Transcription
Simon Turner 0:05
So thanks, everyone, it's a pleasure to be here, particularly discussing this very interesting and exciting topic. Maybe a little bit of background to set the scene. I'm a partner at sofinnova, we've relatively recently launched a strategy dedicated to the digital medicine space. Now, I'll explain what we mean by that, because this is exactly what our panelists will be talking about this intersection that we have between medical devices on the one hand, which have been traditionally single use or single application, versus these new data driven approaches and models that we're seeing leveraging AI and these tools and technologies. So when we think about this space, it's actually something that is totally new, totally revolutionary, because it enables us to generate so many more insights and actionability, that in the past was not actually possible. So we move away from just a hardware function. Now, suddenly, to the software piece, and that intermingling of the two, it opens us up to being able to actually push towards repeat customers repeat business models, opening up to novel people who are actually willing to not only acquire the insights in the data, but also see how they can leverage that for the development of new tools and technologies going forwards. So this panel that we've assembled, shares the vision, but also has gone through pieces of that puzzle, all the way from the very early stages of development, in certain cases, still in that process, all the way through to actual commercialization and achieving the hurdles that we sit there. So maybe if I can hand it over to you, gentlemen, to introduce yourself, how would you like to go first,
Hamed Hanafi 1:25
thank you very much. Thanks for being here. My name is Hamed. I'm the founder and CEO of NovaResp, we focus on obstructive sleep apnea. If anybody in the audience knows about sleep apnea, or CPAP, machines in their family or friends. Usually when someone is prescribed with a CPAP, they don't like using it. We solve that problem with AI, we predict and prevent apnea before they would happen. So the pressure of therapy lowers, sure you've blown up a party balloon before. In the very beginning, it's hard to push then it gets easier. Same with the airway, it gets easier if you know an athlete is coming. You can keep the airway open at lower pressures, make the machines more comfortable. We're purely software company that could turn into a hardware company. So great discussion today.
Simon Turner 2:15
Brilliant. Michael.
Michael Ebner 2:16
Yeah, thanks so much as well for the opportunity here. I'm Michael Ebner, co founder and CEO of Hypervision Surgical. We are a health tech company based in London. And our goal is to provide the surgeon a superhuman level of vision to help them see critical structures with much better clarity and give them certainty in making judgment calls on what to cut and what to spare and where to join this year. So the way we do it is what is offensive technology called hyper spectral imaging basically just means like many more color bands than the human eye can see and use this effectively to characterize tissue and visualize it in a very intuitive way back. So that really is like leader streaming instantly gives a decision making insights and your component of data is really important, because for the first time, we will have well, many more color bands or 10s of color bands, rather than just three at our fingertips during surgery that we can use to extract insights and inform. And that really isn't entirely, you know, fascinating prospect as we, as we think so. Yeah. So due to 200 there as well. So we are I guess on this journey, we are a the pre commercial stage, so very much on this, this conversation will be trying to figure out and I'm very happy as well to share the insights to the points we have been learning them thus far.
Simon Turner 3:38
So a lot to discover still on that path. Exactly. Gabriel. Thanks.
Gabe Jones 3:42
Gabe Jones, co founder CEO of Proprio. Proprio is a series B surgical navigation company out of Seattle. So we're doing real time 3d volumetric navigation for things like spine and cranial surgery. This is an FDA cleared platform for spine surgery. And the first applications are unique and that we can actually track individual vertebral bodies, for example, live in 3d through what's called Light Field rendering. And we're the pioneers of real time light field rendering. What this effectively allows you to do is use machine learning to segment out the entire anatomy, and then track it in 3d like CAD models live. And this gives the surgeon the ability to go from what current navigation is, which is essentially an x-ray stapled to the body, right? You're all familiar with reference frames and arcs and these fiducials that really dominate surgical navigation today, and limit its adoption to maybe 20 and 25%, and a space like spine. So we see this as a huge opportunity, real time 3d volumetric information. Data gives us opportunities to give clinicians really these advantages to perform better, faster and safer. And then on the back end, as we'll talk some more as we go through this really exciting conversation. I think the most interesting panel at the entire LSI Barcelona a little biased on that topic. But literally, I think all of us have the same belief that the future most successful medical device companies absolutely must be data intelligence companies on the back end. And we can discuss that a little further.
Simon Turner 5:11
Excellent. Good. Claude.
