Raising the Bar for Employee Benefits with Ali Diab, CEO, Collective Health

Voiceover:
Welcome to Cowen Insights, a space that brings leading thinkers together to share insights and ideas shaping the world around us. Join us as we converse with the top minds who are influencing our global sectors.
Charles Rhyee:
Hi, my name is Charles Rhyee, Cowen's healthcare technology analyst, and welcome to a special Cowen FutureHealth podcast. Today's podcast is part of a series for our 2021 Cowen FutureHealth Conference, bringing together thought leaders, innovators, and investors to discuss how the convergence of healthcare, technology, and consumerism is changing the way we look at health, healthcare and the healthcare system. And in this episode, we'll be talking about re humanizing the visit to the doctor's office and to discuss the topic with me is Manny Krakaris, CEO of Augmedix, a pioneer in virtual medical documentation in live clinical support. Augmedix converts natural clinical patient conversation into medical documentation and provides live support, including referrals, orders, and reminders so clinicians can focus on what matters most, which is patient care. So, thanks Manny for joining us.
Manny Krakaris:
Hi. Great to be here, Charles. Thank you very much for inviting me.
Charles Rhyee:
Great. Hey, so if we think back now where we are here in 2021, but looking back, let's say over the last 10 years plus, one of the most important developments, I'd say in healthcare, has been really the digitization of health records and the really wide adoption of electronic health records. And, it's fair to say that, I think, this has led to a number of advancements in the US healthcare system, but at the same time, it's probably inadvertently led to additional administrative burdens for providers, potentially leading to other unintended consequences to our system. As you think about that, how do you think about that? Maybe, what are some of those adverse effects that we're seeing in the market as a result?
Manny Krakaris:
Well, Charles, there's no question that the electronic health record system has led to a higher level of healthcare quality, but in the process of doing that, it's imposed a significant amount of burden onto doctors. Today, the average doctor spends between two and three hours on documentation. Some of that time is spent during office hours, but also some of that time is spent at home at nights and on weekends, which has an adverse effect on their quality of life. And as a result, it's leading to a record number of physician burnout, where physicians are leaving the medical practice altogether. That alone is costing the healthcare industry about $4.6 billion a year and it's exacerbating the critical shortage of doctors which already exist in the US and it's estimated that that shortage is going to increase from about 30,000 to about 105,000 doctors by the year 2030. Also, because of how doctors actually take care of documentation, it diverts doctor's attention from the patients to whatever technology piece of hardware they're using to do the documentation during the encounter with patients and that results in a lower level of satisfaction on the part of patients.
Charles Rhyee:
Yeah. I can tell you for my own example, I remember being with a doctor just a few months ago, and he'd asked me questions and he would look at the screen and be typing and he turned halfway away and I understood what he was doing. So, it wasn't bothering me, but I can imagine for others. They would be a little, you feel like you only have five minutes of time. And he spends half the time looking away.
Manny Krakaris:
And that's the typical experience today.
Charles Rhyee:
Yeah. So, if we think about until now, what kind of solutions have providers try to use to alleviate some of that?
Manny Krakaris:
Well, there's a wide array of solutions that have become available in the last few years that grew out of the limitations of the original solution, which is a dictation based solution, which over time has turned out not to save as much time as initially believed, essentially because it is typing with your mouth. In fact, I just heard an anecdote this morning during our weekly sales call with our team where one of the prospects that one of our salespeople was talking to mentioned that they wanted to move away from dictation service because it was leading them horse. And so, they'd spend the first hour or two of the day just dictating notes from the prior day. And then, seeing patients and their voice would give out midway through the day. So, that's not a good solution for doctors. There's a wide range they emanated from a dictation services, but now have evolved to much broader, full service outsource solutions such as the ones that we provided, the Augmedix.
Charles Rhyee:
Yeah. And I think, you might've mentioned it just earlier. But if we think about the market opportunity to address this administrative burdens, it seems it's quite significant.
Manny Krakaris:
Yeah. Look, as far as we can see, there's four major categories of limitations that existing solutions fall into. One. They don't save doctors enough time, so they're not worth the effort. Two. They're too rigid and don't integrate easily into doctor workflows. Three. They're not mobile, which is increasingly important today because doctors don't always practice from a fixed location. And, four. There are limitations, which that pertain to the most prevalent model or a full service solution, which is the in-person model. COVID exposed limitations of that model and healthcare systems are moving away from it.
Charles Rhyee:
Yeah. So maybe, this is a good segue into talking about Augmedix, maybe talk about the Augmedix offerings, Augmedix live and Augmedix notes and how they are differentiated from the current solutions available today.
