Fresenius Medical Care AG (FME) Earnings Call Transcript & Summary

Ladies and gentlemen, thank you for standing by. I'm Haley, your Chorus Call operator. Welcome and thank you for joining the Fresenius Medical Care expert call on mapping a digital future for kidney care. [Operator Instructions] And I would now like to turn the conference over to Dominik, Head of Investor Relations. Please go ahead.

Thank you, Haley. We would like to welcome all of you to the Fresenius Medical Care expert call series 2021, in which we will address 3 different topics over the course of this year. The opening event is about a fascinating future and is titled Mapping a Digital Future for Kidney Care. For those who are joining via telephone today, you can find the presentation slides on the events on our website under Investor Relations, Events & Presentations, Meet the Management. As always, I'm happy to start out the event by mentioning our cautionary language that is in our safe harbor statement on Slide 2 of the presentation. For further details concerning risks and uncertainties, please refer to this document as well as to our SEC filings. Digitization is playing already today a very important role in our business, and this will play an even bigger role going forward. It is fascinating how we can already improve the lives of our patients, thanks to our data analytics. I'm outstandingly happy that I was able to win Dr. Frank Maddux, our Global Chief Medical Officer, to give you some deeper insights today into how this area will further transform the future of kidney care. Please be aware that this call as well as the following Q&A session will not cover the financial background, related costs or revenue questions in this respect. Furthermore, I would like to ask for your understanding that we will not answer any questions about the current COVID-19 situation, such as the vaccination rate, incident rates or excess mortality. We will give an update here in the context of our Q1 earnings release on May 6. We really like to focus in this 1 hour only on the topic of digital future for kidney care and use the precious time of Dr. Frank Maddux for this. I will now hand over to Frank. The floor is yours.

