The a16z Show - When Medicine Goes Virtual

Episode Date: March 19, 2020

We’re at a moment where we are now seeing medicine go virtual faster, and at a scale that it has never done before. In this conversation,  a16z bio general partners Vijay Pande and Julie Yoo, who come from the worlds of bio, technology and care delivery,  talk with Hanne Tidnam all about what exactly virtual care and “telemedicine” is... and what it isn’t; what it works well for, what doesn’t (yet), and where there’s potential; and finally, the current pain points (including regulation), and what we’ll learn from this current moment for the next generation of virtual medicine tools. Stay Updated:Find a16z on YouTube: YouTubeFind a16z on XFind a16z on LinkedInListen to the a16z Show on SpotifyListen to the a16z Show on Apple PodcastsFollow our host: https://twitter.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Hi and welcome to the A16Z podcast. I'm Hannah, and this is another in our series of all remote podcasts. We're at a moment where we're now seeing medicine go virtual faster and at a scale that it has never done before. So in this conversation with A16Z Bio-General partners, Vijay Ponday and Julie Yu, who come from the worlds of biotech and care delivery, we talk about what exactly virtual care and telemedicine is and what it isn't, what it works well for, what doesn't yet, and where there's potential. And finally, the current pain points, including regulation and what will learn from this current moment for the next generation of tools. Stay tuned for another episode soon where we'll also cover the clinical perspective from the field next. I'm going to tell you guys right now that there may be some dog barks and kid stuff in the background. Okay, so we're all getting these messages from all our providers telling us to, you know, use virtual chat to use all these different telemedicine tools. So we're in a moment where medicine is really going virtual at scale.

Can we start by just talking about what virtual medicine or telemedicine actually means, what those different categories are? Is it all the same thing? Like, what are we actually talking about here? Typically what people think about when we say virtual care is probably the traditional sort of video visit where, you know, you have two screens. The patient and the provider are talking to each other live. You know, virtual care, I think, is much more of like a broader paradigm around, you know,

how do you sort of overcome the constraints of the traditional healthcare system, which are largely, I would say two things. One is geography, which is that, you know, typically you as a patient, the demand side of the market only really has access to the supply side that is within a reasonable radius of where they physically are. And then also the physical brick and mortar component of health care. So basically a way to get around the fact that you, at the moment, you have a certain doctor within your geographic range and a certain provider, end of story.

Yeah. And it's more than just video business as well. It could be asynchronous messaging. It could be continuous monitoring. It could be, you know, sensing. Frankly, it could even be a telephone call. I mean, we can go low tech in addition to higher tech. And as an alternative, just everybody goes to the emergency room because they don't know what they're doing, just even the triaging of the telephone call goes a long way. In terms of on the biological side, what works for telemedicine and what doesn't? I can say like, oh, I have a fever and my ear really hurt. I'm highly suspicious of an earache and like probably you could make that diagnosis based on a lot of stuff without actually looking in my eardrum, but like listening to lungs or listening to heart, like how, what are the limitations of what works and what doesn't for this particular medium? For a general practitioner, you could probably do a vast majority of what is done there. I mean, certainly not everything. Let's say just even in the routing function where we're trying to just understand what are the more serious cases and what has to be done, you could probably do a lot of trium.

And that might be the most important thing. And the key thing is, especially in a situation like we have today, you have routing with the benefit of not having to bring someone in physically. And, you know, it's natural thing about what you can't do, but just the unique things of what you can do, reaching people immediately and also keeping them sort of quarantined is particularly intriguing. In many ways, like an ER or a hospital as well as most dangerous places to be just in general. So you're saying essentially it's not really about diagnosis, it's about triage? I think diagnosis goes hand in hand with triage, but that might be one of the biggest wins just to know what to escalate and how to handle it. In some ways, that's what a GP's job is, you know, to say,

