On Friday, WIRED’s editor in chief, Nicholas Thompson, hosted a Facebook Live with ER doctor turned cofounder Caesar Djavaherian, who now serves as the chief medical officer of his telehealth company, Carbon Health. This was the first in a series of four conversations presented by Salesforce in which WIRED will explore what the coronavirus pandemic means for the future of business, education, technology, and health. Hundreds of readers tuned in, and we took questions from viewers in real time. The conversation has been lightly edited for clarity.
Nicholas Thompson: Hello, I'm Nicholas Thompson, and I'm the editor in chief of WIRED. Today we're going to be talking about the future of telemedicine with Caesar Djavaherian.
Caesar is an emergency room physician, has been for many, many years. He's also one of the founders of Carbon Health, which is building one of the first at-home coronavirus tests. He's been at the front lines of combating the coronavirus since the very beginning in the Bay Area. He's on the front lines of helping to build the technology to fight it. He's on the front lines of figuring out the future of telemedicine.
So, Caesar, good morning.
Caesar Djavaherian: Good morning. Thanks for having me.
NT: Let's start with a little bit about you. You started as an emergency room physician. And at some point a couple years ago, you made the decision that the most important thing you could do is try to make medicine more efficient, particularly through telemedicine. Tell me about that choice and a little bit about that transition.
CD: I started out as an ER doctor. I actually trained in New York City, where the pandemic is hitting hardest. And just like many doctors who practice medicine, I became a little bit disenchanted with all of the administrative burdens that physicians have to deal with in their everyday lives. And I wasn't satisfied with the answers to my questions about why we do things this way. I was told, “Well, we always do it this way,” or “This is the way we've traditionally approached X type of health care.” And I thought, you know, in 2013, 2014, 2015, with incredible technology developing in every other industry, why couldn't we take a step back on how health care was being delivered? And really ask ourselves, if we were to create a brand-new health care system today, what would that look like?
And part of it is that you can do so much in person. But there's a lot you can do actually online, outside of the clinics. And being able to take a step back from it, and asking that question, and really trying to strip down health care to its bare bones, almost like, you know, how Elon Musk talks about first principles. So what are the first principles in health care? It's really a provider and a patient and some work that has to happen around that interaction. And once you can break it down to those bare bones, you can then start to build technology that can enable that experience to be much better from the patient perspective, and, frankly, much better from the doctor perspective.
NT: So, you're trying to beat Elon Musk without the tweet storms, right?
CD: Well, I’m not very good at tweeting.
NT: We have a quick first question, which is ‘What is telemedicine?’
CD: Ah. So, you know, the definition of telemedicine is fairly broad. It's anything from taking care of a patient's needs through the telephone—so just the old fashioned way—to having video interactions with your patients, to actually having video interactions with your patients assisted with devices. So whether it's loading information from your Apple Watch, or using a remote stethoscope like Eko Health has built or some of the at-home products like blood pressure cuffs, connected scales, that sort of thing. And there are lots of companies that have gone into this field under the umbrella of telemedicine. But frankly, it means different things to different people.
NT: Ever since the internet was invented, I've been hearing stories about how the internet will allow doctors in rural communities to treat many people. But it hasn't taken off, really, maybe until now. And the differences are, as I understand it, we have much better bandwidth, we have better computers, we’ve got better cameras, we have Zoom, we have wearable devices, which give us more data. What else do we have? What else do we need?
CD: Oh, it's a good question. So first of all, big shout-out to Zoom, because the video quality has gotten much better under the Zoom platform for telemedicine providers. I think without trying it, most patients didn't think they’d be able to get their questions answered. So they've maybe tried telemedicine as an early adopter, tried to onboard, connect with their providers. And what I've heard from patients, actually, is that that was an incredibly frustrating experience for them. They'd much rather just go to a place where they know they can get care. And unfortunately for a lot of them, that meant the emergency department, and that's where I would see them. So patients were showing up with presentations that really could be taken care of at home through telemedicine or, frankly, wait till the next day to see their primary care doctors. And so, one thing that's in place, to your question, is that patients need to try it. And I think this pandemic has really forced a lot of us to try telemedicine for the first time.