Thank you, Dominik, and I appreciate the chance to give you all my perspective on the topic of digitalization. If we can move to Slide 3, Dominik. Across the strategic vision we laid out in the October 2020 at Capital Markets Day, we see influence of digitalization in every aspect of our strategy. Kidney disease care along the renal care continuum is highly dependent on digital influence as it has a quantitative basis as -- at its heart and a need for frequent contact with patients. Insights from longitudinal data and new techniques that we discuss become a catalyst for the evolution of the care model and change that we expect in kidney care and in the marketplace today. The critical care pillar has a need for detailed real-time diagnostics and therapeutic monitoring of acute illness and how to assess and protect organ systems and physiologic processes. And finally, our complementary asset pillar has a strong basis for digitalization as we move toward precision medicine, genomics and engineering of human systems that support chronic disease care. Moving on to Slide 4. Fresenius Medical Care has excelled in scaling access to care as the field has evolved. Using our vertical integration and breadth of our company provides a means to introduce scale into digital innovation, which heretofore has been somewhat of a novelty. We see the approach similar to the introduction of renal replacement therapy, where we set aspirations in markets still evolving and maturing. But in more mature markets, we're catalyzing alternatives that bring power and choice to patients and their caregivers in where and how they receive care using these digital techniques to support a new way of engaging patients and their caregivers. Engaging people where they are in the journey of a life with kidney disease and making choices for care broader than what we've seen in the past 5 to 6 decades of kidney care is one of the objectives here. To do this, we must expand the interest and investment in the space to create more seamless and connected environment in which patients partner with providers for medical, social and behavioral support in an agnostic site of care setting. Moving on to the next slide. Today, I want to share with you sort of a broad view of the digital impact on the core of what we offer in both care to patients, logistics of operations, communications and influence on good health decisions by patients and their providers, but doing this at scale. Not only can we create a more connected and reliable supply chain for products and services, utilizing both machine, device and therapy data and include the recognition of digital techniques to communicate our messaging in a highly engaged world that includes many channels of messaging to patients, physician partners and the uses and power of social media and rapid discussion, dialogue and dissemination of information among patients, families and physicians. The science of kidney care is changing and evolving. We aim to bring the best of these changes to the market quicker than might naturally occur and do it at scale. Moving on to Page 6. Sources of information are wide and varied. They include these 6 frames that provide a framework to discuss the approaches we've taken, incorporating digital data into our decision-making, field influence and direction of how medicine is evolving. The incorporation of data from health records, clinical care, machine and environmental data, along with what is captured from a person's daily life can be incorporated with genetic understanding of how an individual is programmed to be susceptible to disease or might respond to interventions. Kidney disease has progressed from a uniformly fatal condition to a routine treatment with many challenges for patients, but a guarantee that you won't die from your kidney failure alone. The first half century of renal replacement therapy was to assure that all people in need had access to care. Now we are moving to a world that aims to optimize that care and the power and choice that patients have in living with this disease. Much of this is the reason that we are so committed to building out the infrastructure for home dialysis and broader choices of care for our patients, requiring that they live in a connected health environment. Moving on to Slide 7. We divide our data and digital maturity into this evolutionary framework. Analytical data is for a utility and not just to provide access to smarter techniques of what we already do. We aim to make decision-making more reliable, insightful and specific to the individual in front of us. The adage you all have heard is the aim here. The right patient gets the right treatment at the right time while enhancing the information that providers have to make more informed decisions with their patients. These insights are unavailable without the advanced digital and telecommunications frameworks that have matured and been enhanced during the past year of the COVID pandemic. We've expanded dramatically the use of telehealth as an example of the manyfold increases in the use of digitalization and the practice of kidney care during this lockdown period. Moving on to Slide 8. Well, what's required to do this work and assure a practical application mindset? It takes particular skill sets and capabilities, which we share and have germinated across the company. These skills have come in a variety of collaborating experts that don't always have a background in health care but understand techniques that we need to use to both interrogate the data that we have access to and to make it useful. The need to collaborate and coordinate across these and other disciplines and look at the types of relationships and data sources that may seem odd are necessary to unlock some of the insights previously referenced. Let's get to some details and move on to the next slide to talk about one of the things that most practically actually begins to bring this to focus. Let's talk about the #1 associated comorbidity in our clinical and quality agenda. That's cardiovascular health and cardio protection. 80% of our patients have a defined cardiac diagnosis. 51% of the deaths in our patients, except for this past year related to the pandemic, have been patients with advanced kidney disease dying from a cardiovascular cause. 80% of these causes of death are from rhythm disturbances, and the type of cardiovascular disease in patients is different depending upon the stage of the disease. We know more patients die from cardiovascular disease before reaching end-stage kidney disease than ever get to the need for any renal replacement therapy or transplant. Digital care models of the future will demand enhanced observation of heart rhythms, fluid volumes, stress on the delicate muscles and blood vessels that feed the kidneys. The upcoming medications for CKD may well keep more people alive from their cardiovascular disease to continue to enhance the [ popula of ] patients in need for advanced therapies, simply because they live longer. The value of the electronic health record and its evolution is critical in actually capturing this information about this comorbidity that is so important in the population that we care for. Moving on to Slide 10. Where does the source of the frame of our data begin? It's in the data we collect to provide a phenotypic picture of the population of patients we care for. These are large numbers and represent the broadest data set in the world on advanced kidney disease. I've shown you this slide over the years, and you can see that it is growing quickly to scale with representations of the data from across the globe. Let's move on to a couple of examples and move to Slide 11. Let's show you a few examples of the work that has led to our digital framework of clinical care. This source data is expanding, and that expansion is being driven by the passive collection of diagnostic, environmental and operational data to create a more reliable system of working with patients, not requiring them to be physically in a traditional health care facility. These systems of care are moving our traditional health settings and extending them to include the time between treatments as well as the time during an active treatment. The Kinexus system, as described here in this graphic, recognizes that we're taking patient data, both from machines that we're using for therapeutic and diagnostic inputs from passive devices in the environment that the patient lives in and from an Internet of Things that a patient may be having with them in their daily life and managing