oh, this is something I can be dealt with home care or this is something that could be escalated, that needs to be escalated. The telephone call, the call your kind of pediatric nurse stuff, right? Like there's been some level of that already. We're starting to lean on. that a lot more in this particular moment. The video chat is definitely one of the things we're seeing most. Where are we right now in how much we're using these tools and how kind of robust they are in their rollout in the system? When you think about like the traditional phone call, which yes, is a form of virtual care, one of the challenges of that model is that every single caller is treated the same. And when they're waiting in the queue, there's no way to understand how to

effectively triage the ones that might have higher risk versus lower risk. And so there's a whole slew of companies that have a virtual agent or like a chatbot that essentially can ask you questions in a digital form in a self-service way that, you know, sort of prioritize the level of risk of a patient prior to them even engaging with the health care system. That's one thing that will unlock a bunch of capacity is, you know, rather than just brute force putting everyone in a line and waiting until a human answer the phone to figure out where they need to go, these technologies can actually sort of be more intelligent about how to route people in the right direction up front.

And is that happening yet now? When we have this telemedicine conversation, it feels like I'm in line, I'm waiting, and then the nurse pops up and we have a conversation. I may type a few things in, but is it actually being sort of prioritized? In Pockets, so there are a set of larger employers and a set of larger health plans that have partnered with these digital health companies to make those tools available to their members or to their employees, but it's not by any means in the mainstream. But you are seeing, you know, all over the web, all these companies are broadcasting the fact that they have

intelligence in those virtual chatbots that can help people assess what their level of risk is with regards to COVID specifically, but also all sorts of other things. And essentially, allow them to determine whether or not they actually need to come into a physical break of mortar clinic. You know, compared to what we could be doing, it's really intriguing in that it wouldn't take much for people to do vitals at home. If you think about the sort of virtual paradigm is, you know, how can the doctor connect with measuring things? The fact that now you can measure a lot of things at home such that, you know, maybe even $100 would get you a kit that your family could use to get basics plus plus, you know, maybe even includes like a stethoscope that the, that can send the

sounds of your heart and your lungs and so on to the doctor. I think there's a lot more that could be done than what we're doing right now. When you think about the tools that we all, like, the vast majority of Americans at least have like in their household set of things, like thermometer is definitely one where it's pretty much, you can assume that, you know, most citizens have a thermometer in their house. But there's many, many other categories of like tests, essentially that either aren't available to consumers off the shelf at your CVS and actually require either you to send a sample into a central lab somewhere and, you know, wait for the cycle to run to get back the results or that you still need to come in. And that's one of the big challenges

right now that we're seeing with COVID is that there is no at-home test. And you have to actually come into these physical facilities to, you know, both have the sample taken and the lab test run. And that's just exacerbating the supply side problem right now with regards to capacity. And so I think that's going to be a big area where we're already starting to see tremendous movement. I think this moment is highlighting the fact that there is so much more that we need to be investing in as far as innovation to bring those tests truly into a package modality that allows consumers to actually do them in their home. So, Julie, though, are you thinking if it is like just a blood or a urine sample that probably could be collected by a mobile phobotomist as well? Correct, yeah, where essentially like the sample collection kit can be sent to your house. and it's either a urine sample or a simple blood prick or a saliva sample.

And that kid still needs to be sent to a central lab to be actually run. And then again, there's a latency to getting the results back. But you also have other tests. Like, I mean, the pregnancy test is the most canonical example of this where you can run the entire end-to-end test in your home and get immediate results. And so I think that's what we should aspire to is that a larger portion of sort of what we call standard blood tests should be available in that kind of packaging. Yeah, and it would be great if it was done at home, but I think even like sending to a central

lab isn't really that different than what a GP would do. Well, what is, I mean, what are the limits? Because I hear you pointing out a lot of different things that we could theoretically be doing even right now, but where are the limits right now of like, no, this is really this, you know, the opportunity for virtual medicine does end here. You need human to human care delivery. The obvious one is if you need a procedure done, like a surgery. then clearly today that is something that does require coming to a physical operating room type setting,

like a facility that can actually handle that kind of high-risk procedure. But that seems so far out on the spectrum. Yeah, I know. It's probably a while until everyone has like a Da Vinci robot in their home. Right. You know, a surgeon can control remotely. But, you know, hey, we can all dream. But that portion of the market has been unbundling as well in terms of it used to be the case that you had to go to a hospital.