And then what we need more of is to have a broader list of actual ailments that we can take care of virtually. So today, if you're a young woman with urinary tract infection symptoms, that's a perfect case for telemedicine. Now, if you have back pain and fever and vomiting, that's probably not the right case for telemedicine. It might be in the future, depending on the connected devices that patients have in their homes.
NT: And so then, with Covid-19, there are a whole bunch of interesting steps where telemedicine can play a role and a bunch of steps where it can't. So the initial question of: Should I go see a doctor? Should I get a test? That not only can be done via telemedicine, it should be done on telemedicine. Lots of people who weren't positive have gotten infected by going into a hospital to see whether they should get tested. So consultations, definitely. Testing you can't do via telemedicine, but you are working on an at-home test, which I want to talk about for a second. And then there are other stages like treatments—you can't have a ventilator run in your home. So is it the right framework, that with every kind of illness, there are steps where telemedicine can be appropriate and steps where it's not?
CD: Yeah, well, I think we shouldn't try to fit a clinical scenario into the telemedicine box and just try to use telemedicine because it's there or because we want to. The health care system is incredibly complex, and the number of different patient presentations is incredibly diverse. We should use telemedicine in areas where telemedicine works and is effective and can resolve a patient's problems with the same clinical standards as an in-person visit.
But then there are instances where having a patient come into the clinic or into the hospital is more appropriate. And frankly, that's really the angle that Carbon Health has taken, which is that, yes, we do have telemedicine video visits. We can do a lot through that, but we can then connect them into our own clinics, or into a partner hospital, so that from the patient perspective, they're getting the best care possible for every instance. You’re absolutely right, there are telemedicine companies that do monitoring of patients in the intensive care unit so that one doctor can actually see multiple patients at the same time through their monitors. They can look at the vital signs, make recommendations to the nurses, but they can’t perform procedures, they can't put a patient on a ventilator when they need it. So there are limitations out there. And I think, again, going back to some first principles, we know that this is a tool in the toolbox, but we can't have it siloed off from the rest of the health care ecosystem, because we know that there are too many things that need to happen for a patient to be well cared for. And this pandemic has highlighted all of those.
So in this pandemic, you have a disease that is deadly in certain populations. If you're male, 70 to 80 years old, with diabetes and heart condition, this is a very deadly disease. If you're a younger woman, you might just have very mild symptoms, like not being able to taste or smell.
And the reason that telemedicine kind of rose to prominence within this pandemic is that we needed to very efficiently risk-stratify a huge percentage of the population, not just in the US but around the world. And so you can see that for some people, having information about coronavirus was sufficient. For other people, being on a ventilator was the care that they needed. So how do you organize all that? Well, we think through technology you can inform those who need the information and keep them away from the health care system. For a subset of those people, you can provide some care, whether it's symptom control, or guide them to the right testing facility or the right clinics. And then for a subset of those people, they'll need more advanced care and you want them to go to the emergency department or the hospital, but you want them to get there having informed the staff before they arrive so that they can be protected when the patient hits the door.
Now, I know in New York the assumption is that nearly everyone has coronavirus, and that's clearly an unfortunate event. In the rest of the country, the frontline providers mostly assume that patients don't have coronavirus. And so when the patient hits the door, if they're given the heads-up that this patient is at risk for it, they can better protect themselves, they can be more on guard and prevent transmission further, not just to the health care workers but then to their families and the rest of the community.
NT: An amazing question came in that is so dark and cynical that I wish I had thought of it myself. The question is about the business model of telemedicine. Clearly, one business model is: I call Caesar and health insurance pays Caesar. Or I pay Caesar directly, if you give me a consultation or some treatment. Somebody asked: Is there a business model where pharmaceutical companies can insert ads during a telemedicine consultation? Is that something that has come up or that you have seen?
CD: You know, Practice Fusion was a health tech startup that created an electronic health record that doctors could use for free. But in exchange for that free software, they would show advertisements for pharmaceutical products. And that company, unfortunately, I think blurred the lines of ethics and clearly has gotten in trouble with what they've done. There's a recent penalty that the company ended up having to pay. I haven't seen that same model in telemedicine, and I hope I never do. The idea that you can monetize and profit from a doctor-patient relationship is, frankly, disconcerting. And clearly it's not at all within our business model, nor in any of the current telemedicine companies’ business models that I've seen. I think that the industry hopefully has learned a big lesson from the rise and fall of Practice Fusion.