Manny Krakaris:
Sure. So, we start our business with a live offering, which is a real-time solution. Back in 2013, we were the first company to offer virtual real time documentation to address the medical note burden. And we complimented that service with a non-real time service, which we call notes last year. And, under the live service, we have a dedicated specialist to ensure delivery of complete, accurate notes shortly following the conclusion of the patient encounter. Moreover, because it's delivered via a bi-directional or synchronous communication channel, our live service enables an exchange of information between Augmedix and its customers to ensure the highest quality notes by eliminating any ambiguities that might occur or crop up during the patient encounter. And it also enables us to deliver other services such as care reminders, coding suggestions, and pending orders. The value of these additional services is greatest when delivered in real time, which gives us a significant advantage over the non real time services that are out there.
Charles Rhyee:
And I think, maybe some who are familiar with the company might remember, I think you guys started in 2013 using the Google Glass where doctors could be talking to the patients and seeing the information presented by their specialist in real time. Nowadays, it seems like a lot of physicians are using their smartphones to engage with you guys. Is that correct?
Manny Krakaris:
That is true. So, when we started a hundred percent on Google Glass, the first couple of generations of Google Glass were not designed for eight to 10 hours of continuous video and audio streaming so we migrated to smartphones. Today, about 90% of all of our doctors are on smartphones. However, Google released an enterprise, a true enterprise grade version of Google Glass. Last year, we piloted it very successfully in an emergency department of a hospital in Southern California. We're very excited about the performance of this particular piece of hardware. And we're now offering it to physicians, not just in the emergency departments, but for hospitalists and other applications where doctors prefer a hands-free solution that offers point of view to us.
Charles Rhyee:
And speaking of the best successful pilot in the emergency room, you've had a lot of success with major health systems. Maybe, talk about sort of the customers that you have so far and sort of the value that they find in the Augmedix solution.
Manny Krakaris:
Sure. So, we today have under contract for the top 10 and six of the top 20 health care systems in the US as customers, as measured by patient revenue. And, getting into large enterprises is not a trivial matter. It requires a tremendous amount of data protection that they're incredibly sensitive to. So, we not only have to comply with HIPAA standards in terms of data security, but we go well beyond that because the requirements of our big enterprise customers exceed the requirements of HIPAA. And, as you go into enterprises, they will typically start with small cohorts of doctors just to become acclimated with the service, but then as they become acclimated and you're able to demonstrate the ROI, you are able to more deeply penetrate those organizations. And that's been the model that we've been following, in fact, historically, and that's true up through the first quarter of this year. Roughly two-thirds of our growth comes from our existing customers and a third comes from new customers.
Charles Rhyee:
I think, a lot of companies have been challenged because of COVID. Just curious for Augmedix, how did the pandemic affect you? Were there any delays and how clients were responding? I'm sure they're all swamped with dealing with code itself. Maybe, touch on that. How does it look as we're kind of coming out of it now?
Manny Krakaris:
Sure. Well, I think, like most healthcare companies, we were adversely impacted by COVID because our customers patient volumes declined dramatically towards the end of March 2020 through June 2020. They started to pick up towards the latter part of June. And that of course translated into higher volumes for our company. We recovered very quickly. We ended up picking a business from some of our competitors who relied on an in-person model because that model did not work during the pandemic. It does not work during the pandemic. And we ended up having record bookings in Q3 of last year and that momentum is carried into Q4 and the first quarter of this year. And, we see that continuing through the rest of the year, as our customers are less pre-occupied with the pandemic and the impact that it's having on their operations.
Charles Rhyee:
So, we'd love to touch on this in-person type of business, having a [scribe 00:12:15], in physician offices, to do the documentation. I got to think, I mean, when you speak to some of these health systems, I mean, do any of them think of going back to having scribes given the experience with the pandemic?
Manny Krakaris:
Well, that's a great question. It's tough to put the genie back in the bottle once it's out. And I think, what happened during the pandemic, particularly towards the back half of 2020 and continue into this year is that enterprises became accustomed and acclimated to the idea of virtual documentation services and the ease with which, and the efficacy that they are able to deliver. So, I don't think that there's going to be a big rush to go back into in-person scribing on the part of those healthcare systems that have left it. I'm sure there's still going to be a market for it, for in-person scribing. But I think, the vast majority of large healthcare enterprises are migrating much more towards the virtual solutions.