Claude Cohen-Bacrie 5:14
Thank you, Simon. Claude Cohen-Bacrie founder and CEO of e-Scopics is complex is a point of care ultrasound company, we've created a new way to democratize ultrasound, transforming ultrasound into a pure software. So the hardware component of our product is simply a transducer. And this transducer plugged on to a PC where we have a software and this software is an app. And we are developing a specific app for every specific clinical indication. The goal is to go outside radiology upstream from radiology, but also downstream, we are working in intervention. We are a software platform that provides ultrasound and I think the key to democratize ultrasound is to remove the burden of image interpretation. And the way we do that is by providing imaging, quantitative imaging biomarkers. And I think that this is exactly what this decision is about MRI, going from a simple image to an information, which can take the form of a biomarker using data, this biomarker can be an image, but it can also be a completely abstract interpretation of this data, which we find in the AI based algorithms that we develop. So we target specialty clinicians and primary care on one hand, interventionist, on the other hand, to spread ultrasound with this new technology.
Simon Turner 6:39
Excellent. So we all know medical devices, the way we've approached them for decades, now, suddenly, we have this new piece that's suddenly generating data. Gentlemen, I guess my first question is that it is just having the data suddenly enough, and that's the value itself, or is there more? How should we think about this?
Gabe Jones 6:55
You want to take that softball, to start?
Simon Turner 6:58
Go for it?
Gabe Jones 6:59
Yeah, so we were talking about beforehand, I think a useful way for maybe the audience to think about this, as we do is think about like three pillars. So data is very useful already. In clinical settings, I think we all have applications of that, that we've either proven or proving out right now, depending on the stage of the company. So use data to prove something clinically, to help a clinician to solve a problem, whether it's diagnostic or therapeutic. If you think about that as the first pillar, that sort of table stakes to even be in the conversation. That's how we think about it. Second would be can you then leverage the data to either accelerate that clinical treatment making better, faster, safer, perhaps, possibly even cheaper, or to look across the value chain, and then leverage the data to maybe make money or add value to someone else this would be thinking about implant companies, in our case, or other application providers who can connect to your platform, we're very jealous that your demo fits in your pocket. And you can connect to someone else's transducer. In our case, it's a surgical navigation platform. And then we collect over 250 gb of data, which is video navigation, imaging, pre op, post op, what implants were you utilize, and all of those data become this rich data set that then column three, you can start to explore other other ways to monetize for other customers other than the hospital? I think that's where sort of the framework can be applied going forward.
Simon Turner 8:19
Yeah. 100% agree. And I guess one of the the core kind of components of that is it's not just about generating the datasets, but it's a lot about the actionability. The what can we now do with the application setting? So when we think about that, then what are some of the core criteria that each of you have thought about have gone through in terms of your assessment process of Okay, great, we have our first application, but then we potentially have a second or third or fourth, building these services, quite frankly, that can be leveraged on top of your first initial product core, how have you kind of gone about that thinking?
Michael Ebner 8:48
Very happy to you. So of course, like, everything we do, needs to address it by every clinical a patient need. And, of course, as well, that needs to be married up with product market fit whether technology is capable of doing in like a short term, mid term long term. And then of course, trying to now stage this accordingly. In our particular context, what we're having really is the ability to capture information, so a video stream of hyperspectral that is entirely untapped and new. So but we need to demonstrate value of this in the first place. So the way we went about it, like if your wallets actually use this data to provide a first product that can also replicate color vision as a conventional wants to basically be comparable, but also provide a first information that is really invisible to the human eye, that is tissue perfusion. And using this with the same like algorithm, you know, baseline of technology, and video establishing that from the clinically as you've pointed out and the assumption these use cases for colorectal surgery in particular, but then the question is right, if you have shown is now for this indication. And the Holy Grail of careers is tissue differentiation. Like for nerves for for, you know your ducts tumor. But to get there, it's a long journey in terms of fundraising in terms of regulatory. So what we are trying to do is all to how can we always stage from one part to the next Ruby can also with essentially the kind of value chain, we are not interested in developing our own independent vision system commercially. It's just way too complicated. As a startup, it's a leg, it's almost bound to fail.
Gabe Jones 10:34
We know exactly how you feel.
Michael Ebner 10:34
I mean, trying like any conversations, how can we slot into existing vision systems and provide through complementarity, this needs to be step by step evaluated and established. So and from there, we can always gather more data to drive decision from there.