that through a connected gateway into the cloud and then used for both clinical care, operational efficiency and the ability to expand our clinical information systems to where we can use that data to provide advanced trending and other such things. Moving on to Page 12. Let's talk a little bit about the digital genomics registry that we've announced recently. Through Frenova Renal Research, our clinical research arm of the company, we're building an asset that has the capability to catalyze and fuel better, faster understanding of what a person with kidney disease is living with and how the increased investment in nephrology and kidney care will begin to bring us closer to those areas of medicine that have utilized genomic research to their advantage, such as oncology, cardiovascular disease, skin disease and others. Our goal is to build the world's largest registry of curated clinical data, pair it with genomic data, and use this living engine as a way to sustain a strategy that will allow us to recognize that the way we classify and look at kidney disease is evolving and changing from a mere pathologic classification of disease by what we see under the pathologist microscope to a much greater understanding of the mechanisms of disease that drive change. Moving on to Slide 13. The genomics value creation comes from the fact that we'll spur greater investment and research in a variety of areas: that's pharmaceutical R&D research on what new therapeutic drugs and targets will actually change the nature and course of kidney disease illness; academic underlying research in the early diagnosis and monitoring of biomarkers that, in fact, recognize that certain pathways are coming together to cause CKD and progression of kidney disease and looking at ways that, that can be mitigated; and finally, to recognize in patient care, how do we recognize getting that right patient paired with the right therapy and other solutions to try to refine the models of therapy that we have for patient care in this area. This will change how we address economic outcomes and how we, in fact, are looking to actually catalyze more investment in the space. In our value-based care models, there'll be much greater insights for us and other providers to control clinical and economic outcomes, and you'll hear over the coming months the fact that we are initiating a campaign that we call My Reason that will engage patients and research and the gift that they give to future patients with kidney-related challenges to help solve these. Let's move on to Page 14. When we look at renal disease, we fall into the funding bucket of genitourinary diseases. And as you can see, there is a need for greater investment in kidney-related research. 2.6% of research dollars go to genitourinary-related symptoms in the United States. And in fact, as we increase the awareness of kidney disease, recognizing that the spending on kidney disease far outweighs the degree of investment in clinical research, we look to try to adjust this imbalance by fostering an environment where, again, the phenotypic data that we're collecting about kidney disease and the genotypic data that we look to build within this registry will create an environment where there is a much larger precision medicine opportunity in developing approaches towards managing patients with kidney disease and understanding the illnesses. This is all a digital framework because all of this data comes to us in a form that requires advanced analytical knowledge. If we move on to the next slide. Our advanced analytical methods that are destined for use in patients help identify the right patient for a particular intervention. Our goal, and this is especially important to us that in value-based care environments and in those wanting to pay for the right services for the right patients, we've got to create a very practical application of precision and personalized medicine in real life for patients, whether it's in their ability to have certain events predicted to aid in their understanding of disease and what treatments an individual needs, how to recognize the details of what the diagnostic inputs are or whether, in fact, we're modeling human processes to try to test certain interventions, whether medications are tailored to reduce progression of kidney disease, optimize comorbid conditions for treatment or pair patients with transplants that are more likely to provide a long-term solution for patients. These are all actually part of the digital framework that is advancing. If we move on to Slide 16. With new information sources, we have new ways of interrogating the data and creating a better understanding of the care that's to be delivered. Connecting devices in the environment provide us the chance to use these advanced methods of modeling, and we create these environments that are both efficient and personal to the patient. You've heard in prior years our discussions and developments of patient avatars that are used to model clinical care options and that utilize some of these techniques of machine learning, deep learning and artificial intelligence. If we move on to the next slide. Recall several years ago, I explained to you that we were using virtual clinical trials to give us the chance to, in fact, begin to adjust how we manage anemia of chronic disease that's associated with chronic kidney disease. This is an example of digital utility where we can run multiple virtual trials to actually create an environment that changes the way we actually do clinical practice today. If we go on to Slide 18. The simulation of response to erythropoietic-stimulating agents is one of the examples that we've used before in this area where we looked at virtual clinical trials in over 6,700 avatars that we produced, allowed us to model the most efficient use of long-acting ESAs. And this was highly correlated to the clinical outcomes that we would see and that we've been able to improve that showed a remarkable, more efficient use of this particular drug class. As new drug classes hit the market, we'll use similar techniques to know which patients need the drug and what pattern of use provides whatever excess benefit that class has to offer. Moving on to Slide 19. You'll see some new areas where we digitally looked at similar activities. Another example is using imaging data to identify people with dangerous vascular access issues like the development of aneurysms that need immediate attention. These tools identify and risk stratify a problem and highlight who needs that intervention. We started with this project by actually uploading and managing hundreds of thousands of digital images of vascular accesses over time in patients that we saw to then look at the trends of their ability to age and what we saw in the aging of these. These types of passive identification of risk guides us in a value-based care environment. No other company has advanced these types of techniques as far as we have, and they'll change the nature of the way we deliver care. Many companies are testing good ideas of how to connect and engage patients. That is but one part of the story of how a digital framework can enhance care. And the tools for risk identification and response becomes part of the care paradigm. That was all based on the machine learning and image recognition that came from this particular example described in the graphic above. I'll finally end on the next slide, on Page 20, by saying we've had remarkable influences from COVID, but we've used our advanced analytical methods to take over 80 elements in these 5 categories that you see here and look at predicting which patients are at higher risk of becoming persons under investigation and high-risk individuals for COVID to change the way we isolated patients and the way we achieved a variety of safety techniques to try to improve their care throughout the pandemic. In summary, digitalization is a broad concept in which many players will see a glimpse of the enormous potential that it brings to health care. Fresenius Medical Care is compelled to enter this arena so the digital techniques that we have pioneered become standards of care and that the full spectrum of digital benefit is achieved, whether it's to do clinical, operational or risk-related health care business environments and that are addressed at scale so that all patients have access to these enhanced techniques and tools that we engage patients where they are. Dominik, I'll stop there, and I'm happy to take questions related to our approach towards digital framework for the future.