Charles Rhyee:
And apart then, when we think about this opportunity, not only just the organic growth opportunities, but particularly those that are maybe migrating away from in-person scribes to a virtual kind of solution, what does the competitive landscape look like? There's obviously you guys, and I think, some people talked about nuance to a certain degree. I'm just curious, what are the players are out there in the market for this kind of service?
Manny Krakaris:
Well, there are several. And, when we first started back in 2013, there was really just us as a virtual player. You had the big incumbents nuance and that modal with their dictation services. But since then, several companies have cropped up. Many of them on the premise that they can deliver a complete, accurate, cogent medical note without any human intervention, as those propositions have been tested in the market, it's been revealed that they really can't deliver on that promise. And, I'll go into some of the limitations of technology in a moment, but what I think differentiates us from everybody else in the field are these four features of our service that no one else can replicate. We are the only company that offers these four features, which we believe are very important to doctors. One is, we use the ambient conversation, the natural conversation that exists, that occurs between doctors and patients as the input for the note creation process.
Second, we offer mobility to our customers. In other words, the means through which they access the service is not fixed or tied to a specific location. So, if they happen to want to see a patient from any given exam room in their facility, or from their offices, if they want to communicate with us, or if they're doing a virtual visit from their home offices, they don't have that limitation with our service because our service is completely mobile for our customers. Third, we are virtual. We're remote. So, we don't suffer from the limitations of the in-person model as we talked about earlier. And fourth, we offer a real-time service, which is very, very difficult to implement because of the logistical and technical challenges associated with it. So, those are the four differentiating factors, which we believe are very important in the marketplace. And we're the only company that offers that.
Now, in terms of the limitations of the pure AI, let's call it per AI approach to documentation. If you think about conversations that occur between people, they can go in any direction. And that is also the case with a doctor-patient encounter. Clearly, the doctor's trying to understand what the underlying issue is with a patient, but the patient may provide some background or may go on to some other unrelated subject that enters into that conversation. So, when that conversation passes through an ASR model and convert it into text, that text is essentially unstructured data. And no matter how good your algorithms are, it's very difficult to convert unstructured data into an accurate, in cogent medical note. And the requirements in our industry for accuracy are very, very high because the consequences of mistakes can be devastating. So, you have that inherent limitation when you're dealing with an ambient conversation of trying to provide contextual meaning or parameters to what is being said, so that then your technology can take over and make use of that information and generate the note automatically.
So, there are various approaches to addressing this challenge, and some companies deal with it at the very front end, where they asked their doctors and the patients to follow fairly rigid scripts in order for their algorithms to work properly. But that is not how most doctors like to practice medicine. We don't believe that's the case. We think that, that is moving away from re humanizing healthcare. And so, we don't think that technology should be inhibiting doctors from practicing medicine the way they want to practice medicine. We think, technology has to be an enabler, a facilitator to do that. So, we don't believe in that approach.
Charles Rhyee:
[crosstalk 00:18:54] same time, right. I'm sorry. I mean, at the same time, you'd also argue no one patient visit is the same. No one's coming in to complain about the same issue. Everyone have something unique about what's bothering them.
Manny Krakaris:
There are literally, millions of different patient visit types and they vary based on the specialty. So, you can imagine the permutations that you have to deal with when you're building models to generate accurate output from that text when it's converted from speech. So, yeah. Charles, there's all sorts of limitations out there. We don't think that a pure technology solution is going to be effective for a very long time. We need a lot more data, a lot more context in order to get there. And that's going to take a long, long time. In the meantime, we are learning. Everybody out there is learning through machine learning. We're harnessing the data that we capture today to help us understand what the contextual parameters should be with a particular patient visit. And that helps us reduce those permutations. [inaudible 00:20:12] is more manageable for our algorithms to work more effectively with, but it's still not going to be perfect. You're still going to need someone to provide some quality control at the back end of that process to deliver the level of accuracy that's demanded by our healthcare system.
Charles Rhyee:
Then, for Augmedix, maybe talk about, because I think you guys do it a little bit differently than how you bring in technology into the process and AI into it. It seems like you bring it in the middle.
Manny Krakaris:
We do. That's exactly right. So, because our founding principle is on re humanizing healthcare, which means allowing doctors to have an uninhibited conversation with their patients. We recognize the limitations, the inherent limitations of AI. And what we do is we build technology behind that conversation after it occurs. And the technology is built around a tool we call note builder, which we filed several patents. And what note builder does is, it uses, it harnesses the best of what the human brain can do with technology, what technology can bring to the table. And, what the human mind can do that a machine can't do today is provide contextual meaning or parameters to unstructured data. And that's what we ask of our specialists to do. So, they're just listening for keywords during the conversation that encountered between a doctor and a patient and clicking on the selections that are presented to them.