Now we have these ambulatory surgery centers that specialize just, on outpatient surgical procedures. So there are certain components of that that you could predict ultimately make it out to the community. So that's kind of one category. The other thing that that's worth mentioning is when we talk about virtual care, we typically think about the patient to provider interaction

as the component that needs to be virtualized. But there's a whole backend, like provider-to-provider communications still are not virtualized either. Like a lot of what you are seeing out there on social media and physicians, you know, sort of speaking out about what's happening with Corona, is that they themselves don't have the means to communicate with each other in a real-time fashion. So at the moment, where does that break down? So even if you have like a fantastic, you know,

virtual visit with your doctor, you don't have to go in, your doctor can call you antibiotics. Or like, where does the system start failing in that data sharing behind the scenes, provider to provider? Yeah, I would say like the best case scenario is that it just slows things down where you have to have more manual processes in place to aggregate information that, the next provider who you see needs to be able to make the right decision. Worst case scenario is that you actually don't have access to that data and you either are blind to that and therefore make an incorrect or inaccurate decision or that you have to repeat whatever was done to you before,

which obviously adds cost to the system. So I think those are a couple of the examples that we see. That's rampant today in terms of where a lot of the unnecessary costs in the system are is simply because we don't have data liquidity and therefore there's a lot repeat testing and assessment that needs to be done to get a holistic view of every patient at every individual encounter. It's kind of a weird juxtaposition of, you know, kind of the good and the bad that's happening right now, but CMS and ONC just published their interoperability rules that mandate the adoption of certain interoperability standards and technologies for hospitals

to exchange medical information. And, you know, until that is in place, I think one of the biggest constraints to actually virtualizing care models is the exchange of data that enables all of the decentralized players to have access to the same information. You know, it's fun to connect to what Julie is talking about, about virtual care being not just sort of GP to patient at home in that you could imagine having a sort of virtual care where you have a specialist consult done virtually. Exactly. Because right now, often the patient has to reschedule a whole other meeting.

and having that done briefly, virtually would be particularly intriguing. And then another topic that this connects to that's, I think, a broader stuff that we've spoken about in the past is sort of the unbundling of the hospital. It's interesting just to think how far you can unbundle it with the goal of keeping people out of hospitals as much as we can, keep them at home and do as much as you can do at home, do it in local centers as much you can do there, and only escalate to a hospital. If you're having like a triple bypass, I don't think that's ever going to be. done at home, certainly not a time soon.

But like, I kind of want to know, like, what's the farthest you can imagine it, you know, right now with what we've got right now, short of a triple bypass. I think all the reading, you could do a lot of reading at home because you could do the blood test, you could do urine tests, you could do various measurements, but like the writing where you do anything to a person, I think probably that might be just way too far. Although I have to say I had like a weird foot thing, you know, last year and my doctor prescribed a virtual physical therapist to me who like, you know, we had an appointment. We did exercises and it was much easier than going somewhere. Exactly. Physical therapy is probably one of the places where

it's an intervention that traditionally has required going to a clinic, but that is one of the big areas where you start to see kind of at home innovation. Yeah, that's a great point too, especially since, you know, for physical therapy, but for medicine and general compliance is such an issue. Yeah. And if this just helps with compliance, I mean, to some extreme, just having someone watch you to make sure you take your meds and so on, like having a, having a parent or something like that or a buddy. So PT often says about compliance, all this compliance stuff that probably could be improved with sort of virtual care. Part of the problems that were, in many ways, we're not even trying. I think that there's a lot more that we could be doing, but it means