Thank you, Frank. Great presentation, and I will hand over to Haley to open the lines.

[Operator Instructions] And the first question comes from the line of Lisa Clive of Bernstein.

I'm just trying to think of using technology better, being able to stratify patients by risk. Altogether, obviously, the point is to try and reduce hospitalizations, try and reduce mortality. Can you just give us some ideas of sort of what you think this can achieve? I mean I know over the last, say, 10 years, you've managed to improve mortality quite significantly. But there's a big difference between your sort of well-established patients and those patients that sort of crash into dialysis, right? And I think you've said that, that's still about 50% of your patients. Can you give us an idea today of like of those 50% of patients that crash into dialysis, what proportion die within the first 6 months? What proportion die within the first 12 months? What proportion die within 2 years? And if you can really manage them in a much better way using these technologies, kind of how much can that improve?

Yes. So that's a really broad question, Lisa, and I appreciate it. I think let's start with one side of this. One is, I think, remember, we went decades ago from uniformly fatal disease down to this sort of mid-teen area that we live in right now. The [ Healthy 2030 ] program and aspirations are to take this under 10% over the next few years. That's going to require that we fundamentally think about how we approach the patients regardless of how they come to us slightly differently. So the patients that crash into dialysis are going to require more intensive stratification of what was the cause that tipped them over? Was it that they didn't have access to the care that they would have had to appropriately prepare them? Or did they have an overwhelming event that led to their kidney failure? In either case of these, we're going to, number one, want to very quickly know how to get them into a course of therapy that gives them a higher level of engagement, where they're participating and trying to create a solution for this situation that they've unexpectedly fallen into, whether it was event related or it was simply that they didn't access the care system in as timely manner previously as they could have. All of those are substantially going to encourage us to try to be more connected to them, not only during points of treatment but during the other times in their life where they're making either good or poor health decisions. So that's one side of it where I think the digital framework and connecting with them will be ultimately important. We can't be just connected with them when they show up. We've got to be connected with them much more longitudinally in their life. So today, we see that the first 90 days of a patient's therapy is one of the highest-risk times for them. If they can work their way through that first 90 to 120 days, then you'll see that their longevity is substantially enhanced. We've improved that incident mortality rate, I think, some in the last few years. But I think we will see, as we begin to stratify what's the least cardiovascular stress test in starting a patient on dialysis, those kind of techniques will actually continue to reduce what those high rates of sort of early mortality are. I also think that if we attach to patients earlier in CKD, we'll find that we have the opportunity to improve their incident cardiovascular health, which is what they typically die from in those early months of dialysis. So any transition into dialysis offers an opportunity to do that in a much smoother way. And I think we would be able to reduce what we're seeing today probably by 30% to 50% at least over the coming years. Does that help at all?

Yes. That's very helpful. And just my second question, related specifically to sort of the use of digital data in all of this. I remember a previous CMD, in the discussion about the ESCO program and just integrated care, and you mentioned it briefly earlier on in your presentation around having sort of digital avatars for patients. Clearly, more ESCO-like programs would be better. But under the current reimbursement framework where you have a mix of fee-for-service patients and you have a growing number of Medicare Advantage where you can do integrated care in a much more meaningful way, do you have the sort of resources in place to be able to offer this as broadly as you want to? Or will we -- would you ideally be able to offer this to every single patient and you just don't have that financial capability today because of the limitations of fee-for-service?