And once they make those clicks, those selections, then a sentence is automatically generated for them in the appropriate section of the medical note. That sentence will be grammatically correct and medically accurate. And it will, as I mentioned, get populated in the appropriate section of the medical note. So, that's our approach to it. We are through machine learning, reducing the amount of effort and time required by a specialist to provide that conceptual meaning, but they're still necessary because there just isn't enough data today to cover all of the permutations that exist in the real world.
Charles Rhyee:
Do you think that might be possible in the future? Is that still maybe the holy grail of it all is to be-
Manny Krakaris:
Well, it is the holy grail. Everybody's trying to get there. It's just a very difficult problem to solve, different industries, including the one that I came from before I joined Augmedix, have a lower threshold for accuracy and that's perfectly fine for those industries. But in our industry, as I mentioned before, the consequences of making a mistake can be devastating. And so, we can't afford to do that in this industry. It's almost like the autonomous vehicle. There was a great deal of promise with the technology, but at the end of the day, the technology has to be absolutely bulletproof in order for society to accept the consequences of a mistake or a problem that occurs from that technology because the humans are, we're kind of given some latitude in terms of mistakes we make, but if it comes to software, there is no such latitude. People are not as forgiving about mistakes that occur with software.
Charles Rhyee:
I'm sure nobody wants a doctor practicing on autopilot, right?
Manny Krakaris:
No, they don't.
Charles Rhyee:
And obviously this is a big opportunity here for you, but before touching on that, talk about your clients in that sense. Sort of, what are the outcomes they're seeing and what are the indicators that are looking at that deploying Augmedix has been a real positive for them?
Manny Krakaris:
Well, we focus, we have a very data-driven approach when we talk to enterprises. And what we try to do is map out each one of the physicians in these large groups based on a couple of performance metrics that are generally measured by large enterprises, WRBU's, which is a standard measure of productivity used in the healthcare industry for physician's productivity. And then, the other metric is the amount of time each physician spends per patient visit in the EHR. And we map everybody based on those two metrics on a grid and we identified based on our empirical data that we've captured from our existing customers, those doctors, where we believe would benefit the most from our service.
They happened to appear on the steepest slope of that curve, where everybody's plotted against, and we identify those doctors by name. And then, we set up a program where we monitor their performance after they'd been given the service and we measured them against those two criteria. And then, based on the outcome of those initial cohorts, we measure our performance against what we expect to perform, how we expected to perform. And generally, we've done quite well. And that's why we were given more doctors for those particular enterprises. So, it's a very data driven approach to penetrating those organizations and they can measure the effectiveness of our service based on top line improvement.
Charles Rhyee:
That's helpful. So obviously, given the market opportunity that the sector is clearly attracting more and more attention. And I think, the most recent announcement out there was Microsoft's acquisition of Nuance. Do you see others following Microsoft's lead and entering the market as well? How do you see that kind of playing out and not to have you applying too much, but sort of, what do you think the thinking is in terms of when these big tech companies are coming into healthcare, I guess in general, and then this area specifically?
Manny Krakaris:
Yeah. Well, look. We welcome Microsoft century, the market. They've already brought a lot more attention to the documentation problem than existed before, and they're giving further credence to the outsourced solution. So, that's a positive. They're helping to educate the industry and providing credibility to the virtual solutions. I suspect that Microsoft is going to try to leverage Nuance's significant market penetration from their Dragon service. But, we believe, the primary reason why Microsoft bought Nuance or is buying Nuance is because of a service, a full service solution that Nuance introduced last year called DAX, which stands for Dragon Ambient Experience. And, it is basically positioned directly against our notes service. It's a non-real time asynchronous service, but it is full service in terms of delivering a completed note to the customer based on an ambient conversation.
I think, the challenge that Microsoft is going to encounter is that the purchasing decision for a product like Dragon is considerably different than that for a product like DAX. Dragon is an enterprise wide, low cost piece of software that an enterprise can buy for all of its doctors and the people making those decisions are different from the people who are making the decision to buy a full service model. The full service solution is very doctor specific. It is predicated on specific workflows of doctors and it's a much higher priced service, as well.
So, I don't think you can just automatically assume that everybody who's on Dragon is going to upgrade to DAX. I think, that's a big leap. But nevertheless, I think that it's a fantastic move for the industry that Microsoft's in it. And I do think that, it's going to attract other major players. I don't think that it's such a big market. I don't think that other big companies call them, Google, or IBM, or Amazon are going to sit idly by and seed the entire healthcare market to Microsoft. I think, Microsoft's got a good head start, but I don't think that they're going to be the only game in town.