Simon Turner 10:47
Yeah, that's a super interesting. So when we're thinking about it, the way that the game kind of set up the three pillars there, it's it's not only building out across the pillars, but it's also within the verticals of each of those pillars, you've got to keep growing and keep growing, in fact.
Michael Ebner 10:58
Yeah, very much in a temporal nature, because everything like needs to be staggered and predicted to the extent possible in a very unpredictable environment.
Simon Turner 11:06
So how are you doing this?
Claude Cohen-Bacrie 11:08
Well, I think I'm very pragmatic in general. So I think that data contains a lot of hope. But at the end of the day, we are here to sell products. And the way we viewed that was that we can we can do conventional ultrasound imaging with these data, which is something that everybody knows, we can also attract existing quantitative imaging modalities, like measuring stiffness of tissue or measuring physical parameters. We don't invent anything here. And it's done with the same exact data. And this is something that can be sold. This is a very concrete, almost conventional medical product. However, every time we do an exam of the raw data are stored in the cloud. So indeed, after that, we'll be able to extract physical, non physical parameters, but features I would say that you cannot give any meaning about except what you intend to diagnose. So for example, today, we are in the Nash space, in liver imaging, we collect stiffness associated to fibrosis, attenuation associated to statuses, but we still miss inflammation. With all these data, plugging on all the clinical trials that Pharma is doing, we'll be able to retrospectively analyze this data and see whether there's any thing like, I cannot define this thing that is related to inflammation. And I think this is the power of data, but we don't sell that hope we sell what already exists. We've cleared the product, we've CE marked the product, and with this product will be able to generate the data that will then enhanced this product with future biomarkers. So I think we have to go step by step not promised the moon but you know, install and sell the rocket that will bring us to the moon.
Simon Turner 13:00
And I guess the each of these iterations, it almost goes faster and faster, because you have so much retrospective data that you can then look back at and retrain models as and when you go and increase that. I'm really curious, because you're you're sitting implanted on the CPAP machine. Yeah. What's what's the relationship like with the actual hardware vendors as well? Because I mean, that's critical in terms of one year deployment, commercialization of the business model.
Hamed Hanafi 13:21
Great question. And I can relate to the co-panelists here. So for us, it was a unique situation, as we said, you need to figure out what is the market need, we need to figure out what are the regulatory steps towards having this line of products set up. And also, I think money efficiency also comes into this. So for us, it was, you know, we have a software that could go on CPAP machines, and turn it from, you know, old fashioned machines into intelligent machines. But how intelligent, we had to think about the regulatory pathway. If I enable personalization of therapy and learning from patients night after night, of course, that's the future. If I enable it, now, all of a sudden, this software is not doesn't have a 510K pass, it has to go through more gruesome regulatory pathway, which also needs more cash to run. So it was a matter of what does the market care about. For us it was patient adherence, we realize that increasing patient adherence, just patients using their CPAP more than four hours a night in the beginning of being given a CPAP enables Medicare to cover the cost of that machine and the consumables that come up. If we increase that adherents by 10%, that's equivalent to $700 million a year more revenue for the manufacturers. So we decided to shrink the AI model size that would fit in the current chips that are on CPAP. Okay, enable that that huge incentive for manufacturers and for the that give out the see perhaps, they spent quite a bit of money trying to convince the patients to keep the see Pass, and most importantly, to improve quality of life immediately for patients and then line up the rest of the personalization and other SaaS models that could be built out for that.
Simon Turner 15:11
Interesting. So it's a win win situation you're trying to find here with both manufacturer and the healthcare provider, the healthcare payers,
Hamed Hanafi 15:17
Let's put that another win there. Win-win-win. Manufacturer, DME and patient.
Simon Turner 15:22
Very good, very good. Gabe, maybe going to you. Let's maybe go directly into the actual truth subject matter here. We're talking business model and business model innovation. Tell me how are you in your your approach right now valorizing, let's say your data generation in the immediate so right now in terms of actually changing or impacting clinical care, but then in the second phase of thinking about how can you product integration other services on top of that?
