In Conversation with Dr. Neal Sikka and Dr. Colton Hood about Remote Patient Monitoring
Editor’s Note: Dr. Neal Sikka and Dr. Colton Hood are emergency medicine physicians who work in the Innovative Practice & Telemedicine section at George Washington University Hospital (GW). We spoke with them about their experience implementing remote patient monitoring (RPM) programs, GW’s Maritime Medical Access program, and patient safety considerations in the remote environment.
Sarah Mossburg: Welcome, Dr. Sikka and Dr. Hood. Can you both tell us about yourselves and your current roles?
Neal Sikka: I am an emergency physician and professor of emergency medicine at George Washington University (GW). I also lead the Innovative Practice & Telemedicine section at the GW Medical Faculty Associates where we seek solutions that can leverage our physicians’ skills outside of the traditional boundaries of the hospital walls. Those areas include things like telehealth, remote patient monitoring (RPM), and other types of innovative solutions that meet the needs of our community, both locally and around the world.
Colton Hood: I'm an assistant professor of emergency medicine at GW. I am board-certified in emergency medicine and clinical informatics, and I co-direct the fellowship in telemedicine and digital health at GW. I’m also physician champion for telehealth at the GW Medical Faculty Associates and a member of the Innovative Practice & Telemedicine section in our department. I am interested in developing and using new technology to improve clinical care such as RPM or using augmented reality for remote support during bedside procedures.1
Sarah Mossburg: Today we are interested in talking about remote patient monitoring or RPM. For readers who are not familiar, can you give us a brief explanation of RPM?
Neal Sikka: Underneath the umbrella of telehealth, we think more specifically about telemedicine and remote patient monitoring. Telemedicine is remote diagnosis and treatment using some kind of information communication technology. RPM allows you to better monitor or manage an acute or chronic health condition from a distance over time. We think of RPM as the collection and transmission of patients’ physiologic data from home or some other environment and then the management of the patient using the data that are transmitted.
Sarah Mossburg: Could you tell us about how RPM has evolved over time? What are some ways that RPM was used in early adoption versus present day? How has the technology changed as part of managing patient care? How has patient safety been part of that evolution?
Colton Hood: Originally patient monitoring was pen, paper, and maybe a blood pressure cuff. A patient would call their doctor and tell them their blood pressure or bring it to them written on paper. In a more modern environment, we have increasingly connected medical devices that automatically send information to healthcare providers that can be processed accordingly. Cellular connected medical devices are always available. As long as you have reception, the data will go to your physician.
Neal Sikka: One of the earliest use cases of RPM was cardiac monitoring. Patients went home from the hospital with a Holter monitor or other implanted device, and data were uploaded and then monitored. Another way to think about the evolution is that we are using this methodology to gather information that occurs in between visits for individual health.
When deploying RPM-connected technologies to a population, you start with a group of people with a similar disease. For example, you start thinking about population health and population management. Now you can leverage technology to look at data for a population and look for outliers and trends to see where individual or organizational approaches to managing a condition could impact that population's health.
Sarah Mossburg: With technology, sometimes we find glitches and problems once we start using it. With the evolution from pen and paper to connected devices, have you run into problems?
Colton Hood: Electronic devices are now connected and available 24 hours a day. Previously you would have to call someone in the office or leave a message, and work is done on a different time scale. The difference now is that, when you take your blood pressure and you have a connected device, the results can be transmitted instantly. Because the data are available immediately, that expediency and expectation to read results and provide care is much different. It is both helpful and can be overwhelming at the same time.
Neal Sikka: Many patients do not know what to expect. You have to provide a clear expectation that while these data are going to come to us right away, a clinician will react to a trend rather than each data point. This is how these systems should be utilized, so managing that expectation is important.
There is also the benefit that you get potentially more accurate data. At times, patients will want to please their health care provider, and with the pen and paper method they may not have as consistent or accurate data. For example, if a patient has an appointment for a physical, but forgot to record their blood pressure readings for the last week, that patient may not remember exactly what those values are. They may report values that are better or worse than what they actually were.
