Alan Pitt: Welcome to Healthcare Pittstop. Today I have a friend with me, Dr. Andy Barbash. As many of you know, I’m a neuroradiologist but I work very closely with neurologists. Andy is one of the premier neurologists, but he practices in the cloud. He’s a virtual neurologist. Welcome to Healthcare Pittstop, Andy.
Andy Barbash: Thank you.
AP: You had a very traditional training at premier institutions in the country and you started your practice like most neurologists would, but then you moved to take care of people virtually. How did that happen?
AB: I did outpatient work at Kaiser Permanente in the mid-Atlantic area when I also helped run the electronic medical records (EMRs). And even then, I was becoming very aware of the power of the EMR system as a way to help people communicate with each other asynchronously about patients. So it actually started there. I began to see that, even in the early days of technology: the power of even simple Microsoft-based technology. Then, when I ran a neuroscience program at a hospital in a suburban area in the Washington D.C. area, having gotten familiar with the way things were evolving in terms of communications, messaging, collaboration tools, and video conferencing—it wasn’t about being in a rural area. I lived in a somewhat urban area. I felt a great deal of responsibility for helping our nurse practitioners and emergency rooms deal with issues.
By 5:30 at night in the D.C. area, it didn’t matter that the hospital was five miles away—it was an hour away. And so I began to explore ways to use secure chat. We used practical tools around and videoconferencing even in the early days, so that I could bring my expertise into the Emergency Room and into the Intensive Care Unit to meet with families, for example, after patients had been admitted and I worked collaboratively with nurses. So that’s how I began to make the migration.
Then I got to the point where I felt that I had built the program, a neuroscience program at the hospital system, part of a larger health system. I thought it was time for a younger person with different goals and objectives to take the project on and really grow it. I wanted to move on and take this virtual model and have that become the way I practice in order to be able to bring my expertise to both hospital settings and also to the outpatient environment.
AP: You know, Andy, I completely agree. People often think of telemedicine as helping rural populations that are hundreds of miles away. But I very much agree that it can be just a couple of miles away, even in a hospital, where you can provide a lot of value to patients. That’s a big deal. A lot of patients worry that virtual care is not as good as in-person care; that you can’t do the whole exam. Do you have any thoughts about that?
AB: Yes, without being too cynical. The evolution of healthcare, if you take telehealth out of it and you look at the average transaction going on in a hospital or an office setting, the nurses or the doctors, one of the biggest complaints patients have, is that they’re spending more time interacting with their computers then they are with the patient. The physical exam is done by more machines. The physical exam is becoming a smaller element because we are fundamentally becoming communicators and information managers. And so if you accept that as a fundamental transformation of the practice of medicine, whether it’s in cancer care, psychiatry, radiology, etc., then it doesn’t remove the importance of the physical part, but it focuses more on communication, collaboration, shared decision making, and working together.
The rest of the world is doing that virtual model. In fact, when we move to a virtual practice model or a telehealth model (whether it’s phone, texting, messaging, by video, or just looking at files, etc.) in some ways we’re allowing ourselves to focus on what is at the essence of what we do now, which is managing information, data, evidence. I think you can make a very good case that for an increasing fraction of healthcare, a virtual model is actually a better model. But part of that virtual model is now learning to work with remote people on site like a nurse.
AP: I agree with that too. It’s also a different kind of practice. It’s not necessarily better or worse it’s just different.
AB: It’s different. And the nature of our practice is different than it was 20 years ago when we didn’t have the EMRs. And so it’s just reflecting the fact that our nature of medical practice is evolving anyway.
AP: It’s a response. I’ve talked to you in the past, and something that isn’t well understood by a lot of families and patients is that there’s a risk and a benefit to every decision we make as doctors. So there’s no real free lunch. So sometimes when we give you a drug, we have to subsume those risks. You’ve talked a lot about shared decision making. How do you incorporate shared decision making in your practice? And are there things that the telemedicine community could do better to help you with that?
