always how we have these amazing executives who all care and want the same goal, but sometimes 319 00:30:21,119 --> 00:30:23,184 You know, it's just so complex. 320 00:30:23,184 --> 00:30:31,122 So how do you how would you break it down on how you approach these different systems and what does the value provide and to which leaders? 321 00:30:31,122 --> 00:30:42,531 we have fantastic relationships with our health system organizations working at really two levels, the population health level as well as the virtual care and digital health level. 322 00:30:42,531 --> 00:30:46,174 And we work with our partners in two ways. 323 00:30:46,174 --> 00:30:58,703 One way is, and like you mentioned, the 90 day episodes and the readmission windows, we work with our partners a lot on heart failure, readmissions work. 324 00:30:58,703 --> 00:31:12,235 So, and one of the key reasons why we have a number of heart failure readmission programs where we provide technology to our health system partners to be able to have reductions 325 00:31:12,235 --> 00:31:16,458 of readmission rates is we do a couple things really well. 326 00:31:16,679 --> 00:31:20,423 One is we integrate data back very cleanly. 327 00:31:20,423 --> 00:31:25,847 Two is we've innovated device hardware and firmware. 328 00:31:26,001 --> 00:31:27,532 So we identified pain points. 329 00:31:27,532 --> 00:31:39,116 For example, in heart failure management, it's part of evidence-based guidelines to collect not just weight, but also, and blood pressure, but also symptom surveys. 330 00:31:39,376 --> 00:31:50,780 And what we saw in the market with our partners was through their existing portals, they would get 30, 35% adherence from their patients to these symptom surveys. 331 00:31:50,780 --> 00:31:55,122 And that lack of adherence means there's less visibility into the patient health, which means 332 00:31:55,174 --> 00:31:56,554 higher readmission rates. 333 00:31:56,554 --> 00:32:05,917 What we've done, for example, is that we've designed our blood pressure monitor to be able to, on the blood pressure monitor, believe it not, collect symptom surveys. 334 00:32:05,917 --> 00:32:10,678 The result of that is that we have an 89% adherence to symptom surveys. 335 00:32:10,678 --> 00:32:17,080 So we've been able to design hardware and firmware to meet the needs of specific use cases. 336 00:32:17,080 --> 00:32:21,181 Heart failure is one of them, which is really big with population and virtual care teams. 337 00:32:21,181 --> 00:32:23,622 The other one is hypertension management. 338 00:32:24,153 --> 00:32:37,664 For patients who are higher acuity hypertensive and polychronic and would benefit from more integration with the healthcare system, we have on that blood pressure monitor 339 00:32:37,664 --> 00:32:43,559 the ability to better educate on how to take a blood pressure reading. 340 00:32:43,559 --> 00:32:50,434 Also on the blood pressure monitor, we have the opportunity for patients to be able to do automated retake. 341 00:32:50,434 --> 00:32:52,836 And these features are really interesting, Eddie, right? 342 00:32:53,092 --> 00:33:00,792 When we have our fee for service organizations who speak to us, they say, wow, this is so efficient. 343 00:33:00,852 --> 00:33:05,832 That's horrible, because it's going to reduce our activity billing because we're making fewer phone calls. 344 00:33:05,832 --> 00:33:14,072 When we have value based organizations who look at this feature, they say, this is amazing because I'm going to save my clinicians time calling for retake. 345 00:33:14,072 --> 00:33:17,252 So we're thinking about these small pain points to make. 346 00:33:17,272 --> 00:33:23,237 virtual management more efficient and effective and that's been widely popular with our health system partners. 347 00:33:23,237 --> 00:33:36,367 Now that we are doing lower acuity care ourselves, now we're working a lot more with population health and virtual care, but through the way of how do we identify who's eligible and 348 00:33:36,367 --> 00:33:39,429 who's really going to benefit from this type of program. 349 00:33:39,490 --> 00:33:44,264 But also for patients that we enroll that may not have known the system before, right? 350 00:33:44,264 --> 00:33:47,446 Because we have 5 million users around the country. 351 00:33:47,544 --> 00:33:49,905 How do we get them back into the health system? 352 00:33:49,905 --> 00:33:53,577 If they need an echo, can we do a direct referral back? 353 00:33:53,577 --> 00:33:56,668 If they need a cardiology visit, can we do a direct referral back? 354 00:33:56,668 --> 00:34:06,872 So it's been a very exciting scope of partnerships and how we work with population health and virtual care leaders. 355 00:34:07,773 --> 00:34:08,823 That's amazing. 356 00:34:08,823 --> 00:34:19,456 And, you know, again, really speaks to when you can align incentives from this sort of policy lens, it really does drive