Colton can also speak a little bit more about managing remotely monitored data, and how information systems and clinical entities handle these data. Even though the data are coming from devices, some of which are regulated by the Food & Drug Administration, we consider it patient-generated data. These data tend to fall in a different category than the data we collect in the clinical setting, like when we take your blood pressure at a doctor’s office.
Colton Hood: When setting up these systems, you have to decide the data governance structure on an organizational level. You can decide to accept the data as fact and put it into the medical record as data coming from the patient, or you can decide to have a clinician review the data that comes through, and then accept that data as fact. There is not a single correct answer for this. It is an organizational decision to how you want to accept and incorporate the data into the medical record. For example, when you can graph the vitals, you can highlight the data supplied by the patient and the ones that were taken in the clinic. Understanding what to do with the patient-supplied data is important.
Sarah Mossburg: How much variability do you see in the data supplied by patients versus the data supplied during clinical care?
Colton Hood: It depends. Some clinicians are skeptical of home medical devices and their accuracy. Many patients may have what is known as white coat hypertension, where they may have higher blood pressure in the clinic versus when they are relaxing at home. There can be discrepancies. You have to take a bird's eye view of the data. There are often variations in blood pressure during the day. For the most part, the medical devices are accurate, and the values are going to be similar. There is a chance that someone can misuse the medical device, or for example someone other than the patient could use the device, and that data will go into the medical record. Ideally, patients would notify us in those cases and we could correct that data.
Sarah Mossburg: As a nurse who worked in the hospital, one challenge that I saw was occasionally inaccurate readings from medical devices. For example, a pulse oximeter may not have been accurately measuring oxygen levels and show a very low level, but when looking at the patient, you see that the patient is fine and not in respiratory distress. Have you come across these scenarios in remote monitoring? Is it more challenging to manage potentially false readings given that you are not able to easily walk into the patient's room to assess them?
Colton Hood: That actually happened at the start of the COVID-19 public health emergency (PHE). We were completing acute monitoring for patients that were discharged with COVID-19 and the pulse oximeter readings could occasionally be problematic. The most important part of RPM is providing good education to the patients about the devices, and what they should do when they get particularly concerning readings, like an out-of-range result. We received many phone calls from patients where we identified that the equipment was not being used correctly. For example, the patient did not warm up their fingers or was putting the device on incorrectly. Occasionally there were concerning readings, and if they remained concerning, we would activate an emergency plan. By creating emergency plans in telehealth and RPM, a patient and clinician will know what to do in case monitoring shows very out-of-bounds readings.2
Neal Sikka: When COVID-19 was an emerging disease, we were initially seeing patients who actually looked well, but had low oxygen levels and high heart rates. It was even more difficult for us to interpret that information. At the end of the day, you often will think of your device data as objective, and the patient symptoms as subjective, and will need to act on that information. It comes down to your protocols: How do you think about the pathophysiology of the disease you are monitoring, the way you are collecting these data, and how are you going to react to the individual data or react to the trends? If a spurious value makes you concerned, what is your escalation plan and what is the timeframe for action to occur? All these things become complicated when you start to unwind them. If a person wakes up in the middle of the night and takes a blood pressure that is either high or low, what do you do with that information? Did someone actually see it? Is there a liability associated with that as well? It becomes complicated and comes back to setting expectations on both the patient and provider side.
Colton Hood: Remote monitoring almost needs to be personalized per patient and per disease. There are some conditions in which you may tolerate slightly out-of-range readings, but someone else may look at that data without being familiar with that program or the patient and may not have the same assessment. It takes thought and planning.