Gabe Jones 15:44
Yeah, so got a friend from intuitive right here in the front here. So I was thinking about different applications of the data, such as the amount of data you guys are collecting, you're able to go back and look at clinical performance efficiency, how the surgeons are utilizing it. And dude, it's got a great application for surgeons, I think that's great. Tell me if I'm wrong here. But if you figured out how to monetize that, but maybe it makes the product much more sticky and drives a lot of engagement. I think that's kind of layer one, right? Can we plug data back in that makes it more engaging for whoever the customer or the user is not always the same person that I think applications clinically that give us kind of unfair advantage, competitive advantages against whoever the players in the space right now, the example for us would be thinking about Medtronic or Stryker or one of these players in spine and orthopedics. The dominant paradigm and navigation is to put little markers on things. So you can track it in 3d space with infrared cameras if you're familiar. So that's what everyone is going to do. So unfair technology advantage based on the data is, well, we can build up 3d models of everything in the space and track it independently. Okay, all of a sudden, you now in a world where you can add value to someone else's implant, that is otherwise no offense, not smart. We won't say dumb, but not smart. It's not enabled in some way, you're effectively digitizing that product, building up a CAD model, then tracking it live in 3d. And through the performance of a surgery, you're able to say, hey, I can make your product measurably and demonstrably more effective. Because coming out of the ad with a navigation system and visualization, you get a data set that says this product performed better than that one. Now I've given you marketing materials and data to go sell your product better. And that gives you incentive to be on my platform. So you should probably pay me for that kind of an app store model. But thinking about both hardware and software as app. This is essentially chasing the pain, pain in the workflow pain for the surgeon and the piles of money. And then back to that third kind of column of framework there. What did we just observe in that incident in that intervention? Well, we saw preoperative imaging, pathology, diagnosis and intervention, surgically, we saw the implants and the techniques and the approaches and the navigation, the angles, everything that were used to tackle that and to treat it, otherwise known as a longitudinal data set of surgery. So that's valuable to any third party player who would want to come in and understand what was used to achieve what and what should be paid for. So now you have inventory information, you have billables, coding, operative notes, everything kind of embedded in a singular data set. That's valuable to anyone in the value chain. So we're two of the three steps through that process. Now, monetizing the third column to RCM companies and insurers, I think our next.
Simon Turner 18:31
So that's the next stage of it. Claude, how have you been kind of thinking about that.
Claude Cohen-Bacrie 18:35
we've observed the same kind of move, we, the software platform that we created for ultrasound imaging is also licensed to third party manufacturers, I mean, that can be licensed to a company that does indeed image guided robotics, but we license it to a company in the US that is commercializing medical devices for intervention, pure hardware, you know, you pay by the by the device kind of business model. And that goal is really to become a data companies specifically to to create more growth and to basically have their customer, I mean, create differentiation compared to the competitors. And they thought that the imaging data that we could provide could could add what you exactly describe, which is, what do we learn when we do the intervention from this imaging device that could help us not only do a better job, and our customer will will buy our device instead of others because we have this accessory that helps us do a better job, but also what we learn during the intervention, observing what we do, can actually help us go closer to the patient upstream. We would learn from an intervention, let's say on the properties of a vascular wall, then we could commercialize and distribute early on the day Vice that can characterize this vascular wall early enough so that the customers, the patient are oriented towards our medtech. Medical Device sooner. So I think it is both a way to differentiate yourself when you're a conventional medtech company to move into the data space. It's also a way to create more value. But I think at the end of the day, this is going to become a true barrier to entry for competition. We've talked a lot about all these AI models that are based on a data set, let's say a plug on to a big clinics that has a huge data set, and I create an AI model. What is my buyer barrier to entry, it is the data set. If now you create a device or you sell the device that helps you create this data set, you've increased the barrier to entry for competition, I think.
Simon Turner 20:56
Very interesting. So maybe taking a bit of a different swing here, there's there's another piece of the puzzle that we haven't actually addressed yet, which I find super interesting. It's all four of you, you built fundamental platforms, if I if I should have likened it to anything. There's of course, the direct monetization direct business models that you can build off of that. But then there's also thinking about the wider ecosystem play here, because let's say classical medical devices, they were relatively vertical it was take medical device, sell the hospital, implant and patient jobs done. Here, though, you're one creating insights and actionability on your own approaches. But is there an interaction to be had maybe with the rest of the ecosystem? I'm thinking, for example, with other surgical applications, or with you know, novel AI approach is now being developed for for sleep, if it's in OSA, if it's in sleep apnea, if it's in other modulations, and things? How are you thinking about that? And is it part of the strategy now? Or is this more of a next phase of this entire data driven business model iteration? Do you think?