Yes. So the avatars that we have used today have been to actually model something that has allowed us to adjust the algorithms that we use. And I actually think the impact of these 6,700 avatars have, in fact, been on all the patients that we treat, certainly in our North American population. Similarly, in our European population, we've used a machine learning technique to similarly manage anemia. So all of our patients across our U.S. enterprise, as an example, have benefited from the way we use the avatars to essentially run hundreds of virtual clinical trials to minute changes to the algorithms that we're actually using in clinical care today. That's the side where I think that digital data has impacted everybody already. And that led to a much more efficient use of ESAs, which was a huge cost savings that you recall from a couple of years ago, and sort of changing the benchmarking and baseline. We've seen this in bone and mineral disease as well. We've seen this in some of the other comorbidities that we have. And so I think it will be something that we'll continue to extend. And our resources that we have to support this, it's a relatively small team, but they are all focused on developing applied solutions to these things. So I'm still quite comfortable that I think these digital techniques can have broad application even outside of the ESCO market.

The next question is from Veronika Dubajova of Goldman Sachs.

Two questions for me, please, sort of both slightly bigger picture, I guess. The first one is undoubtedly the case that you guys have access to a tremendous amount of data and information about ESRD patients. And I'm just curious, I mean, obviously, beyond the use in things like value-based care to help you on the predictive analytics side, what other opportunities do you see, if any, to monetize the data that you have and the knowledge base that you have? And if you kind of look forward, maybe not tomorrow, but on a 3- to 10-year basis, are there opportunities that you think could drive incremental revenues for you guys? And then my second question is, and I appreciate this is a really tricky one. But obviously, there is a tremendous information -- amount of information and data. And I think part of the challenge that a lot of health care providers have is how to digest it all because it's coming from so many different systems. I presume that's one of your challenges. But if there was a magic wand you could wave, what sort of problem do you have in this digital landscape today other than digesting that kind of data that you would like solve, if you could?

Sure. So on the first, I think there are a number of ways in which we look to monetize the value of the digital data. One is, I think, operational efficiencies are one thing. And these operational efficiencies work towards where can we reduce costs by doing things more efficiently or doing them just to the right individual. That I think is sort of a passive way. On the other front, we, today, license some of our phenotypic data, and we want to combine that with our genotypic data to those that have invested in research. And we think the catalyst for that will be to believe that the amount of data, when we start looking at precision medicine and nephrology, that is going to be needed to unlock sort of the variation that exists in individual people and various groups of folks will be very powerful in being able to license that data to academic researchers, to industry researchers and to those that, in fact, actually see that there is great utility. Today, many of our studies in the genomic area, for example, are on 5,000, 10,000, 15,000 patients. I think that the complexity of the genetics that's involved with kidney disease is going to require that we do data sets that are 100,000 patients, 200,000 patients. And that's a strategy in which, through our clinical research arm, we think there is substantial opportunity to evolve biomarkers, to evolve the understanding of profiles of patients that researchers can look at and actually bring new innovations to the field through greater investment in that field. And then finally, some of the tools like the Kinexus tool that I described to you, our models were in our product set. We can provide and sell with products the ability to have this connected health ecosystem related to sensors in the environment, whether they are things that patients would wear on them that sends diagnostic information, whether they are actually passive in the environment and they understand how mobile is a patient in their own home, and has that changed over time. Those are all things that become service side opportunities to build out the service set that's creating remote monitoring of patients and those kinds of things. And then on your second question, Veronika, I think digesting data is certainly one thing. We are moving our environment into a much more substantial cloud-based environment that allows us to scale up the resources for computing that we need when we need it. When you begin to get into the interrogation of data at the genetics level, where an individual has a data set that has 2 trillion items to it, then suddenly, you need to actually make sure you have access to scaled infrastructure that's quite different. And so I would say digesting the data is trying to actually recognize that we're not going to be able to predict which data points are the important ones. We need to absorb them into an environment and then have a broad multidisciplinary group of people, like I described in that one slide of the different skill sets that we need, that are capable of interrogating it. And we've got to partner with others to recognize that we're not going to have all of the skills do that interrogation internally. We're going to have to develop relationships with others that allow us to do that. That's probably the biggest pain point that I have today is we can't get enough of the relationships together fast enough to do all of the things that potentially could be done in this area. And frankly, some of the scientific techniques for interrogating some of the data are still to be developed. Today, the genome, the coding part of the genome is around, let's just call it, 1% of the genome. But the 99% of the other part of the genome surely has an impact on how people express different variations of their own genetic code, and the scientific ability to interrogate that is really still developing. So those would be the areas that I am sort of watching very carefully on trying to both ingest the data and then do something practical with it.

The next question is from the line of Oliver Metzger of Commerzbank.