Neal Sikka: In 2012, our team was awarded a Health Care Innovation Award by the Centers for Medicare & Medicaid Innovation Center for RPM. We monitored the blood pressures of home peritoneal dialysis patients. We worked with nephrologists to set unique blood pressure parameters for every patient. We knew their trends and that some of them had typically higher blood pressures and some of them had typically lower blood pressures, so we created individual thresholds.1,3,4
Sarah Mossburg: It sounds like at least one of the key processes is thinking through protocols for specific diseases, and even individual patients, as you are setting up and managing these programs. What other key processes or practices are integral to setting up these programs?
Neal Sikka: On the protocol side, patient selection is important on the front end: Who is really going to benefit from remote monitoring and for what condition? You need to set up those expectations and make sure that patients have the digital literacy to be able to manage transmission of the data. While we hope it's always plug-and-play with medical devices, we have not found that to necessarily be the case. For example, being able to manage a connected device, connect that device to some type of transmission device like a smartphone or data hub, and then ensure that the data get uploaded to the cloud and to the monitoring station. From there, you need to know what types of trained staff are going to monitor that data. We often try to use emergency medical technicians or paramedics. Nurses or physicians might be involved. How do you supervise them and be available for escalation? What are the protocols they will follow? How do you document these interactions? We have to break down all the different parts of the program.
Colton Hood: Good information systems can flag abnormal findings timely and flag the data that need to be prioritized. If you implement RPM at a larger scale, then a lot of information needs to be processed and you need to use technology to help urgently identify important values and have workflows to manage it in place.
Sarah Mossburg: Do you explicitly think about patient safety as part of this process when you are setting up a new system?
Colton Hood: Yes, it is important to have good patient education so that they know when readings are normal and abnormal, and what to do about it. When readings are abnormal, patients need to know when to dial 911 or call their physician. A solid education leads into patient engagement. We need to provide education and be available to assist patients until providing care like this is the new normal.
Sarah Mossburg: What are differences in approaches to remote monitoring for acute or chronic conditions? RPM has been used for a long time for more chronic conditions. Are there differences in managing acute versus chronic diseases?
Neal Sikka: RPM was initially designed to manage chronic diseases. For example, with hypertension you need blood pressure monitoring, and with diabetes you need glucose monitoring. For congestive heart failure, you would need to monitor daily weights in addition to blood pressure. Patients may also have multiple co-morbid diseases, so you may be tracking their blood pressure and glucose. RPM is slowly evolving to expand into other chronic diseases as well. For chronic obstructive pulmonary disease (COPD), you can monitor breathing, pulse oximeter numbers, and inhaler use.
One of the first acute scenarios in which we used RPM was COVID-19. Even in 2020, RPM was an emerging technology. Since then, we have talked about a number of more acute conditions that could be monitored. For example, instead of admitting a patient with pneumonia to the hospital, you could monitor that acute episode at home with pulse oximetry and other vital signs. That situation extends to the concept of hospital-at-home, which functions as a remote monitoring solution. Instead of admitting patients to the hospital for certain conditions, patients could go home and get telehealth visits or in-person visits, and then have remote monitoring of vital signs to track the progression of that disease and make interventions. That may be one way to think of that spectrum of managing chronic diseases to acute diseases.
Sarah Mossburg: As payment models evolve, is it correct to think there will be more opportunities for using RPM?
Neal Sikka: Yes. In a traditional, fee-for-service (FFS) model, you can use RPM billing codes. These codes are similar to telehealth. You must navigate a patchwork of payment from different payers. You have Federal payment programs, like Medicare FFS, Medicare Advantage, and various types of Medicaid. There may be variability in payment among those payers, as well as the commercial payers. It also varies by state. GW’s patient population is across three states (District of Columbia, Maryland, and Virginia), so we have to navigate the rules of those three states, and it becomes difficult to do these complicated and operationally intensive interventions on our population.
Within an accountable care organization (ACO), the incentives might be different because we have financial incentives for the improvement of population health. You may be able to reduce some of the barriers. For example, if we have a population with a high prevalence of hypertension, then maybe we could implement a hypertension monitoring program among that population. We can do that at scale with some cost efficiency to get the outcomes we are looking for. We are not there yet, but that is the way to start and may be the way we see programs within integrated health systems implement RPM as well.