AB: Well, yes. Fortunately, my exposure to telemedicine has been largely focused around the acute care telestroke/acute stroke decision making model. But I can give you many examples where the same workflow process would relate to other kinds of conditions. The beauty of that is that, at the end of the day, if you think about why I am brought into a telestroke environment, my role is not to administer a drug, but to gather information. To build a relationship rapidly with whatever resources are available—that could be the family present, the family on the phone, no family, a patient who can communicate and talk, a patient who can barely communicate and talk, a nurse who knows the patient, or a nurse who doesn’t know the patient, a doctor who is directly involved, or a doctors who is not. So one of the things we do is we assess our environment quickly. We need to understand who we are interacting with. It’s not a given. And the primary reason is for us to help guide a shared decision making process. We are clearly all finding that unless you literally have somebody who’s unconscious or can’t communicate (which is becoming a smaller fraction of the time), the patients and families are increasingly comfortable with the idea that we’re quoting statistics, that we are not treating the patient as the statistic, but they understand that if we don’t actually treat them as a statistic, we’re mis-serving them.
We are bringing the same kind of evidence discussion into play every single time we see a patient, regardless of where they are geographically, what race they are, what language they speak. You become a real believer in the model of shared decision making. In reality, with the exception of people who just can’t decide, you cannot give this clot-busting drug unless you’ve had the family or patient reach a decision that they’re comfortable with it. And the only way to get there is to discuss benefits and risks in general, and then render your opinion and perspective of the relative benefit in their situation, and the relative risk in their situation.
An increasing number of things we deal with are not black and white statistics. There’s a gray zone. That all emanates from the relationship we’re building. So when a patient says, “Dr. Barbash, do you think I’m at a higher risk of bleeding because I had a nosebleed last week?” I may respond with, “There may not be very good evidence about that, but I might have an opinion relative to how bad you look right now.” Families and patients are engaged in assessing what they believe their own deficit is relative to what they consider to be a problem for them.
AP: I think informed consent is not well done by most in the medical community. It’s kind of a perfunctory “check the box.” I think it’s hard for patients to truly give informed consent, and I applaud your efforts to get closer to a real decision about what to do next.
AB: We’re not asking them to consent to something. We’re literally asking them to make a decision with our assistance.
AP: There’s a big change in your area and stroke. There have been two recent studies- the DAWN study came out in the New England Journal of Medicine in December and, most recently, there was the DEFUSE 3 trial that came out of Stanford. With newer technology for triaging patients, both studies suggested that rather than treat patients only within a six-hour window after stroke onset, we could expand that window to 24 hours after stroke, either with clot-busting agents or with thrombectomy, actually pulling the clot out of the brain. Do you have any thoughts about that?
AB: When I was studying physics—and this is not as much a digression as you might think—I remember being told, “Well, the speed of light is the speed of light, and everything else in the world is relative.” And I remember thinking, Well, what happens if one day we find out, well, it’s not quite the speed of light? Right?
In the field of stroke, the fact that we’ve all been practicing with the assumption that there’s this hard line—so if the time they were last normal was not, say, 4.5 hours—even with everybody knowing how uncertain that was, because it was completely based on some observer’s perception of the time when something happened—but it allowed us to focus on processes and literally drop the ball and say, “No, no, it’s been four hours and 32 minutes instead of 4 hours and 30 minutes, hence don’t give this drug. I think fundamentally that allowed us to put stroke care in a box.
We have limited numbers of interventions. We would only take patients and send them for procedures (other than a clot-busting medicine) if they met very strict criteria: time limit, age, etc. In the process of giving TPA, even before the DAWN trial, over the last ten years, everything we thought was an absolute reason you couldn’t give it has turned into a relative issue (with a few exceptions, obviously). Now, with these new trials, it’s a natural evolution. We are now being forced to take a more personalized view, a little less statistical view. It’s supported by current evidence, but the evidence is going to evolve as the field evolves. We now look at it as not a black or white 3 hours, 4 hours, yes, no, you have these conditions or you don’t; you get it or you don’t.
Now it’s going to be more, well, your stroke is of this severity, you’re of this age, you have a CT scan with this kind of abnormality… Now we’ve had we have more patients getting an advanced study like a CT angiogram or CT perfusion study. Now there is more personalized information relative to this particular person’s situation.