the right outcomes 357 00:34:19,456 --> 00:34:29,899 and the right alignment across the entire network from patients all the way to the payers, providers, clinicians, and everybody along the way to help make that happen. 358 00:34:29,899 --> 00:34:37,245 now, you know, as we start to think about all the data. 359 00:34:37,245 --> 00:34:41,818 that's being created and what are the insights that you're most excited about? 360 00:34:41,818 --> 00:34:52,833 What's, you know, that they'll help unlock now that we are able to start you know, getting closer and closer the patient in the moment that we we have some sort of insight 361 00:34:52,833 --> 00:34:58,787 to actually the care delivery, which some it sounds like some sometimes Withings can provide. 362 00:34:58,787 --> 00:35:05,250 And then also going all the way to, you know, a specialized higher acuity care need as well. 363 00:35:05,250 --> 00:35:07,121 What do you think it's going to unlock? 364 00:35:08,788 --> 00:35:11,108 It's a fantastic question. 365 00:35:11,349 --> 00:35:21,441 know, from at the end of the day, time to control for most of this population is, think, all of our guiding principle. 366 00:35:21,441 --> 00:35:32,534 Now, readmissions are a very important metric, but it's a downstream effect of getting a heart failure patient under control and stabilized. 367 00:35:32,594 --> 00:35:37,636 So we're always thinking about that component to it. 368 00:35:37,636 --> 00:35:38,936 What we see 369 00:35:39,128 --> 00:35:51,562 is when the patient is engaged in their own care, and our belief at Withings is that the device is a core piece of that care that they get, and that data is helping inform 370 00:35:51,562 --> 00:35:58,053 clinician decisions, we can accelerate time to control tremendously. 371 00:35:58,614 --> 00:36:07,606 The other piece of this, is as patients become more consumers and are thinking about 372 00:36:07,606 --> 00:36:09,837 What experience of care do they want? 373 00:36:10,117 --> 00:36:15,620 We're thinking a lot as well about patient satisfaction as being one of the core pillars. 374 00:36:15,620 --> 00:36:20,463 Not just getting patients under control, but were they happy with the experience? 375 00:36:20,463 --> 00:36:25,145 Is that an experience that they are choosing with their feet, right? 376 00:36:25,145 --> 00:36:27,507 And to join. 377 00:36:27,507 --> 00:36:33,510 We waive co-pays and our partners often waive co-pays on things like this as well. 378 00:36:33,510 --> 00:36:37,872 But NPS is becoming more more important. 379 00:36:37,995 --> 00:36:43,636 One thing is our approach to that is we design a premium experience with our devices. 380 00:36:43,636 --> 00:36:53,969 I don't know, Eddie, if you have, but it's a premium experience that we make available to all income brackets, all lines of business. 381 00:36:53,969 --> 00:37:07,922 So much like we have our MedStar partnership in concierge medicine where patients in those

programs have resources to engage in higher touch care at MedStar. 382 00:37:07,934 --> 00:37:15,856 We also have FQHCs using the same technology for the same use cases, right, in the Medicaid population. 383 00:37:16,016 --> 00:37:17,967 And that's been our guiding principle, right? 384 00:37:17,967 --> 00:37:30,301 So great data, FDA-cleared data to the patient, to the clinician, and an experience that the patient is going to want to engage in, regardless of means to pay, is what we are 385 00:37:30,301 --> 00:37:31,667 always thinking about here. 386 00:37:31,667 --> 00:37:35,002 And we've seen that with high activation rates. 387 00:37:35,002 --> 00:37:35,848 So... 388 00:37:35,848 --> 00:37:44,563 When a patient signs up, historically in my life, I've seen a number of programs where even when they want to sign up, they don't start the program because once they sign up, 389 00:37:44,563 --> 00:37:47,084 they're kind of like, what am I getting myself into? 390 00:37:47,084 --> 00:37:55,869 At Withings, 94% of those who enroll in our partners programs, hypertension and heart failure, start the program. 391 00:37:55,869 --> 00:37:57,290 They are collecting data. 392 00:37:57,290 --> 00:38:05,094 92% engage month over month in a clinically meaningful way. 393 00:38:05,128 --> 00:38:17,088 and over 90% still use their device at the end of 12 months, that enables a much more effective care model that the patient is gonna come back to the system and is gonna come 394 00:38:17,088 --> 00:38:23,228 back and feel like they're being heard and that they're gonna want more support, more services from that system. 