Michael Ebner 21:53
So it's something that is very much on my mind, because these questions are really important. I mean, all to do to think about it perhaps more holistically. I mean, it's always fantastic. Like, I'd be able to think about data within medicine surgery, and how do we get business models from there, but it's also to we're looking into just for where I've recently seen it, and where it's a bit more established, say, for reimbursement of AI based technology in radiology, I mean, two hundreds product to be cleared at this point, and only six thus far have their own reimbursement code. Of course, this has all like knock on effect, what this means for potential business models, per se, we know that there is a lot more towards digital surgery, in robotics, trying to have an ecosystem where all these technologies can interact with each other. But still, the stakeholder fitting and how it can be priced in those particular kind of like streams and how these individual stakeholders can work with each other is very, very difficult to get right in practice. And I think we're just scratching the surface on that. So for a startup to actually work on that and try to figure it out is a hell of a journey. So for us, what we're trying to do now is the really kind of like the spine, advanced visualization, still has been very much on something that is subjective in one way or another, you know, surgeon make decision 'A' and then afterwards kind of like looking okay, well, what's the right position. In our case, we we provide a bit capture, and we can analyze an objective data stream on how tissue looks in the spectral domain, we can literally say, at the time of insertion here, there is now the level of perfusion at that point of organization that should very much help as well to around the perioperative periods to be able to have preoperative imaging, see through the surgical procedure, and afterwards post operative outcomes and being able to connect this with intraoperative data. So it's a very new way of of seeing it. And that's very appealing, of course, in ecosystem of seeing robotics, where it has many more ways of capturing data. So again, for us is to figure out how do we fit in there, right, we don't want to have no new imaging system that is thought of that as in like a separate one, because it can't be it needs to be tied into this value chain. So what you're trying to do is again, like understanding complementarity, and in our case, we do things a lot about augmentation. And then the question is, how can this be billed separately? And we are actually very keen as well as a question for the audience. Like experiences people have had, and I'm very, very happy as well to have conversations afterwards to to exchange notes I guess, in one way or another, but it is like to mine, it's like if it was straightforward, companies would have done it. So we are essentially pre digital environment and have a lot of opportunities that we see outside of medicine coming into medicine. So how do we do that? Not sure what I answered a question.
Simon Turner 24:57
So my takeaway is it's fair to say That one, we're still at the cusp of beginning to enable this fully. You guys are at the kind of forefront of this. In fact, with two companies
Gabe Jones 25:08
just teed up a really good thought, though, how can we look at other adjacent industries, maybe I'm just biased because I live in Seattle. And I have not only Starbucks, I apologize for the bad coffee that we have projected on the world. But we also have Microsoft and Amazon, increasingly Apple and Google Cloud. It's really the cloud capital of the world. So we, we have the flywheel business models kind of all around us, right? Whether it's Microsoft's flywheel or Amazon's, we understand how the cloud and Amazon Prime and the marketplace can all work together to drive more efficiency and frankly, value for the consumer of those services. I don't know if you guys are all use Amazon. But we're fully addicted to Amazon Prime, everything is a two hour delivery in Seattle. And it changes your behaviors. And so think about that in the context of medicine and healthcare. Again, back to the kind of the pillars, we must be clinical relevant, clinically relevant, we must impact the clinical decisions and make things better, faster, safer, that's table stakes, think that, you know gives you an opportunity to be there in the OR as you were you were alluding to, what can you do to add something to that? So now you collect data? In our case, we'd say a case like, or a condition like scoliosis is currently on addressable? And how do we know it's on addressable? Because the literature tells us only 35% of the time, is the care team able to achieve their goals in terms of correcting the anatomy of the spine into a healthy state within it's very measurable within five degrees in 3d space. Why are they unable to do that, because they don't have real time 3d information. So we're solving that on the front end. Think of that as the mini gear that you add to navigation in our case that accelerate that flywheel a little bit flywheels need to be built intentionally. So the third piece is data and access to data to third parties. This is implant companies, we're adding digitization to otherwise, maybe not smart product. And then the third piece that accelerates the whole thing is providing external third party access to the insights that come out from that. So whether that's, you know, RCM billing, coding, all those kind of applications I was discussing before, the belief is that will accelerate even the clinical product development, right. So the flywheel then drives reinforcement learning from a technical standpoint, back into the clinical front end, and the whole thing moves a lot faster.