sort of capital outlay to get the programs going and then to get rolling. But that, you know, so that's on the provider side. On the page, side, you could argue, too, that there are a lot of people that don't take advantage of even what we could do now. And that part of what is powerful about the doctor's visit is just sitting there with the doctor and how much that is a form of medicine and so on. Right. I think, you know, they've done these interesting tests where they have, you know, sort of these different variants of the doctor visit and just the doctor giving attention has a huge sort of placebo-like effect or sort of positive effect. And so we wouldn't want that human connection part to be lost and how to

do that is tricky. On the other hand, maybe even just cultural things change and it just becomes much more of a norm to connect to people virtually. That may change. That may change a lot. I remember even like when the history of the telephone, people originally thought the telephone would be seen as such an impersonal kind of like no one would ever want to use that versus, you know, connecting in person. I think we sort of just got over that and got I mean, ironically, we're doing this remotely. None of us are in the same room right now, but I feel like it would be just like it would be the same as we were.

Yeah, almost the same. Yeah. Yeah. If we can just get over it and get used to it, that may actually still incur a lot of those benefits. What else in terms of stress points? Because we're starting to see, you know, this is going to be like a fast, big, hard rollout of a whole bunch of stuff for a bunch of people that have never used telemedicine

in immediate use all over the place. So what are some of the other stress points that you think we're going to start see popping up that, like, well, you know, next generation of virtual medicine tools will learn from? Well, two structural things that we should definitely mention are on the regulatory and like the payment side where if I'm a doctor and I see what's going on and I have a motivation to spin up my own virtual care practice. It's very non-trivial to do that on the fly because of regulation around licensure.

So it is not the case that I can treat virtually every patient across all 50 states unless I am appropriately licensed in the states where the patients are located. It's definitely a source of friction that prevents a lot of companies from actually turning this on from day one. So in other words, even though it's virtual, it's still very local. It's still very local, exactly. There's actually an interesting study that came out that showed that there are literally no two states in the U.S. that have the same policies. And even within the states, some of the policies conflict between like state law and Medicaid law. And it's very convoluted. So that whole, you know, sort of jungle of policy is one big thing that there's been talk of change.

And I wonder, you know, given the current situation, how much that will rise to the top as a potential regulatory change that might be put on the table. What sort of change do you think could happen? Well, just to relax the constraints on licensure so that there's essentially like imagine like a common app type construct. where you could apply once and have coverage across multiple states. And then the reimbursement one is interesting because we just saw that the White House signed an emergency bill that relaxed the constraints on reimbursement for telehealth services for the Medicare population because historically that's been another huge constraint that it was only reimbursable under very specific circumstances.

For instance, like if patients were located in rural areas that were deemed, you know, sort of low access, those were the only situation. where you could get reimbursed, but now they've taken that that off the table. We already see positive tailwinds there, but that historically has also been a big challenge is just getting paid for doing the service. Interesting. How about scalability? One of the things I'm very curious about is how this could help scaling.

And there's different versions of scaling. One of the real challenges is just how do you schedule and sort of do the people matching problem? And if you had just a bank of virtual doctors or art, ends that could then be much more easily routed to anyone throughout the country. You could do load balancing between regions and so on in a way where, you know, everyone would be at very high capacity. And in situations where there is just extreme need, you could have a five-minute virtual visit that maybe gets the basics done in a way that that just really wouldn't be possible to do in person where you're just with the rooms packed and the parking packed and the

roads fact. There's just things that you can do at scale that you couldn't do in other ways. And I think that gets to a broader point of like when healthcare goes virtual, you don't think about just like taking the way that things work in the physical world and then just like translating it to a virtual version of that. But you can sort of like reinvent from the ground up the actual operating model of how that works. Today's healthcare system is the patient has to have like everything is optimized for the provider's schedule. Even the notion of like pre-booking an appointment and making it work for us, us meaning patients going out of our way to accommodate the schedules on the supply side. You could entirely flip that on its head,

