Listen and followApple PodcastsSpotifyYouTubeMore apps · Browse all episodes
A patient with a urinary tract infection calls at 9 p.m. and gets told to call back in the morning. Frustrated patients default to the emergency room instead. Tod Stillson is a family physician, health care entrepreneur, and founder and CEO of ChatRx. This episode is based on his article “Why medicine is late to asynchronous telehealth,” published on KevinMD. You will hear why traditional telemedicine did not close this gap, because it is still appointment-based and still synchronous. You will hear how orchestrated AI agents and standing orders can tee up a low-acuity case so the decision itself takes seconds, why he calls that augmented clinical automation rather than autonomous AI, and what happens when he reviews a prescription the next morning and disagrees with it. You will also hear where he draws the line between this and the marketing funnels that route patients to doctors who rubber-stamp prescriptions. Press play for the case he makes for asynchronous care with a physician still in the loop, and the conditions he says never belong there.
Partner with me on the KevinMD platform. With over three million monthly readers and half a million social media followers, I give you direct access to the doctors and patients who matter most. Whether you need a sponsored article, email campaign, video interview, or a spot right here on the podcast, I offer the trusted space your brand deserves to be heard. Let’s work together to tell your story.
PARTNER WITH KEVINMD → https://kevinmd.com/influencer
SUBSCRIBE TO THE PODCAST → https://www.kevinmd.com/podcast
Transcript
Kevin Pho: Hi, and welcome to the show. Subscribe at KevinMD.com/podcast. Today we welcome back Tod Stillson, family physician, health care entrepreneur, and founder of ChatRx. Today’s KevinMD article is “Why medicine is late to asynchronous telehealth.” Tod, welcome back to the show.
Tod Stillson: Kevin, it’s great to be back with you and your audience. I love this ongoing dialog that is really happening in the world of medicine, talking about different ways we can deliver health care, because the world is changing, and I’m excited to talk to you about that in the context of asynchronous care today.
Kevin Pho: All right. So, why did you decide to share this particular facet in your article?
Tod Stillson: Well, first of all, because I think it’s a relatively new component to the delivery of care for physicians, and it’s evolving and growing, right? And so whether a physician is in a health system and is seeing their message box grow, because now patients have ongoing access to them and can throw any question they have in the box, so to speak, or whether it be what we’re seeing happening with other industries in our own experience, when it comes to banking, when it comes to travel, when it comes to food delivery, we’re seeing this growth of asynchronous care.
And then even pull forward the idea of remote patient monitoring, right? Patients, from the Oura Ring to their watch, are getting all this data, and they’re sort of wanting to vet it through their physician, you know, to help interpret it. So we’re getting all of these pieces of information that are coming at us as doctors asynchronously, not in real time, not in the exam room, but we’re having to figure out how we’re going to work through that and what we’re going to do with it. And I think this is a conversation point for physicians, but it also brings forth the opportunity for us to think about how we might deliver care in that same context.
Real physician voices, twice a week
Free, and one click to unsubscribe.
Kevin Pho: So, in your article, you start off with a patient who called at 9 p.m. with a urinary tract infection. So talk about how this case launches into some of the other care delivery options that we’re talking about today.
Tod Stillson: Yeah. So I think everybody in the audience who’s a physician, especially who’s been in primary care, knows this story, right? Whether you’re on call, or the patient tells you the next day, or you see them at Walmart: “Hey, I called, and I had a UTI at 9:00 at night,” right? In our system of health care, which doesn’t cover after-hours and weekends very well, it usually ends up with this kind of call-center notion. Sometimes that gets filtered to us, sometimes, but for the patients, it feels like it gets kind of blocked. They sort of get into this never-ending tree of “What do I do next?” And at the end of the day, what patients end up doing when they get frustrated by this, or have not a good navigational experience, is they default to what’s available 24/7: the ER, OK? And that’s just not the right level of care for these common conditions. The patients are put in this very costly, time-consuming, and really kind of wrong location for care, but they just don’t know what to do, and they default to that place.