NT: So let's talk about a product that you have, which is the at-home coronavirus test. You prepared it, you had some interesting scientific way of doing it that wasn't being done by others. The FDA said, “Wait, you can't do this right now.” You're in the process of negotiating with the FDA. I won't ask you to talk about that. But tell me how your test worked, what was different, and why it was optimized for personal at-home testing.
CD: So we had our first coronavirus patient come into our clinics, or a highly suspicious coronavirus patient, on January 23, before all of this started to happen. And from day one, we mobilized our company to try to address the needs. We also realized that effective testing wasn't available even to public health officials. So in retrospect, now that I'm reading more reports about it, we were unfortunately not prepared for this pandemic, even though we had seen other countries go through it. So the idea was: OK, we have a disease that's highly transmissible, that's putting not just our health care workers at risk but also using up all of our resources when it comes to personal protective equipment, and also has put a lot of strain on the supply chain for just the basic testing components.
We also realized that the test itself is frankly old. You know, when I was in college, I used the RT-PCR technique in my research endeavors, and that was 25 years ago. So, again, going back to the first principles, what is it that we need to get accomplished? And what are the different ways we can accomplish it?
The at-home test doesn't rely on the typical supply chain. It uses a different type of swab. And the idea of being able to administer the test within the patient's house, where they're not at risk of infecting others, was incredibly compelling to us. So we worked with our lab partner to identify a workflow where patients who are at risk for coronavirus infection would be evaluated by one of our doctors at scale. We can look at hundreds, if not thousands, if not tens of thousands of patients’ symptoms at the same time. And just like you’re analyzing something in a dashboard, you can identify who would benefit from an at-home test. And the doctor can go through those patient responses, order the tests appropriately, have the test delivered to the house. The patient then self-swabs—and the swab that we're using is actually a buccal mucosa, so it takes saliva from the inside of the cheeks and the gums, and the patient would put it back into this tube and ship it back through the US Postal Service to our lab partner for analysis. We would get those results in 24 to 72 hours, and then communicate to them again, and then initiate further video visits with the patients who are positive. We can focus on communities that are being hit hardest, without having to put our health care workers at risk.
So we were working within FDA guidelines when we launched the test. We immediately stopped when the FDA updated its guidelines around the testing, and we are continuing to work with the FDA to obtain approval, because the officials have said publicly that they support the idea of home testing. They want more data around it. And frankly, we're hoping that between the FDA and the local governments in each state, that we can get this test approved, because we know that it performs incredibly well compared to the standard that's out there.
NT: How do you know that? I mean, you know that the test is effective because you've tested it. But you haven't tested the process where people actually have to put it in the tube and put it in the mail, and they forget to put stamps in it, or their kid puts a peanut butter and jelly sandwich on it, right? Like all the things that happen at home.
CD: That’s a great point. And frankly, that clinical study hasn't been done with the at-home test. But it hasn't been done with any of the existing tests in this pandemic. So when you ask your doctor, you're going to run a nasopharyngeal—so you're going to put a swab through my nose into the back of my throat—and you're going to send it off to Labcorp or Quest, what's the sensitivity? How do I know how accurate this is? You won't get a straight answer. And the reason is that clinical data don't exist for any of these tests. So the real-world data doesn't exist, whether it's for the at-home test or for the commercially available testing.
We do know that, to your point, that the process of extracting the DNA and getting an appropriate sample for the RT-PCR machine is highly effective and replicable. But how good are people at swabbing their own mouths and noses and putting it in the mail? We do have data that we've shared with the FDA. And so we're optimistic that it is an effective way of doing it.
NT: We have a good question here. What do you view as the biggest bottleneck in telemedicine that needs to be urgently resolved?