First one is a more general one. So data collection is, I think, it's a huge field of massive potential. But I'll focus my question more on the data which is generated during dialysis. So could you elaborate how the data collection for treatment has evolved over the last years? Do you think that you now basically collect all potential data which are necessary during dialysis or -- and that the potential from now comes mainly from more of the analysis or of a prediction -- of predictive analytics? Or do you think that going -- going forward, you might even generate more data? The second question is on the genomic analysis. So it's a big source of information. Could you elaborate about opportunities to react? So I think a higher genomic disposition towards certain diseases is definitely quite interesting information. But as long as you're not execute the scanning or direct a patient to scanning, you have just more information. So I clearly understand the link towards earlier detection of risk and also a higher sensibility to the disease of patients. But how can you make sure that a patient is treated adequately and it -- so in a pre-active phase. So in the past, you owned a hospitalist business, where basically you had a much clearer link to -- into the hospitals where you could also look that the respective patients with higher risk is treated adequately. But that's basically the past. So it would be quite interesting to know what do you think about this.

Yes. Thanks, Oliver. So let me start with the beginning of your question and talk about the evolution of data collection. One is I think during dialysis, our machinery generates a fair amount of information about the treatment characteristics of the actual therapeutic intervention that's going on. And we collect from that about 500 data points per patient per treatment. Some of those data points are actually data points about patient physiology. And that is one of the areas where things are continuing to expand rather substantially. So over the course of time and certainly, over the past 10 years, we have begun to access more data from each treatment that we have. And we are beginning to look at our device platforms. Our therapeutic platforms is also our diagnostic platforms. So within those diagnostic platforms, for example, we're now beginning to measure certain things that give us a picture of cardiovascular health during the course of that treatment. And some of that is related to not just typical things like heart rate and blood pressure, but also oxygen saturation and other things. One example would be in measuring oxygen saturation, when we got down to where we were measuring it on a very, very rapid basis where we're measuring multiple samples of oxygen saturation every minute, we began to upload that into the cloud environment. And we uncovered that about 15% to 17% of our patients go through a period of oxygen desaturation during the treatment. And if they do this consistently through at least 30% of the treatment, these are people that need certain things. And that's not just supplemental oxygen, but they need to actually have their lungs looked at and look at what is it that is generating this particular hypoxic state that they have during part of their treatments. Those are things that have led to us changing the way we behave around certain patients and the kinds of other physicians that we get in, all of these designed to try to, again, create a more efficient environment, a cost-efficient environment in getting a patient with a particular medical need to a particular specialty for evaluation, in this case, a pulmonologist, so that you can avoid a hospitalization or avoid some other really more terrible event. As part of your second question, though, I would say that in the proactive phase, we still have high visibility of patients that are in the CKD arena. And when we look at all of our physician relationships, we see that our doctors are taking care of patients that are coming into the system. And so whether it's through our active engagement of our provider network, if it's through our risk-based models that include chronic kidney disease care or it includes our work in the acute care arena. We still have a fair amount of visibility of what's happening in both the early stages of disease and in those patients that have sort of an acute insult. And the more and more we get closer to value-based care environments, the more those become actually critical for us to have greater involvement in both the things that influence the protection of their kidneys that influence the protection of their heart lungs and vascular system. And ultimately, that will protect their cognitive function during this time.

The next question is from the line of Tom Jones of Berenberg.

I have a couple of questions. All this kind of data analysis, data collection is great. But unless you can turn it into some kind of action, it's just data. And I guess the data tells you the path is -- the evolution of dialysis in all these data is to tell you what to do. But how to do it is, I guess, the question that then follows on from that. And I guess the question I have is when you use this data to potentially change treatment processes, change the way things are done, where do the bottlenecks appear? Are they at the patient level? Are they at the payer or payment structure level? Or are they at the physicians'/nurses' dialysis test level? I'm just trying to give a sense for how easy it is to turn the output of all this data management into meaningful clinical interventions that actually have a bearing on patient outcomes. It's all well and good knowing what you should be doing, but actually getting it done is a different question, I would say. And then I have a follow-up after that.