And so, you know, having been in primary care for thirty years, Kevin, and you’re in primary care, you know what happens. Say that patient call got through to you as a doctor. Our group always had a policy: We didn’t call in antibiotics on call, because we didn’t know the patient; we didn’t know their context. We just kept that policy in place. And so the patient would be like, “Well, I know I have a UTI. What do you want me to do, then?” And my answer would always be, “Well, call the office first thing in the morning.” And the reply would always be, “But I can’t get in when I call at 8:00 in the morning.” And you and I both know that’s how that goes down, right?
And so I would just kind of leave them in that lurch of, “Well, call the office first thing and hope you get an appointment.” And then the patient’s just back in the same stuck box. So there are a lot of low-acuity conditions that we as physicians encounter all the time, and patients encounter, and our current appointment-based system, which is synchronous in nature and very scheduled in terms of delivery time, just doesn’t accommodate those low-acuity conditions that we all get.
And so I think this is again part of this conversation point, where we start to think about how we can provide care to patients not just in that box of synchronous care in an appointment-based system, but in something outside of it. We’re going to help improve their quality of life. We’re also potentially going to improve our quality of life, if we lean into technology, because quite honestly, you’re not going to get that phone call at 10:00, which is going to disrupt your life at the time. So we have the opportunity to develop systems and technology that support both parties, the doctor and the patient, in that context.
So what could that look like in the future? The patient at 9 p.m. gets into a software-based program that has a physician in the middle, maybe your own practice. It uses AI to guide the patient through a conversation. It identifies a diagnosis and perhaps even asynchronously sends it to you to act on, and you document that and send in a prescription for the patient within, you know, ten, fifteen minutes. You don’t have to wait to see the patient in the morning, you don’t have to wait to put them in a queue, and it doesn’t gum up your schedule, because the next day, when you have those thirty patients in your schedule that are high-acuity patients who do need your time, now you’ve got the space to see them. So this is the kind of challenge that doctors face, right? What to do with these low-acuity conditions.
Kevin Pho: Yeah. And it’s not just doctors. I think, like you said, the medical system is so antiquated. That scenario you described, that patient calling at 9 p.m. They have to call the next morning and hope that there’s going to be an appointment available. And I did a podcast last week where we contrasted that particular scenario with pretty much everything else in our society. If you want a meal, right, you call DoorDash. If you want a ride, you call Uber, right? And if you want some type of medical appointment or interaction with the health care system, there’s really very little way that we can do that now.
Tod Stillson: That’s right. That’s right. And so here’s what the doctors are going to say, and we know the system has sort of adjusted a little bit. What they’re going to say is, “Well, what about virtual care, right? Well, now we have this telemedicine opportunity, right?” And that’s a great solution. But what I want to remind our doctors listening to this podcast is the following: Telemedicine, what I call telemedicine 2.0, or traditional telemedicine, is still appointment-based. It’s still synchronous in many cases, and it still puts the patient in the same bucket of trying to find some availability and time. Even though it separated things so that it’s not where you’re located or they’re located that’s required, it still has some of the very essence of what synchronous, appointment-based care does. And when we now layer in asynchronous care, Kevin, just like those other industries have done in terms of their asynchronous communication and interaction, it brings a whole other opportunity level for us physicians to deliver care.
Kevin Pho: So let’s go deeper into this particular example. Let’s say they have that tool, some AI-based tool that goes through their symptoms, asks them questions, and decides, “Hey, this patient may need an antibiotic for a UTI.” Tell us what happens next. There’s going to be, I presume, a clinician in the loop who actually does that prescription. So how does that work?
Tod Stillson: Yeah. So to disclose to you, I have a company that does this, and so that is something I’ve engineered and orchestrated, and we are currently doing it in the marketplace. So what does that look like? We have the AI kind of replicating what happened in my clinic for years, and that is, my staff would organize the visit so that by the time the patient and I got in the room, I would be in the sweet spot of my medical decision-making, right? I need to make a decision about what’s going on, and I’m the one responsible. My nurses and my staff aren’t. But if you do it well and organize your office visit to lead to that point, you’re going to use the least amount of time needed for your brain and your effort to get to whatever issue the patient is addressing.