Simon Turner 27:22
And I guess you can even start pushing this into education. So even when physicians are still being trained up, suddenly they're familiar with this, they're getting the best insights generated for them, they come into the hospital segment, they already know your product solutions, and suddenly bang, it's a much better adoption against.
Gabe Jones 27:37
Unfortunately, that's not a great business model, right? If you just had a clinical education, product or company, it's hard to monetize, we've learned that. So you have to kind of fit it into that your own ecosystem, while making money for the other players in the system. Very good point.
Claude Cohen-Bacrie 27:51
I like your point on education, because that's what data can be useful to. When, of course, we educate our customers to perform an ultrasound, it's a very targeted exam, the first customers, we have our diabetologist, pathologist, we teach them where to position the probe between the ribs, there's no issue about that, and where to position the region of interest to do their measurements. But we've also developed a software that runs on the data that are stored in the cloud, for our pharma clients that are performing clinical trials, looking for patients. And they want to have a visibility on how the exam is done to control the quality of what is done. So we've developed based on these data algorithms to automatically segment deliver, and basically create tools that that are going to be used for customers to do a better job and for us to do a better education, and ultimately, a completely digitized education, thanks to the leverage of this data that we have stored. So I think it's data. Of course, data is useful when you have a huge volume of data to I think we've not spoken about that. But this is the key, good data, well, triage data and the volume of data. So to get a lot of data, you have to get a lot of procedures if you're in intervention, or you have to democratize it's used very upstream in the in the chain if you're in the diagnosis or screening arena. And going back to the business, although we've chose a subscription based business model now for this ultrasound system, where you subscribe to the service. And you have, you know, the probe is just shipped via regular mail, you download the software, and you can generate your data. And so I think business model has to evolve also with this type of new type of business, subscription or pay per use, because we also trigger, we have specific billing codes that exists and we don't invent anything is a key to generate this volume of data. Very cool.
Hamed Hanafi 29:54
I can add to this. So, again, data, a lot of data for AI is it important. So for us, we could have focused on non invasive ventilators. for COPD patients at home, we could have focused on ICU, ventilators, anesthesia machines. The funny thing is, a lot of the angel investors who originally invested in NovaResp were actually physicians, because it wasn't a surprise to them that we could predict the respiratory failure prevented, because they see it with their eyes on the monitor, and they can't do anything about it until the next bad sign shows up. And that's what machine learning AI is you, you see something with your own eyes, it repeats, then you collect 1000s of samples of that, and you develop AI models with that. So we had to focus as an early stage company, we focused on sleep and decided to just collect that data.
Simon Turner 30:47
Interesting. So if I'm kind of summarizing this, and I realized we only got about seven minutes left, the actionability component of all of this is the critical piece. In fact, it's getting that initial actionability of the data that you've generated to then have an impact somewhere in some stage of the healthcare cycle, or even product development as such. To the next point, though, that flywheel getting that flywheel going, I'm curious to understand where do you see the biggest hurdles right now, because we've got this title and its future or pipe dream for data driven business models. I'm a firm believer that this is happening already now. So I'd say it's, it's pretty much the present will become something standard in the future. But what do you think the biggest hurdles are for that to actually be true, let's say across the entire industry,
Gabe Jones 31:30
I can start with when we're dealing with our customers, our primary customer doesn't really know the value of the data that we are collecting or that they already have. So a hospital system, we're dealing with one right now, I will mention the very, very large one, has been collecting physical forms on surgeries for the last seven years or so, actually going to surgeons and asking them to check boxes to say, you know, what,
Simon Turner 31:54
What did you do so huge filing cabinets paper, literally.
Gabe Jones 31:57
So 400,000 surgeries, spinal fusions, and you can narrow down which hospital system I'm talking about, it's one or two of the largest. But they've been asking surgeons to fill this out yet another documentation step, whether it's EMR EHR, coding, etc, that we're now leaning on the clinicians even further, who spend 80 85% of their time not doing clinical interventions to help patients because we ask them to do so much paperwork. Now, here's another form you have to fill out. So it's not surprising that the adherence rate if you will, for the surgeons, is fairly low. Thus, the data stream that's coming out of those surgeries is quite limited, then we ask them to go fill out an operative note. And maybe they do it a week or two later when they do their coding. And then somebody bothers them because they didn't fill out the coding forms correctly. This is really how it works, right? So even a very large system like this that does have 10s of 1000s of spinal fusions, every year across the US hasn't figured out how to capture even that simplest of information. What did you do in the surgery that drives so many other outputs? And so they don't know what that should cost? either. If I offered them a service, would it be $500 for surgery, in a surgery that cost 40 to $75,000 to level spinal fusion, depending on where you do it in North America, they don't even know the value of that thing. They know it's very hard to get clinicians to fill out that form. They know that they want that information. They don't know themselves how to monetize it with they could get perfect information.