especially if you think about a world in which you're continuously monitoring patients. It's not the patient sort of determining that he or she needs to go see a doctor, but rather the data saying, hey, this patient needs intervention and actually having the provider side of the market, you know, reach out to the consumer side. So I think there's like lots of opportunity there to make it much more patient-centric as well as much more proactive so that it's not, again, it's not up to the burden is not on the patient to have to figure out when it's appropriate to go in. So it's not just you reaching out to your doctor to get virtual care. It's the virtual care reaching out to you when you need it and you may not even realize it. Yeah. The whole notion of like

provider networks and even like what is a provider sort of changes fundamentally where this is also potentially a cultural shift where in order to do like really intelligent load balancing, it might be the case that you're not necessarily going to have an established long-term relationship with like a single human being, but kind of more of a care team. And this is a model that's been talked about, you know, for a while, this notion of like the medical home or medical neighborhoods where you have more of a care team model and therefore you're not constrained by any one individual player in the system, but rather can tap into multiple resources. And that's, that's, I think that'll be a big cultural change, at least here in the

US. It's funny because, you know, we do, I, you know, I have a dermatologist and I have like, I don't know, dentist and I have a PCP. But it does seem that the way you develop a relationship with one PCP and you sort of assume they know you and they're looking out for the 360 degrees of you, whether or not that's actually true. That is like a very, that does feel like an important cultural an emotional thing in this particular culture. Is there a way to do both in the virtual? I mean, we've seen this in other aspects of our lives, right? Like, you can still have a phenomenal customer experience

when you have like really good CRM and just really good 360 data on who you are as a consumer and like retail and, you know, like every time I interact with an airline, like they know my whole history and all that kind of stuff. I mean, we've seen this in O.B. you know, I think these days most, at least the larger hospitals and larger OB clinics, knowing that it might not be your OB who's actually delivering you based on, you know, when you go into labor, they try to actually introduce you to the entire care team as part of

the prenatal experience so that no matter who ends up being there, like, at what, at, during game time, so to speak, you're going to have at least some established relationship with them. I think it's more about like, how are you setting expectations to the patient? upfront and giving them the room to actually meet everyone in a low risk way, such that when stuff hits the van, you have that preexisting relationship. But it does. And even in your example with the OBs, it sort of reminds me of like, okay, so yeah, I sort of knew all along that whoever I got at the hospital was going to be the person

I got at the hospital. And in a way, that was a totally separate event from my pregnancy and, like, tracking me through the pregnancy and understanding what was going on with me. you know, in a way, it's almost unbundling the experience, right? Yep. You know, you could take this virtual idea even one step further, whether it's like a really good chatbot, a chat bot that could answer sort of standard questions

of chatbot that knows your history and connect to that and naturally would and could escalate to a human being and beyond. That gets really interesting in terms of scale. And the question is, in my mind, can that type of service, you know, what fraction of the, of what a doctor does could be serviced by something like that. Obviously, there's a lot that couldn't. But in terms of just having that with you at any time that you could just asynchronously connect with to ask any sort of health question and with the sort of knowledge that the answers

are completely accurate and so on, which is a very high bar to make sure that way we reach, that's a whole other direction. And you can see how that just gets smarter and smarter as time goes on. You guys have heard me talk about Baymax from the movie Big Hero 6. Like, you're all going to have a Baymax at some point. Japan is like way. ahead on this. So if you actually look at like the landscape of companion robots in Japan, they actually have like pet robots, right? And in some ways, like a lot of why the promise of like humanoid robots has kind of, you know, fallen short is like we have such high expectations for the level of intelligence that those quote unquote human robots have. Whereas like if it's a

pet, you know, your bar goes down. But there's still like a tremendous amount of therapeutic benefit to having that kind of companion. And so like especially with the silver tsunami and the elderly population and what have you. There's like a pretty meaty set of things that you could do in a very basic form that are not clinical in nature that don't require clinical judgment that would still hugely benefit the system, both in terms of just like reducing anxiety, right, at a basic level. Yeah. Wasn't there some kind of study where they had baby seal robots in elder care facility for the, for the, no, they did for for the empathy and the comfort as a particular clinical need in that setting? So if we're at the tip of the tip.