Yes. So let me start with that one, Tom. Thank you. It is the fact that the development of these tools leads to insights that generate care model changes. And as care model changes, there are things that have to be adopted by prescribing physicians. And the care model changes maybe the selection of an algorithm or a course of care that fundamentally changes the way a patient is approached because of the condition that they're in and the insights that you gain. I think that the pairing of the data with our provider relationships that we have is a way to, in fact, actually change both what we do and why we do it for those patients. We are understanding why we're addressing certain things to patients in a different way. So that I wouldn't call it a bottleneck, but I think that the way you actually make it applied is there's got to be an adequate technical solution. So the way you capture the data, create the insight and then get the insight in front of the provider or the patient, whoever needs to be making that decision, those are sort of the technical components of this. The practical components of this are you've got to validate that what you've done and have that trusting relationship with the provider and the patient to make sure that they know this becomes part of their decision-making. And that's where I think we've got probably several dozen examples where we've actually been able to utilize these digital techniques. Now the use that we probably excelled at so far is in helping provide new information or new methods for our providers. We are still on the journey to get more access and more connection into our patients. And I think that's where many of the things like we've got the patient hub. We've got the [ One for All ] app. These are directly related to applications that engage patients and help them make decisions that actually are going to be in their best interest. But that bottleneck is the last-mile solution, essentially. It is making sure that the provider knows that this information is available to them in recommending a course of action for a patient or a patient has information that says, "Oh, I need to make an adjustment or maybe do something a little different in my daily life than what I'm doing today." Those are where those challenges are. You had another question, Tom?

Yes. It was on the genomics question, really. I mean with the cost of whole genome sequencing now coming down to a point where it's kind of on par with a single dialysis treatment, are you aware or participating in the kind of whole genome sequencing projects within the CKD or the ESRD populations? And if so, do you think anything meaningful is going to come out of them in the not-too-distant future? Or is the kind of understanding of the CKD, ESRD relevant parts of the genome just not yet to the level it needs to be to make whole genome sequencing a worthwhile endeavor?

Yes. So the studies to date and the scientific interrogation to date on most of the genome has been with what's called whole exome sequencing. And the whole exomes and what are called SNPs or single nucleotide polymorphisms are the section of the coding regions. And that's where most of the research has been done today. There are a number of groups that have been -- begun to utilize whole genome sequencing. And I think part of our registry bet here is that, in fact, we ultimately will need that whole genome to understand fully the variations in what's called the epigenetics that actually leads to the disease states that our patients see. So I think there are examples. There are a number of young biotech startups that have begun to use whole genome sequencing, and there's still quite a bit of whole exome sequencing interrogation. The number of people across the world that are capable of doing whole genome interrogation is one that needs to expand. It's an area where we need more people and more scientists looking at that area. But I think with kidney disease, given that it's such a systemic nature of the disease, it covers a lot of other organs than just the kidney because of the way the kidney is used in the endocrine system and the cardiovascular control in other parts of the body, I think we're going to find that the variance and the variant calling that's required is going to require really quite large numbers of patients to sort of unlock the potential of precision medicine there. So I do think there are examples of a handful of young biotech companies that are doing whole genome, and there's a fair amount of work that's being done in the field on whole exomes right now.

Yes. And then I have kind of one follow-up question, if I may, which would pull the 2 together. But a lot of what you've described sort of focuses on the dialysis patient. But to what extent is that to apply or capabilities there -- or opportunity to apply all this to the CKD population? Because in a way, what would be good, and this is kind of the first upside down statement. But if you could get more people to dialysis, in effect, by preventing them from suffering mortality events related to other conditions, therefore, they make it on to dialysis. It's a terrible thing to say to be -- get more people on dialysis, but I think you see where I'm coming from. So how we view this kind of data management to take up into the CKD population to either improve the number or improve the health of people arriving at your doors?

So I think the whole goal of this genomics project is to attract an understanding of who is destined to have kidney problems in the CKD arena as well as how they're going to respond to particular treatments as it advances. There's no question, more people die of cardiovascular disease than ever get to end-stage kidney disease. So drug and drug classes like the SGLT2 inhibitors is a typical drug class where I think its greatest impact will not be in slowing CKD progression, although that is a positive impact for that. I think its greatest impact is when we're beginning to see the positive impact on cardiovascular health in the mid-stage CKD population. As we look at more precision approach where today, for example, with our pathologic classification of disease, when we biopsy a patient that we think has diabetic kidney disease, close to 30% of them actually have some other kidney disease going on. So we aren't as accurate as we need to be, and that's why we need these genomic and digital data sets to stratify people so they're getting actually the therapy that's most likely to help them, and we're actually treating pathways and mechanisms of injury in CKD rather than just the ultimate fibrosis that occurs to the kidney. So I think in pharma, there will be a huge play in CKD that will get much more specific with precision medicine. I think there are substantial opportunities to develop biomarkers that will understand which -- how we actually measure success and identify patients that are good potential candidates for certain drug classes and drug targets. And instead of just treating people with these very genetic immunosuppressive type agents, I think it will be a much more targeted approach to deal with the mechanisms of injury. That's true about CKD. It will be true about the drugs that we use for patients with end-stage kidney disease and the treatments we provide, and it will certainly be true in the transplant population in trying to recognize that we've got to become much closer to that world as well.