395 00:38:23,228 --> 00:38:32,988 So that's how we think about it a lot, Eddie, is get the outcomes but do it in a way that the patient experience is also top of mind. 396 00:38:33,990 --> 00:38:34,940 No, that's amazing. 397 00:38:34,940 --> 00:38:39,352 And such an again, an amazing way to look at the entire experience. 398 00:38:39,352 --> 00:38:47,754 It's if you listen to any of the other episodes of our podcast, I really talk a lot about my personal mission around access to care. 399 00:38:47,754 --> 00:38:59,567 And I think that's fundamentally what you all are driving towards with experience also still front of mind, but access so that we can make it more available to the people who 400 00:38:59,567 --> 00:39:00,938 need it the most as well. 401 00:39:00,938 --> 00:39:02,650 that's 402 00:39:02,650 --> 00:39:03,401 so powerful. 403 00:39:03,401 --> 00:39:06,255 I it's it making me feel like I'm missing out on something. 404 00:39:06,255 --> 00:39:12,642 I'm gonna I'm gonna have to go to Withings and try to see what what I can get for myself and find out more about my own data. 405 00:39:12,703 --> 00:39:14,216 maybe we'll 406 00:39:14,216 --> 00:39:14,958 going to be possible. 407 00:39:14,958 --> 00:39:24,515 The more and more payers are going to start covering for these types of ACCESS services, we'll all be eligible for these, not just original Medicare. 408 00:39:24,515 --> 00:39:31,798 And I think the difference between remote monitoring with health systems and this ACCESS model is worth speaking about too. 409 00:39:31,798 --> 00:39:36,180 So I think it touches on what is 410 00:39:36,819 --> 00:39:42,963 how integrated or sorry, how tied to the hospital should this patient be? 411 00:39:43,424 --> 00:39:51,148 For a patient that's coming out, they have multiple hospitalizations, multiple ED visits, remote monitoring is gonna be the best way to do this. 412 00:39:51,148 --> 00:39:57,832 We want really small patient panels, very closely tied to that discharging hospital. 413 00:39:58,093 --> 00:40:05,957 And for patients who may not have an ED visit or an acute event, but like many of us, 414 00:40:06,504 --> 00:40:21,317 going down the risk spectrum from obesity into hypertension to diabetes, ACCESS becomes a fantastic service to have more preventive healthcare at a very low cost structure that is 415 00:40:21,317 --> 00:40:22,658 no copay to patients. 416 00:40:22,658 --> 00:40:34,638 So I think also the type of access to care is being kind more formalized, which has been very, very clarifying on the Withings side. 417 00:40:35,637 --> 00:40:36,958 No, that's so exciting. 418 00:40:36,958 --> 00:40:40,030 And I I'm excited to see where it keeps going. 419 00:40:40,030 --> 00:40:50,096 And it's it's great to see these enablement models instead of, you know, the responses that have come out in the past, which is not yet another model. 420 00:40:50,257 --> 00:40:50,947 so it's great 421 00:40:50,947 --> 00:40:56,519 the ones that actually get to the heart of what we're trying to achieve have from that like larger vision. 422 00:40:56,519 --> 00:41:04,764 as we start to think about, you know, wrapping up this incredibly exciting podcast, and I could keep talking about this for hours. 423 00:41:04,796 --> 00:41:10,384 what are you most optimistic about in the next three to five years? 424 00:41:12,708 --> 00:41:13,828 I 425 00:41:14,969 --> 00:41:28,737 think we're on a path right now where technology-based companies like Withings, who are both entering healthcare or enabling, and Withings we do both, 426 00:41:28,737 --> 00:41:39,630 We enable through remote monitoring, we go directly to patients for lower acuity, are now having a more meaningful seat at the table of value-based care. 427 00:41:39,630 --> 00:41:52,614 whereas I felt a stagnation in this transformation that we've been part of since the mid 2010s, now I see this reinvigoration of the value-based care movement 428 00:41:52,614 --> 00:41:56,805 because technology companies are having a seat at the table. 429 00:41:58,139 --> 00:41:59,690 not overlapping. 430 00:41:59,690 --> 00:42:09,137 So we are not, we are focused on the outcomes and we partner with our organizations who are focused on the cost of care, quality measures, utilization based. 431 00:42:09,137 -->