Simon Turner 33:25
So it's even a question of how do you actually implement this to collect the data to then be able to
Gabe Jones 33:30
Back by question or about intuitive is it a way to drive more value from existing customers and make the product a little more sticky, because something we do in our businesses, once you introduce it, you can't charge for it incrementally, there's no way in hell, your customer base is going to turn that off once they have access to it. I think there's everything from that to here's a new thing, it costs $50,000 a month for every subscription, it layers on top of the platform, we get to kind of explore those because we're at the tip of this business model collecting that really valuable data.
Simon Turner 33:59
Very interesting.
Claude Cohen-Bacrie 34:00
I see three hurdles, when you have to deal with that. Number one we've talked about it's curation of the data, you have to have someone qualifies the data that you store so that you can actually use it. Complimentary data that you would want. I'm getting an image but I want the result of the biopsy for example to correlate that. And thirdly is the data ownership or data privacy and data. You know, we use so for all these three things. I think the key for us is a very strong partnership with our customers. And we every time we sell a device we really have to engage the customer that will become a partner for future value creation of our products. So it's to me involving customers is key.
Simon Turner 34:50
Very clear. Michael any last thoughts on this topic?
Michael Ebner 34:53
Having the burden of the last question. To me, as I touched on earlier is like I think as a as a community as a whole, like game from from pre digital era in some way transitioning towards that, there's a long way to go still. And how to connect all these stakeholders that essentially can can be complimentary, I think requires a fair degree of standardization. Otherwise, we will, I think end up in a situation where big players will have some form of monopoly, where to have like access to data, and they never ecosystem access to data and for for startups, like in some way driver of innovation, figure out how to be able to inject into that and to derive value from there for to have to have a chance as well to be to be viable thing. There's a lot on how as a community can be designed that the business models as well are very much like future proofed. And I don't at this point, like we I think it's fair to say we're we're still trying to figure it out. So it's too early for us, of course, you do have a public opinion on that. But it is very, very, it's, it's I think it's a long, long way to go still to make sure we are ready for I guess AI data driven business models, because we should all have data driven business models seem even gets more traditional complex, very,
Simon Turner 36:11
Very clear. So if I summarize, it's, it sounds like you need to have identified a clear first utility, something there, this model will apply an additional healthcare benefit health care outcome or something. The second is that you can eventually then start thinking about creating this flywheel this data flywheel on the one hand, but of course, Insight generation on the other, and then seeing how you move from an first pillar into a second one, which eventually can even lead to development of new products or automation, or at least added functionalities that you could then possibly dream up. But to your point, it's sometimes the systems are just not built for that data generation. So you need to find a way of, I guess, incentivizing that to take place to then be able to ultimately monetize on it. Is that a fair summary?
Gabe Jones 36:53
Yeah. And that may lead to a "Me Too" of that product where the paywall you, maybe you're giving it away first, because there's such a pain point to acquire this information in a form that they that is standardized, and templatized, not the most exciting and sexy product on the front end. But if it gets you access to a full ecosystem of data lifecycle of every procedure, well that you have to believe you can monetize because everybody wants access to that information. And you get the honor or the opportunity to go attempt to monetize it with each of the players who benefits from it. So in that case, it's really worthwhile to build that front end product. It sits on the platform, even if it's not an incremental lift in revenue day one, that you're in the game and you have an opportunity to monetize it.
Simon Turner 37:38
So almost a Google search function being totally free. But then of course, all the marketing inside and the revenues could potentially build on top of that.
Gabe Jones 37:43
That's a historically wonderful business model.
Simon Turner 37:47
Excellent. Gentlemen, I think that's the end of this session. But thank you very much for taking part. Thanks.
© 2024 Life Science Intelligence, Inc., All Rights Reserved. | Privacy Policy