The next question is from the line of Ed Ridley-Day of Redburn.

First question, a little bit more detail. If you'd speak to your relationship with Livongo, which you announced last year, a little bit more idea about how that is working and how you said going forward, given that you have a lot of the data and so -- and they're kind of the new kid on the block, if you will. So interesting -- interested in the balance of that relationship and maybe compress you on why you feel you need to use them rather than you would want to develop that capability in-house. And the second one is related to some of the previous questions. With all this data, this always seems to me that Fresenius Medical Care has been in a position where you can be the wider provider of care in the diabetes continuum, but it never really happened. And one area, particularly for what you've spoken to today, where it was seen, as part of the ongoing strategic review, there is opportunity would be in cardiovascular care and building maybe that interventional cardiology, interventional radiology business out. But if you could speak to sort of that, sort of how that may be also an opportunity, that will be helpful.

Yes. Sure. So I would say that the diabetes care and the relationship with Livongo is a good representation of when I described I think we've got to create stronger and better partnerships with areas that are innovating things that have an opportunity to become a best-of-breed type of activity to support, whether directly or a comorbid disease. The engagement of patients and the ability to recognize that -- the impact of diabetes, when they have CKD, is very different than somebody who is -- just been diagnosed with diabetes and is trying to getting to understand how their lifestyle is going to change from that disorder. So the population of diabetics that we're interested in with Livongo is how do we engage people in good health care decisions in their regular life. And our feeling was that we needed to use some of the digital techniques and some of the engagement techniques that -- recognize they had shown success in some of the other cohorts of diabetic patients that Livongo had engaged with. And I think that these processes are still germinating in how we actually begin to utilize them. I think that as we again approach our value-based care environment, these become the sort of learning ground where we, in fact, can engage in how do we impact a health decision that a patient is making, not that a provider necessarily is making. And those are behavioral psychology techniques and activities. They aren't necessarily just medical physiology techniques and tactics. I think in the cardiovascular arena, it is important to recognize that the cardiovascular disease in the patient with kidney failure is very different based on the stage of kidney disease and the vintage of the patient with dealing with their disease. So what you see in earlier stages of kidney disease and in the -- in CKD prior to end-stage kidney disease is very different than the type of cardiovascular disease that you see in a patient who's been on dialysis for many years. Early in the course, you have much more impact from atherosclerotic vascular disease. Later in the course, you begin to see more impact from pressure and cardiovascular stress and the development of left ventricular hypertrophy and rhythm disturbances from an overburdened muscular heart. So there are a lot of things that, I think, if you were to go into the field of cardiovascular disease, you would want to look at it from the frame of the patient with advancing kidney disease because it's not going to look like the general population. And I think if we lump everything into just what would you do for the general cardiology field, we won't get the results that we want. So I think it is sort of a subfield that is quite ripe for us to be progressively more involved in. Likewise, I would say the next field beyond that is how we look at protection of the brain and the impact on cognitive function in patients over time. That will be the next horizon to try to meet up with once we get a better read on how to protect cardiovascular health across the whole continuum. But these are good areas that, and I think both of these are going to require that we don't do this completely independently on our own. We've got to do this with partnerships and looking at where the science is taking us.

And there are no more questions at this time. I would like to hand back to Dominik for closing comments.

Thank you, Frank, so much for your time that you spent with us and the insights you provided. Thank you, everyone, for participating today. As Haley said, we have no more questions. I hope you have been able to increase your understanding of the digital future for kidney care and our business a bit more. Please be informed that our next expert series call will be with our Chief Executive Officer for Global Research and Development, Dr. Olaf Schermeier, and it's not so soon. It is on September 20. And he will be giving you an insight on innovating dialysis, and we hope you will join us again. And of course, we look forward to speaking with you on May 6 for our Q1 earnings call. And with that, I'm closing the call and say thank you, and goodbye.

Ladies and gentlemen, the conference has now concluded, and you may disconnect your telephone. Thank you for joining, and have a pleasant day. Goodbye.