Sarah Mossburg: It seems like there are a lot of challenges in setting up these systems related to costs and billing. Could you speak about staffing for RPM?
Colton Hood: If you are in the FFS realm, to be able to bill for the service, the provider or someone who is under general supervision by the provider, like a nurse, must spend 20 minutes evaluating the information. Their job description must include the ability to handle the physiologic information that is being sent to them. You also need one real-time, audio-video capable evaluation during the month. You also need information technology (IT) staffing to set up the system and make sure it works. You may also need staff to train patients as well, like a medical assistant or nurse. We actually find that the person who receives the data, typically a nurse, physician assistant, nurse practitioner, or physician, will be doing much of the work because it can be a lot of information.
Neal Sikka: An ideal place to launch a small RPM program is maybe a smaller practice that is still growing and has some excess capacity within their existing staff. At organizations that are larger, everyone may be working at full capacity or over capacity. When adding an RPM program, you have to justify the revenue to support adding a new staff member.
A number of companies will provide off-site remote monitoring services. It may allow the physician or treating providers to simply identify the appropriate patients for monitoring and write an order. The company will help deploy the technology, do the monitoring, and transmit documentation of the twenty minutes of monitoring for billing purposes. They would send the data to the physician to make treatment management decisions. That can facilitate the establishment of RPM programs. You have to figure out what financial agreements work for your practice.
Sarah Mossburg: GW has a Maritime Medical Access Program that started in 1989, which provides emergency services in remote locations. Can you tell us how you use RPM in that program?
Neal Sikka: In that program, there are patients who are in very austere environments, such as a ship at sea, and can be days away from land or definitive care. Most our vessels have a trained provider on board. This training can vary and can be as simple as someone who has taken a first-responder course to facilitate the conversation that we have with the patient. It can also be an EMT, paramedic, nurse, advanced practice provider, or physician. Historically, these are acute problems that arise where there is illness or injury. With the COVID-19 PHE, vessels have had a difficult time coming to port because of restrictions, so they changed their crews out less frequently. As a result, we are being asked to treat more patients with chronic diseases like hypertension or diabetes.
Generally, in acute cases, we work through the medical officer to gather a history and physical exam like you would through telemedicine, but often the technology we are using is limited due to their environment, such as their bandwidth. We might be limited to a broken conversation over a satellite phone. In some scenarios they will have intermittent email access for communication. In other scenarios, we can use various video telemedicine and video systems that can provide two-way real-time access. In these scenarios our clients have equipment that can allow for real-time transmission of vital sign data, and we can do RPM.
In traditional in-person, FFS care, you can see your doctor one day and it may be hard to check-in the next day. In this program, we will have check-ins very frequently. Our command center will facilitate a 4-hour follow-up, and then an 8-hour follow-up, or at a minimum, 24-hour to 48-hour follow-up. We will touch base with our patients to make changes in medical decisions or see how their disease is either improving or getting worse.
In the RPM realm, when the vessel has capacity, we might continuously monitor the patient until they get to port. For example, we could have serial electrocardiograms sent to us or stream vital signs. We would then be able to help the medical officer watch the patient and react to objective information that might indicate the patient’s condition is worsening.
Sarah Mossburg: How can some of that technology be used in rural areas of the United States? Do you see that model working similarly?
Neal Sikka: We learned from these remote and austere environments that telehealth processes are applicable to both rural and urban environments. It is thinking about your patient, understanding that you have some distance—whether that's a horizontal or vertical distance—to see your patient. The technology you use to communicate with your patient and gather the information you need, could be similar to a medical officer on a ship, for example a family caregiver who can help relay or extend the eyes, ears, and hands of the treating clinician to gather more information to make diagnosis.