Charles Rhyee:
Yeah. So, maybe just around out here then, talk about what's next for Augmedix. What should investors be looking out for as we think about the back half of this year, looking out further, anything to highlight and keep an eye on?
Manny Krakaris:
Sure. So as I mentioned, we're continuing to build out on our automation technology. We're adding services and more features to our core offerings. At some point, some of those additional services will be fully automated. So, we're very excited about that. We're actually working on full automation of some of those additional features and services.
Charles Rhyee:
[crosstalk 00:30:28] rough examples.
Manny Krakaris:
Sure. So, I have to be careful in terms of what we're able to disclose, but if you think about care reminders, which is something that we offer today, we do it manually. The documentation specialist has access to the EHR, the patient's EHR, and they will look at the flags that are already in there and ensure that the doctor's aware of the key flags during the encounter so that they don't miss anything from care gap perspective. Now going forward, we believe we can automate those notifications to the physician without the need for a specialist to go into the EHR and pull them out manually. There are other services like coding, for example, that are very amenable to automation. There's already several stand-alone coding companies. There are different levels of coding that the industry needs to undertake. And, we believe we can offer at least some of that on a fully automated basis.
We're also going to leverage our unique asset, which is the synchronous communication channel we have with the point of care that is unique. We own that. And, it enables us to deliver other types of relevant medical information in a more targeted and efficient manner than is possible through traditional media and delivery mechanisms today. One thing about the healthcare industry, which you know, Charles, is that there's a tremendous amount of data that's out there, but it's siloed. It's housed in these very discreet silos across the industry, and much of it is transmitted or delivered indiscriminately. And so, there's a tremendous amount of inefficiency inherent in the industry in terms of how to deliver that critical information, where it is needed most. And we represent a very efficient distribution channel to the point of care. And there's some information that is really critical to be delivered at the point of care during a patient encounter. And that's what we want to take full advantage of.
Charles Rhyee:
If I remember right, with the augmented solution, you were already doing a lot of that. A physician is speaking with this patient and the specialist can toggle up the most recent lab results and make sure the physician has seen them, things like that. When you're talking about this type of additional information that you could bring to the point of care, maybe give an example of that. But, I'm curious how much, is this now enabled because of the inter-operability final rules and the getting rid of information blocking. Is that enabled some of these services to be easier to deliver through Augmedix than before?
Manny Krakaris:
That's a great question. So, I believe it is. We're not just talking about information that we would originate Augmedix, but information that is originated by third parties. And they're searching. Some of these third parties are searching for more efficient distribution channels to get the information to the point of care where it matters most either to the doctor or to the patient. And we just happen to represent that more efficient communication channel. Information could be from big pharma, for example, in terms of eliciting enrollment in clinical trials for oncology. So, there's all sorts of information that is out there that is trying to get to the right home in the most efficient manner. And we happen to be sitting on it.
Charles Rhyee:
Yeah. That sounds really exciting. So, it sounds like the future is really pretty exciting there and we look forward to really learning more and hearing more about Augmedix in the future here.
Manny Krakaris:
Well, we're excited as well, Charles. And, we're just building the company one step at a time, but the future looks pretty good.
Charles Rhyee:
Yeah. It certainly does. And so, I think we'll end it here. And Manny, thanks. Thanks a lot for joining us today for this podcast and look forward to hearing more updates in the future.
Manny Krakaris:
Thanks for having us on the show.
Voiceover:
Thanks for joining us. Stay tuned for the next episode of Cowen Insights.

This podcast should not be copied, distributed, published or reproduced, in whole or in part. The information contained in this recording was obtained from publicly available sources, has not been independently verified by TD Securities, may not be current, and TD Securities has no obligation to provide any updates or changes. All price references and market forecasts are as of the date of recording. The views and opinions expressed in this podcast are not necessarily those of TD Securities and may differ from the views and opinions of other departments or divisions of TD Securities and its affiliates. TD Securities is not providing any financial, economic, legal, accounting, or tax advice or recommendations in this podcast. The information contained in this podcast does not constitute investment advice or an offer to buy or sell securities or any other product and should not be relied upon to evaluate any potential transaction. Neither TD Securities nor any of its affiliates makes any representation or warranty, express or implied, as to the accuracy or completeness of the statements or any information contained in this podcast and any liability therefore (including in respect of direct, indirect or consequential loss or damage) is expressly disclaimed.