00:42:13,359 But because we have our own seat at the table now, so much more can be done. 432 00:42:13,359 --> 00:42:25,197 And what that so much more can be done, I think of things like, you know, our ability to predict GLP-1 outcomes, our ability to predict heart failure congestion, it allows us to 433 00:42:25,197 --> 00:42:27,829 really invest a lot more into 434 00:42:27,837 --> 00:42:38,485 not just the individual biomarkers, but that overall AI sensing layer about what do these FDA approved markers mean, so we can be more proactive in care. 435 00:42:38,485 --> 00:42:45,740 And that's really enabled because we have a seat at the table now, because now we can safely invest into these technologies. 436 00:42:45,740 --> 00:42:54,316 And we've seen it already in the past nine months to a year, what we can build once there's clarity about our role in the value-based care movement. 437 00:42:54,316 --> 00:42:56,817 So that's been very exciting. 438 00:42:57,116 --> 00:43:08,681 I don't want to say what's going to happen in three to five years, but I think there are ways for us to be able to really bend the cost curve on chronic conditions. 439 00:43:08,681 --> 00:43:17,395 And I think there are ways that we can get everyone engaged in some preventive healthcare, most notably hypertension. 440 00:43:17,395 --> 00:43:22,003 I think in three to five years, the idea of having, you 441 00:43:22,003 --> 00:43:33,183 Over 40, 50% of hypertensive uncontrolled, I think it's gonna look a little bit crazy in three to five years once people all have the technology to monitor and the access to care 442 00:43:33,183 --> 00:43:36,263 to properly treat a condition like hypertension. 443 00:43:37,648 --> 00:43:39,772 I feel like that's the future I can get behind. 444 00:43:40,837 --> 00:43:46,882 Patrick, what have we not talked about that you won't that we want to cover before we wrap up today? 445 00:43:48,338 --> 00:43:50,979 I think this was an amazing conversation, Eddie. 446 00:43:50,979 --> 00:43:53,119 It's refreshing to speak with you. 447 00:43:53,119 --> 00:43:54,859 It's great to see you again. 448 00:43:55,399 --> 00:43:59,459 I would love to hear from your perspective, right? 449 00:43:59,459 --> 00:44:02,939 Because you deal in lot of episode-based models. 450 00:44:02,939 --> 00:44:06,819 One of the things about ACCESS is it's a longitudinal-based model. 451 00:44:07,279 --> 00:44:16,891 Maybe we can talk about this over coffee another time, but the way the overlap of episode-based models and longitudinal models and... 452 00:44:16,891 --> 00:44:23,609 I'd love to hear more from you about what you think about that, but that could be a separate conversation altogether. 453 00:44:24,176 --> 00:44:26,156 Well, it's good for us to tease it here. 454 00:44:26,156 --> 00:44:38,924 I think just on a high level, I just wanna leave that's a good place to leave everybody, which is I think there is so much space and so much white space actually for us to build 455 00:44:38,924 --> 00:44:45,067 on in both areas and drive better value, both in the short term where we see 456 00:44:45,196 --> 00:44:54,160 an acute event that's taking place and how to manage that versus really how do you manage a better life. 457 00:44:54,160 --> 00:44:55,121 as a whole. 458 00:44:55,121 --> 00:45:01,708 And I think both of those have the same goal, but they're just driving at different 459 00:45:01,708 --> 00:45:02,267 stages. 460 00:45:02,267 --> 00:45:06,150 And I think the value is spread out over time. 461 00:45:06,150 --> 00:45:20,002 It's it reminds me of when there was a big literature finding back again in the early 2010s that said only about 30% of your care and your outcomes are dependent 462 00:45:20,002 --> 00:45:22,424 on clinical information. 463 00:45:22,446 --> 00:45:23,877 The rest of it happens at the home. 464 00:45:23,877 --> 00:45:25,288 It happens in social settings. 465 00:45:25,288 --> 00:45:27,180 It happens in everything else. 466 00:45:27,360 --> 00:45:33,866 And what was always amazing to me about that is we only capture that 30% on the clinical side. 467 00:45:33,866 --> 00:45:45,386 And now with technologies that are coming out, I think we have that ability to stretch that into that remaining 70% and strive for that overarching goal that we may never get to 468 00:45:45,386 --> 00:45:46,307 of a hundred percent. 469 00:45:46,307 --> 00:45:47,678 But hey, that's that's 470 00:45:47,678 --> 00:45:50,350 the frontier that we keep striving towards. 471 00:45:51,458 --> 00:45:52,058 I love it. 472 00:45:52,058 --> 00:46:00,041 Yeah, I think the if I think about ACCESS V2 right in 2027, 2028, I think this 473 00:46:00,047 --> 00:46:06,927 know, this integration between episode-based models and this longitudinal model is going to be really exciting. 474 00:46:07,087 --> 00:46:11,807 And what does it mean for a transition of care back into the home environment? 475 00:46:12,247 --> 00:46:20,227 I think it's a point that we're going to be thinking about, and it's going to be a really exciting time on this. 476 00:46:20,227 --> 00:46:23,307 So thank you very much, Eddie, for having me on here. 477 00:46:23,307 --> 00:46:28,927 And I look forward to diving deeper with you over coffee another time as well. 478 00:46:29,530 --> 00:46:31,659 Patrick, thanks so much for coming on.