CD: So my concern, not just with telemedicine, but with health care in general, is that the regulatory network is an old one. If you're a doctor and you graduated from a medical school in the US, and you've gone to residency in the US, and New York state gives you a license to practice medicine—and then you need to practice medicine in Pennsylvania or Connecticut or New Jersey, you actually have to go through the same six-, eight-, or nine-month process to get approval to practice medicine in that other state. It makes zero sense. When you get on Metro North in Manhattan and you come out in Greenwich, your DNA doesn't change, your body doesn't change at all. You're still a human being, but the doctor that could treat you in New York can't treat you in Connecticut. And that's a problem. And telemedicine faces the exact same problem, where a provider who is licensed in California can't provide services in Idaho.
And I know there's a concern about, for example, treating patients in underserved communities or rural communities where health care providers don't necessarily live, because there isn't a population density that's large enough. Well, we need to take a step back and ask ourselves, “What are these regulations good for?” Is New York state that much better at figuring out whether I'm a good doctor than California is? Let's identify the state in the country that's the hardest to get acceptance in, and let's say that if you get accepted in North Carolina, then you can be a doctor anywhere in the country. That's what we're looking for.
And having parity with the telemedicine visit, meaning if you're seen via telemedicine or in-clinic, insurance companies should pay roughly the same amount. I would say that is also an important barrier, because, currently, if you look at incentives for providers, it's to ask the patient to come into the office, because they won't get paid otherwise.
NT: They get paid zero for a telemedicine consultation? Or they get paid half, they get paid a quarter?
CD: Yeah, it's state by state, and it's insurance company by insurance company. So you have to read your plan to find out what your benefits are.
NT: But what's the range? On average, it's 10 percent, or on average it's 92 percent?
CD: I would say that, on average, it's zero. Unless you have a specific kind of telemedicine clause in your benefits. And, well, this is why what Trump said, you know, early on, which was, there's now parity. You can see a Medicare patient who you've never seen before, so you haven't established with that patient, you can see them and you can get paid for it. That was novel. But what does it do? Again, what are we trying to achieve here? What we're trying to achieve is that a patient gets care. If it's appropriate for telemedicine, it's appropriate for telemedicine. If it's inappropriate for telemedicine, the doctor shouldn't be providing that care through telemedicine. The payment should be secondary.
And yet, we've gated it with these regulations and with these insurance contracts, and states started to, in 2018, say, “Well, we want parity between telemedicine and in-clinic visits.” However, there's no stick to that. So the caveats were if you had an established patient, you can then, next time, see them via telemedicine. You needed to do a video visit versus just an asynchronous visit where the patient fills out their information, the provider looks at it later on. There were all these nuances to it, where in the practical world, it didn't make telemedicine a viable option for many. And it incentivized doctors to say, “Oh, you want that prescription refill for the blood pressure medicine that you've been on for years? You have to come into the clinic for me to see you.” If you've been on the receiving end of that comment, you now know why. It's because your provider is not incentivized to take care of you remotely and say, “I understand that it's hard for you to come in to get that prescription refill. I know you need it. I'd like you to send me your latest blood pressure from the home blood pressure monitor that you took the other day. And as long as it's in range, I'm going to refill your medicine.”
NT: Ok, so I'm just going to restate for everyone in the audience that this is absolutely insane. If there are any congressmen, governors, legislative aides listening: If doctors do telemedicine consultations, they should be paid. And I agree with Caesar, if I go to see you and you give me good advice about something that can be handled via telemedicine, you should be paid. If you try to instruct me on performing my own surgery with a saw, that's inappropriate, you should not be paid. So I think that's where we should be going.
We have a whole slew of really great questions. So one of them is: What about telemedicine in prisons? I’d like to broaden that out a little bit, but that's an excellent question. Where are the populations, or surprising areas, where telemedicine is particularly appropriate?
CD: The prison population is one that’s ripe for telemedicine. I work in an emergency department that's very close to a state prison as well as the county jail. I think also rural communities that may not have access to a dermatologist or a specialist—telemedicine is perfect for those fields, especially when it comes to things like stroke care. Most of the decisions made around stroke care can be done remotely. And it's a service that can really change someone's life.
So the prison population is a great example. The local jails will contract with medical groups that just do jail care. And that's a very limited amount of payment for mostly chronic issues, and they'll send out to the local emergency departments for the more acute issues.