Before the pandemic, most telehealth was limited to rural areas and designated medically underserved populations due to reimbursement rules. In an urban area, we must consider different barriers for our patients. We have patients who may only have to travel a few miles across town, but even distance can be a significant barrier for a variety of different reasons, including time off from work or access to transportation. Additionally, in urban areas you have vertical distances. Some patients with mobility limitations who are living in high-rise building or high-density housing have significant difficulty seeing their doctor. GW has a homebound patient program where the doctors go to the patient because we see these barriers to patients receiving care as if they were in a rural area. It is not very far, or is a vertical distance, but it is just as challenging for patients to get the care that they need.
On the RPM side, our experience would be more related to acute care monitoring, like a hospital-at-home scenario, where we see real-time vital signs, utilize telehealth, and work with a family caregiver to change a treatment plan. We are doing less chronic disease management on the ships, except manually managing blood pressure medicines, setting expectations, and giving advice on how to manage blood pressure until the vessel gets to port. Typically, our clients do not have good ways to connect RPM and send blood pressure automatically through a system every day or every week, so our remote monitoring is through the phone or email. We will get the data and then react to it.
Sarah Mossburg: Are there other settings where RPM could potentially help manage patient care?
Neal Sikka: One that comes to mind would be places like assisted living because they are not necessarily clinical settings like a nursing homes or long term care where they might already have the ability to do vital signs. We have done projects with patients who have memory care issues. Using devices for vital sign monitoring could potentially help improve their care and provide safety to patients who may not be able to manage their environment in a safe manner.
Colton Hood: Correctional care could also benefit from RPM. Medical care can be challenging in that situation. You may have to be creative because of logistical issues there.
Neal Sikka: Also, we already do a lot of self-tracking. We monitor our own steps, heart rate, exercise, and weight. All sorts of measurements are enabled through consumer-grade wearable devices. Eventually some of that data could become connected into the healthcare system more tightly. In certain scenarios, you can submit some of that information you collect on your own into your own electronic health record.
Sarah Mossburg: You mentioned that the paradigm changed after the onset of the COVID-19 PHE because it was the first acute use of RPM across the industry. I would imagine it was pretty important to put RPM in place quickly, and there was high demand. How did you manage that immediate need?
Neal Sikka: We were lucky at the start of the pandemic because we already had this infrastructure from our maritime program. We had staff in place and knew this was going to be different from hypertension monitoring where you could look at a trend over a week. We needed to react to that data in real time, and having that capacity is what allowed us to do it. Otherwise, it would have been very difficult for us.
We also had experience with telehealth. We had experience with thinking through the use of technologies. We had been thinking about COVID-19 before it spread to the United States. When we were actually seeing patients with this problem, the pulse oximetry was the obvious physiologic parameter to monitor. We knew there were hospital capacity problems and a lot of different pressures and stakeholders that aligned to make this something that we could try as soon as we could.
Colton Hood: We were pretty hands-on, and I think that allowed it to be successful. When we gave out the medical devices, we could call patients. We were busy, but not too busy to call patients when they submitted abnormal information, help troubleshoot issues, or make sure they were doing okay. There were so many unknowns that we had to be very hands-on because we did not know what the course of illness for COVID infection was going to look like. We ended up sending about 6% of patients back to the hospital. We learned a lot very quickly.
Neal Sikka: The telehealth waivers during the PHE facilitated supporting the program from a revenue perspective. Typically, with RPM you need to have a patient transmit 16 days of data to be able to bill for that service. Through the waivers, you had to transmit one day of data. While this was not a program designed to generate revenue, the revenue did offset some of the costs.
These programs were very time intensive. We had a lot of people available, like fellows, in our telemedicine digital health fellowship. We also had medical students who were available to support at the onset of the PHE. Many other institutions tried to do RPM, and a lot of it was a nexus of opportunity. We had a defined need. We had a biometric parameter that made sense. There was extra capacity of people around to help support these programs. It allowed us to develop these programs because we were all focused on this one disease at that time.
Sarah Mossburg: It sounds like one of the ways that you maintained patient safety with the increased volume was that hands-on aspect of the program as you implemented it.