So when it comes to low-acuity care, we use a system of orchestrated AI agents, not a single agent, but multiple agents separating various aspects of that patient encounter, just like they would be replicated in an exam room, OK, or in a medical clinic. And what that does is eventually lead you to that decision tree, where your system presents to you the case, the patient, the information, literally right in front of you. And in low-acuity conditions, Kevin, you and I both know: How much time does it take for us to make a medical decision about that? Maybe five seconds, maybe four seconds, right? We kind of know, given the data on low-acuity conditions, that we can make that decision, and very quickly make the decision about prescribing a medicine.
That’s what the architecture of my system does: It puts me in the position to amp up the capacity that I have for managing low-acuity conditions. And part of the program that I’ve created is that it identifies which ones are suitable for that, and the ones that are not suitable, that do need more time, attention, or face-to-face care, are offloaded into a different pathway, OK? So what we do is allow this patient at 9:00 at night to truly receive the care that they need in the moment, and it’s still delivered by me as the doctor in the loop.
So, key distinction: This is not autonomous AI, autonomous doctor AI work. I know that’s coming. People are going to push for that in the future, but this is a doctor in the loop making the decision, just like in my clinic. It replicates getting me that information to make that moment of critical decision about the condition. That’s it. That’s how it works.
Kevin Pho: Now, if there’s a human clinician in the loop, by definition, there has to be a little bit of a delay, because you have to sleep, you know. If you get a prescription request from the system at like 3:00 in the morning, how long does it normally take for you to act on that?
Tod Stillson: Yeah. So it depends on the actual company and business that is doing this, right? There are different ways to do it. We have what we would call an automated system that follows standing orders in our program. So in that after-hours version, our system is working and doing it and actually creating the prescription and sending it in, and then I’m reviewing it asynchronously the next morning: reviewing the case, reviewing the information, reviewing the prescription, and verifying its accuracy, OK?
And so it’s much like the standing orders that you and I operate with in our clinic, right? We tell our nursing staff. Let’s say a patient calls and says, “I need to refill my Synthroid,” OK? And the staff looks and goes, “Well, you haven’t been here for a year, and you’re due for a TSH. And so we’re going to send in thirty days’ worth of prescription, and we’re going to ask you to get a TSH and schedule an appointment.” But that’s a standing order. Your staff executes that because they don’t want to just send you a message to say, “Kevin, what do you want me to do right now?” And you’re going to tell them the very thing I just said: standing orders, right?
And so our system is able to allow our agents to orchestrate and follow the standing orders that are in place for every condition that we treat in our system. And that allows it to really function not in an autonomous way, but in an automated way. Automated is different than autonomous, OK?
Kevin Pho: So what happens in a scenario where you review a prescription that was sent in, and you don’t agree with it? The prescription has already been sent, right? Isn’t the cat already? So what happens in that scenario?
Tod Stillson: So I can tell you, first of all, the prescription can be canceled. The patient can be contacted, and you can initiate synchronous care as needed to support that and change that, right? And so that’s definitely part of the review and responsibility, the liability responsibility, that we have as physicians. However, I can also tell you that in my particular case, having had well over six thousand encounters in those environments, the system error rate is basically negligible. It’s nearly 100 percent accurate, OK? And so that is a very, very small subset, because in this particular case, we’re talking about low-acuity conditions and treating infections, etc.
The standing order process is really relatively straightforward, OK? Back in the day in medicine, we called this the cherry. You know, those are the easy visits, right? You and I know what these visits are like. They don’t take much time. They’re very easy. You end up talking about things that are nonmedical with the patient when they’re in the exam room, to fill the time and space, which is great relationship building, but your medical decision-making knowledge is very small. So patients want this sort of access to asynchronous care in the marketplace, and innovation allows that to come into play. And so we’re going to continue to see the evolution of this beyond my company. Other companies are doing similar things as well.