Although don’t have a much wider array of interventions, but we have more specific interventions. It’s like the difference between giving chemotherapy that kills on all your cells and chemotherapy that only kills the right cells.
Now we have an intervention that literally mechanically goes after the clot and has a whole different risk-benefit profile in terms of individual patients. What’s happening is that it’s pushing us to be more sophisticated, more tailored to the clinical situation of the patient, their medications, their decisions, their deficit, and how they were before. So we have more pieces of information to assemble in a shorter period of time in order to get to potentially offering the patient a more intensive interaction that might not be available in person.
AP: For folks listening out there, this means that for many people who came in with a stroke, there was nothing we could do as doctors. Now many more people should get an evaluation to figure out which one of these buckets they fit in and if there is a therapy. Is it a fair summary?
AB: Yes. To be honest, we were seeing this well before the DAWN trial. We were seeing more and more hospitals under more and more pressure to make sure they never ever miss the opportunity to have given TPA to someone when it turns out they had a stroke on an MRI scan the next day. We’ve all seen an increasing number of patients on an urgent basis because they might possibly be candidates for TPA. Now we’re broadening the definition; even if they are not candidates for TPA, a much larger number of people may be candidates for mechanical interventions.
AP: I think the absolute number of people being treated for strokes may not go up that much, the number of people requiring evaluation for possible treatment will rise dramatically. I’m not sure the healthcare systems are prepared to handle this.
AB: We need to get smarter about how we do it. We need to get smarter about how we triage scenarios. We need to be more sophisticated in a more timely fashion in order to literally allocate a given patient in a given situation into a given kind of benefit-risk profile. They will still be shared decision making but it is going to be a little bit more sophisticated. There are more data points that are relevant.
AP: You’ve been a leader in moving your practice to the cloud. Now that the tools of telehealth more accessible, do you have any words of advice for other clinical specialties thinking about moving their practice to a virtual space?
AB: What I would pose to any specialist is to have to answer the question why making themselves available on a more virtual basis is not be a good thing for both themselves and their patients, and the people who refer patients to them. The vast majority of specialists—endocrinologists, G.I. doctors, orthopedic surgeons, when everybody is involved in helping to assess the history, the story, the medications, the underlying conditions, the environment, their previous baseline function, etc., and to try to determine whether this person is going to be a good candidate for treatment. This is possible way in advance of the hands-on kind of stuff. This is abetted by more and more objective tools—apps on people’s phones, remote devices, to monitor data, etc.
There are certain areas where I already see telemedicine: psychiatry, mental health. Why? The need is enormous. Our healthcare system really doesn’t have the resources or wherewithal to address mental health on a large scale. The ramifications of not treating people are very serious. We take it seriously in neurology—it’s the brain. Suicide is a pretty serious thing as well.
There’s a gross shortage of professionals who have experience and training. A tremendous percentage of these acute psychiatric interactions can be done by video telehealth assessment, etc.
In the outpatient environment, this becomes even more true. Endocrinology is a good example. If you talk to most hospitals, they will probably tell you: We have patients in the ICU but we can’t get an endocrinologist to come to the intensive care unit! And yet, you have patients where you would just do anything if you could get an endocrinologist to be in their office, allocate half hour of time, get on video with you, look at the EMR with you and look and some data and decide what to do. That would make all the difference in the world. I think most specialists would probably find it advantageous to look at what they do and ask themselves why they wouldn’t want to provide expertise in a more virtualized manner.
AP: I think most doctors would respond that they can’t get the money because there’s no billing, but clearly you’ve figured it out and I think it might be interesting to have a follow-up conversation around how people make money with a virtual practice. I have a lot of thoughts about that. I’ve seen a lot of successful practitioners do it, but that should probably wait for another day.
Andy, I very much appreciate your time and certainly found this very insightful. I hope people realize that we can do more for stroke now. It’s really incumbent on our hospital systems to offer the right care at the right time. Whether that’s in person, or, if need be, virtually. Thank you so much.
AB: My pleasure.