If you know of local uses for telemedicine, and you're wondering, why haven't they been adopted, I would say look at the payment scheme, and look at the incentives for that system and how money is allocated. And I guarantee you that there's waste there. It's one of the frustrating parts of being in the field; you see a ton of waste. You see an incredible amount of press around the cost of health care, and yet very few people are doing anything innovative about it, to change it. There's just too much money on the other side of that equation.
NT: I'm going to read something from a viewer. This is something I think you're going to agree with, because it maps closely to something you just said. This is what Rich says: “I have been using telemedicine for 10 years. Regulations and lack of parity from payers are what's been holding it back, not the technology. Can't be used for everything, no. But patients love it and are not going to tolerate it going away after Covid-19. Payers need to pay for it ongoing, and restrictions across state lines need to go away. My patients live in two states and travel all over. I should be able to help all of them via telemedicine, not just the state where I'm licensed.” Godspeed, Rich. Thank you for that excellent comment.
Now I want to go back to something you said. You were talking about telemedicine in rural areas, which brings me to a question from three viewers on a similar topic. And that is basically: In rural areas, there's a need for telemedicine, but sometimes there isn't broadband. What can you do?
CD: Well, I'm hoping 5G and the additional capabilities will be there eventually. I know that the T-Mobile/Sprint merger happened partially on the basis of providing broadband to these rural communities. I've been able to do telemedicine consultations through my cell phone on a soccer field when there's been an urgent condition that a patient needed to be seen for. So I think we're less reliant on broadband issues than we were a few years ago.
I would say also that there's a subset of telemedicine called asynchronous telemedicine, where a patient can input some information about what they're going through. So, for instance: My age is 25, I'm female, no other medical problems, and I'm having burning when I urinate. What should I do? So you input that information as the patient, and your doctor at some nonsynchronous time later on reads that information and says, “Sounds like you have a urinary tract infection. You should have an antibiotic sent to your pharmacy.” And does so. And that allows you to not have the best broadband service but some sort of access to the internet. It allows your provider to see many patients at the same time. And it's evidence based. It turns out that in the clinical scenario I just gave you, having a urine test and a urine culture doesn't affect your treatment of the patient, as we once thought it did. And so it's better for the patient, because it gets them treatment before the infection gets to their kidneys, and it's better for the provider because now they can see more patients at the same time and do so efficiently. And then spend their time following up and instructing the patient on their condition, rather than the administration of the health care.
NT: Okay, we have a question from Facebook: With Carbon Health telemedicine and Covid-19 testing, how much would it cost to get at-home tested once it's approved by the FDA?
CD: That’s an awesome question. So I think I mentioned that on January 23, we had our first patient in our clinics with what was suspected to be coronavirus. We had two patients, from Wuhan city, tachycardic and with fever. From that moment on, we rallied the entire company, and part of rallying the company was: What can we contribute to this pandemic? And part of what we're contributing is our doctor services for free, for the at-home test. So the only cost to patients will be the cost of the test itself and the shipping. We've gotten the payments down to around $167 for the at-home test. The idea is that if our politicians are being straightforward with us, and that these tests, which should be covered by insurance, will be free to patients, the doctor visit which Carbon Health provides, we won't charge for.
And so we're trying to do everything we can to get widespread testing to patients to follow the same trajectory as, you know, South Korea and Germany have done. We understand that payment is a problem. So we've made our contribution to the cause.
NT: Let's look at your crystal ball about the future of telemedicine. As with many other things in the world, coronavirus has accelerated trends that existed before, like working from home and communicating via video conference. Look into the future five years from now, I would imagine it'll be much more telemedicine for certain kinds of treatment. I would imagine you would then need differently designed hospitals, because you need less space for consultations and maybe a higher percentage of space for invasive treatments. So you need to redesign hospitals. I would imagine that there will be different kinds of doctors who succeed—you know, a person who's very good at telemedicine, is very good at communicating through Zoom and has good lighting, as you do Caesar, will have an advantage versus some of the soft skills you have for in-person consultations. I would imagine that the number of times people go to the hospital for no consequence will go down. What else is going to happen because of telemedicine?
CD: Well, you know, my prediction is that costs will go down as well. At Carbon Health we've been hyper-focused on access plus cost. So if you create a better patient experience and a better doctor experience and use technology to make that experience more seamless, and then take away the administrative burden, what that does is it reduces the cost of care. And we've seen that in our company, in our clinics. For example, we now have mental health services, we have pediatric services, we do primary care through our virtual platform.