And I advocate for the fact that a physician does need to be in the middle, actually architecting the KB, or knowledge base, that supports that whole process and understanding how it works, but also being in the loop to verify the accuracy and safety of the medicine. And unfortunately, in the marketplace, we also see marketing companies using similar models of funnels and AI to get patients to doctors who rubber-stamp prescriptions, so to speak, right? And that’s where a lot of pushback comes in: Is this really medical care, or is this marketing? And I think, done responsibly and orchestrated the right way, you can end up with great antibiotic stewardship and great delivery of care, high quality, high safety, and today’s technology allows that to happen.
Kevin Pho: Now, you said that the system, when it comes to you, is near 100 percent accurate. And I know that you’ve been skeptical of autonomous AI prescribing, but you hear it from the venture capital companies and private equity: If it’s almost 100 percent, or negligibly close to 100 percent, what’s stopping autonomous prescribing eventually, and removing that physician from the loop?
Tod Stillson: I think it’s going to be legislation and regulation that will keep that from happening. But what we realize is that the capacity issue for doctors and the doctor shortage are becoming such an issue that it really is sort of a leverage point for this. And so again, I make the point for augmented clinical automation as opposed to autonomous AI. They’re distinct. They’re close; they’re in this similar space, but one has a doctor in the loop, and the other doesn’t have a doctor in the loop. So I’m very much morally, ethically, and professionally opposed to autonomous AI.
But you’re going to see these venture capital groups and other people continue to push for it, because the marketplace access and demands are so high that we’re going to continue to see that pushed against. But again, I’d rather create a company that does it differently and pushes in the opposite direction, but still enters the gray space, Kevin, to deliver care that is asynchronous and technologically supported and allows the patients to have the experience that they need and want, just like we all experience elsewhere in the marketplace.
Kevin Pho: So we’re talking about UTI as a common low-hanging fruit. What are some other conditions that would lend themselves to this approach?
Tod Stillson: Yeah. Well, we have forty-plus conditions that we treat, but, you know, pink eye would be another classic example, right? Impetigo, bronchitis. Now, not viral. Well, we do cover viral bronchitis, but we don’t give antibiotics for it. That’s a no-no, right? So we have antibiotic stewardship around that. And we have clinical guidelines based upon national criteria that allow us to make and formulate the right diagnosis for that.
Acute bronchitis is just a great example to me, if we really want to get into the weeds on this. We know most bronchitises are viral, right? They don’t need antibiotics. I fully agree and support that, but we know viral bronchitis gets treated with antibiotics all over the country all day long. So we have the ability to now orchestrate an AI system that allows us to accurately determine viral versus bacterial based upon symptom guidelines that are, again, national-criteria based, and really put people in the right buckets to decide which is the best way to treat that.
But I mean, you name it, we can treat, you know, athlete’s foot, tinea cruris. I mean, there are a number of things that we can treat based upon asynchronous symptoms and national guidelines. So again, forty-plus things: scabies, lice. There’s a lot of stuff that is just garden-variety, low-acuity conditions that patients get.
Kevin Pho: We’re talking to Tod Stillson, family physician and health care entrepreneur. Today’s KevinMD article is “Why medicine is late to asynchronous telehealth.” Tod, as always, we’ll end with your take-home messages to the KevinMD audience.
Tod Stillson: Yeah, I think the take-home message to the audience is that asynchronous care is growing in the world of medicine, and Fabric Health, which is one of the leaders of this in the marketplace, demonstrates that 90 percent of patients prefer the entry point to be asynchronous in nature. That’s what patients are looking for. They’re interested in that. And so I think it’s wise for us, if we are patient-centered as doctors, to begin to look for ways to support what patients are seeking already, so that we can provide the highest-quality, safest level of care and actually put us in the sweet spot to use our brains to make medical decisions.
Asynchronous low-acuity care is just one of those buckets. I do not want anybody listening to this podcast today to think that this is meant to replace all the other levels of care that require in-person care, hands-on labs, X-rays, etc. As I say to people all the time, I can’t deliver a baby online, right? So there are certain things that are always going to require in-person care.
Kevin Pho: Tod, as always, thank you so much for sharing your perspective and insight, and thanks again for coming back on the show.
Tod Stillson: Kevin, it’s great to be with you, and thanks for having me.
Listen and followApple PodcastsSpotifyYouTubeMore apps · Browse all episodes