Colton Hood: Yes. Engagement is probably the most important part of RPM moving beyond the actual clinical outcome—engaging your patient with their health and getting them interested and checking and understanding things like their blood pressure.
Neal Sikka: We had operators in our command center call patients who did not transmit data to try and make sure it wasn't because they couldn't transmit data. That is one of the challenges with an acute vital sign like pulse oximetry. You worry that if that value is too low, then maybe someone cannot transmit the data, and you have to check on them to make sure they are okay. We implemented wellness checks on patients whom we could not get a hold of because we were worried when a person was not responding. Was it because they were disengaged or because they could not physically respond?
Sarah Mossburg: It sounds like in this acute COVID-19 scenario that you were able to improve patient safety using RPM. Do you see that with other chronic diseases as well? What are some other potential improvements in safety?
Neal Sikka: Sometimes those safety improvements may not be completely obvious. For example, if you manage high blood pressure better, you can improve outcomes for patients, which could reduce the risk of heart attack and stroke that may lead to potential disability and falls, among other things. Another potential way patient safety could be improved is with medication compliance, maybe by picking up on problems with how patients are reacting to medications, and if they are making them dizzy or lightheaded, or if there are opportunities for you to pick up on challenges that patients are having that could be harmful to them in between visits separated by a significant amount of time. Identifying symptoms and effects early and reacting to them is a core component. You can change treatment plans based remote monitoring data. You can also augment that with telehealth where you get a view into someone's home and see if there are tripping or falling hazards, or other issues at home that might impact care. If you put all those things together, you have a great impact on health outcomes and patient safety.
Colton Hood: Another way is being able to detect conditions earlier. For example, you might be able to catch asymptomatic hypertension that is out of bounds earlier. If you only check a patient’s blood pressure once every six months, but it is regularly high for those six months, then you could potentially intervene earlier than at the next physical appointment.
Neal Sikka: In our 2012 program, we monitored patients for two to three years and we noticed that monitoring led to better control of blood pressure. The patient's feedback was that they were motivated to transmit data because they knew that someone was going to look at the number. They were motivated to make sure they were taking their medicine. You would see patients take serial readings in a short frequency to try to see if that was really an accurate number. Knowing that your providers engage in your health is engaging for a patient.
Sarah Mossburg: That’s really interesting. Is there anything else you would like to share about RPM?
Colton Hood: Like any technology intervention, RPM is a tool that impacts patient care. It might be hard to define how it improves and benefits it and then put a monetary value on it. You can potentially use indirect measures like in value-based care and ACOs, or you can calculate savings for keeping someone out of the emergency department or the hospital. Direct cost savings can be hard to calculate. If you can invest in the technology, there are many potential benefits to patient care.
Sarah Mossburg: Thank you. This has been an interesting conversation and discussion. I really appreciate your time today.
1. Lew SQ, Sikka N, Thompson C, Magnus M. Impact of remote biometric monitoring on cost and hospitalization outcomes in peritoneal dialysis. J Telemed Telecare. 2019;25(10):581-586. doi:10.1177/1357633X18784417
2. Aalam AA, Hood C, Donelan C, Rutenberg A, Kane EM, Sikka N. Remote patient monitoring for ED discharges in the COVID-19 pandemic. Emerg Med J. 2021;38(3):229-231. doi:10.1136/emermed-2020-210022
3. Rosner MH, Lew SQ, Conway P, et al. Perspectives from the kidney health initiative on advancing technologies to facilitate remote monitoring of patient self-care in RRT. Clin J Am Soc Nephrol. 2017;12(11):1900-1909. doi:10.2215/CJN.12781216
4. Magnus M, Sikka N, Cherian T, Lew SQ. Satisfaction and improvements in peritoneal dialysis outcomes associated with telehealth. Appl Clin Inform. 2017;8(1):214-225. doi:10.4338/ACI-2016-09-RA-0154