point. If this moment is sort of the tipping point for for virtual medicine, two questions. One, why did it take a pandemic to get here? And then two, what do you think is the most, the sort of most immediate near-term things that we're going to see start rolling out right now, as, you know, not just video chat, but are there other things we're going to start seeing today, tomorrow, with COVID going on that we're going to all be getting more familiar with? It's unfortunate that this had to be the forcing event to sort of bring all of this to light, but there are a number of tailwinds that have been in motion that enabled us to actually respond in a way that's reasonable in light of this kind of crisis, which definitely would not have been the case, I would say,

you know, five or six years ago, like just like the very visceral understanding that costs are spiraling out of control and the way that we deliver medicine in the physical world today is just not sustainable to patients just being at their width's end with regards to access and convenience and therefore being willing to adopt these types of novel technologies, combined with what we talked about earlier, in other parts of our lives, we are now getting much more comfortable with the notion of either asynchronous communication or video-based communication. And then now, like, the actual virtual care platform technology is mature enough to actually be delivered at scale. I was like one of the very early adopters

of some of the early telehealth solutions. And it was super choppy. Like, you know, the, audio quality was bad and it was just not a smooth experience, but if you do it today, it's very, very streamlined. So I think the confidence of all those things, like, had to be in place such that we could respond in a situation like this in the way that we are. But do you think that we would be doing it without something like this to push us over into it? Do you think it would have taken longer? I think it would have just taken longer. I think it's, the forcing function is not just like adoption, but it's also, again, like, there's top-down regulatory change that's enabling reimbursement. There's, you know, I hope more relaxing of the regulation around

like medical licensure. And I think we will see like in the next year a tremendous uptick in adoption by at least the enterprise side of the market for access to virtual care services, which has always been like an emerging area. It's definitely gotten a lot of early uptake. But, you know, this could be the thing that pushes it into the mainstream. The UIs for these things are kind of clunky, especially in a world where people have like Google and Facebook and things where these consumer products have really elegant UIs. And it's clear of the utility, if we can get people to use them and be comfortable with it. I'm just imagining while we're talking like a Facebook-like feed where I'm chatting with,

you know, my various doctors and everything's in there and like my records are there and these are coming up as posts and I can just look through it. Maybe I can even look through my kids feed to see how their medicine is going. that really wouldn't be that hard to do in principle and practice, you know, UIs and art and so on. But I think if we can force the tools to sort of come up to speed with what people's expectations are, I have a feeling the engagement could be comparable to engagement in other sort of consumer-like products. And it's interesting because it seems to me like right now, you know, the sort of incentive to do so is aligned on both sides, right? Like nobody actually wants to go into their doctor's office right now.

Doctors also don't want you to go into the doctor's office right now. Like, it's unusual for everybody to be aligned in that way. We're all incentivized to you something like this at the same time. Well, I'll add one more thing, which is that these difficult times often create some of the most exciting startups. And that we have this combination now for sort of the fire to do that combined with all of these things just in our face. You don't need a world pandemic for it to be an individual crisis. and what can we do to handle my crisis in the future? Hopefully this will give us a model for that.

A whole new set of tools. Virtual care is actually in some ways like the perfect solution at a time when we need lower cost ways to deliver care because the actual way by which you can just eliminate so much of the cost structure of the traditional healthcare system. Thank you so much for joining us both on the A16Z podcast goes remote. Thank you. Stay healthy.