And so exactly what you said is what I believe too. And the reason that you've come to that conclusion, I've come to that conclusion, is that it just seems like it makes sense that whatever we can take care of through great technology, and remotely, we should. And those things that you can't do through telemedicine or technology should happen in-person. And that's kind of what clinics will look like. And that's what hospitals will look like—they’ll be for procedures and more acutely ill patients, and then we'll reduce cost.
I would say one caveat, though, is that it's not in your hands or mine what that future looks like. It's really going to end up being how much your employer demands that type of care to the insurance companies who decide which providers get paid for the care that they provide. So I think that one of the caveats in health care that's not obvious is that even if you create the most amazing product in the world, patients don't necessarily come unless that product is approved by a gatekeeper (the insurance company or the government). So that's the one caveat that's frustrating. But I think there is an opportunity in this pandemic, where if you're an insurance executive or an employer who does self-insurance or the government, it's hard to argue that, “Oh, technology is not useful in health care.”
NT: And then, digital disruption often sounds, in theory, fantastic and wonderful. And there are huge benefits. You can look at different industries: the music industry, where we had digital disruption and Spotify is a great way to listen to music, but it drove a lot of record labels out of business, drove some bands out of business, made more people tour. We all know those effects. Digital disruption has been great for journalism. Look at what we’re doing. We have Facebook distributing the conversation that WIRED is having. That’s so cool. But it’s also changed the advertising market, which is complicated. Journalism has changed dramatically. So one of the things we know about digital disruption is that it turns things upside down, in ways that are very hard to predict. So with that premise, give me some more predictions about what it does to hospitals, insurers, doctors.
CD: I think digital disruption has already happened in health care, but in the exact opposite way as it has in other industries. When you look at time spent, the doctor’s time, and how we spend it during the day, frankly most of it is in front of the computer. The numbers are staggering. It’s up to 150 percent of the time that you spend with your patient; you actually spend 150 percent of that time documenting the visit. And of course that depends on what specialty you’re in, but the point is that doctors have become really attached to, burdened by the administrative needs of documenting the visit. So typing out: Nick Thompson, male, came in for XYZ reasons. That takes time and effort, so now you’re spending one of the highest-paid profession’s time on documentation. So the digital disruption has happened, and it’s led to this terrible, expensive health care system that you see today.
Now, for version 2.0 or 3.0, whatever it is in this life cycle, I think that can change by having smarter technology in play. So at Carbon Health, we look at how much time the doctors spend documenting, how much time they spend with patients, how much time they spend after their shift is over documenting. Typically, if you look at Epic Systems, which is a very well run company, it has software in many of the hospitals across the country, if not the world. What happens is that typically there’s a peak in log-ins during hospital hours, and then there’s a lull around dinner time, and then there’s another peak late at night. And what’s happening is that the doctors are spending time with their families, then logging back in and finishing up their work that they started during the daytime. That is incredibly disruptive to the doctor-patient relationship, to job satisfaction for providers, etc. When we look at our own numbers at Carbon Health, we see that typically there’s about a 15-30 minute period of time after the shift is over, where the provider finishes up all their charts, and then there's no more log-in until the next day. And that to us is success. And hopefully there are more and more companies like ours who can use technology in a smart fashion—to disrupt the disruption, frankly—and get us back to why people always wanted to go into medicine in the first place, which is: I love spending time with my patients, I love getting to know people, understanding what they need and trying to meet those needs. Rather than: I spent some time with my patients and then I spent a lot of time with my computer to document everything.
NT: Ok, we’re going to wrap it up here. Thank you so much for everybody who joined in. Thank you for all those fantastic questions that came in through Zoom, Facebook, other channels. Thank you to our audience and to Caesar. We’ll see you at the next one.
- What if it returns every year, like the common cold?
- “Here in spirit”: an oral history of faith amid the pandemic
- We need a vaccine—let’s get it right the first time
- Un-miracle drugs could help tame the pandemic
- WIRED Q&A: We are in the midst of the outbreak. Now what?
- Read all of our coronavirus coverage here
