Good morning everybody. Thank you all for joining us here this morning. I know we're running up against sepsis in Manny Rivers, so we have a small crowd, but nonetheless, we're excited to have you here this morning for our telecritical care session. My name is Chris Palmer. I'll be the moderator here for today's session. I'm from Wash U in St. Louis, emergency medicine and intensivist, also co-director for our telemedicine program. I'm really excited for you guys to join us here today. I'm excited for our speakers here to kind of share their experiences and their knowledge with you guys so that you can bring back some of their ideas and things that they're using at their programs to your programs and institute some really novel ideas. We have a multidisciplinary panel here. We have Jeremy Pamplin, as he refers to, Army critical care guy, a doctor, Teresa Rincon from UMass, critical care nurse, and Desiree Kosmicki from Atrium Health of Farm D. So each of them has a tremendous amount of experience and expertise in their area, so I'm excited for them to share with you their experiences. And we're going to hold questions until the end, so after each one of their talks, if you can just hold your questions until the end. And with that, we'll get started with Jeremy. Thank you so much. Good morning, everyone. And as Chris said, I'm an Army critical care doc, currently stationed at the Army's and really kind of the DOD's or the military health systems telemedicine and advanced technologies research center. But I spent most of my career doing trauma and critical care across the military and certainly across the world through several deployments. So I don't have any disclosures that I need to provide other than to say that the opinions that I'm going to share with you today are those of my own and not of the uniform that I wear or the government that I represent when I'm wearing it. And I'm hoping to answer a few questions for everyone here today, particularly what is telemedicine really, what has telemedicine done for us today, and what can potentially telemedicine and indeed technology to a larger extent do for us in the future. To get us started, I'd like to start with this quote. That will never come into general use, notwithstanding its value. It is extremely doubtful because its beneficial application requires much time and gives a good bit of trouble to both the patient and the practitioner. And because its hue and its character are foreign and opposed to all our habits and associations, it is just not going to get used. And that is from an 1834 article in the London Times referring to the stethoscope. But I think it's incredibly applicable to our world today because when we are approached with new technology solutions, it is oftentimes that we are constrained by our current reality and we can't imagine what the future might look like. And I hope to break some of those perspectives today. So it's really not about the technology itself, right? If you think about a cell phone or a television screen, the cell phone as a device really doesn't do a whole lot. The television screen really doesn't do a whole lot. It's not until you start using the applications that are associated with that device or that technology or watching the programming that's available on that television screen that all of a sudden you get to get a flavor and understanding about what that technology's use might actually be. And it's probable that in many cases the technology as it actually gets used in the world around us is imagined and is implemented in ways that the original designer of that technology really never imagined. If you think about this technology, it was designed to do one thing. It was designed to measure pressures within the cardiovascular system. And as such, it works very, very well. But when you start thinking about how it's used, how it's implemented by humans in the critical care environment, it clearly changes the behavior that we manifest. It changes our practice patterns. It changes the things that we do. But it may not have the intended outcomes at all times for all patients that we try to study this device to do. So really it's about the human element. And if you forget about how the human element is used with the technology, then really the technology is just a cool gadget that looks pretty on a wall or looks pretty in your hand and doesn't do a whole lot. It's really the impact of how that tool works with the human that makes the difference. So what does technology do? It enables us to do more than we can otherwise do ourselves. It extends human capability and capacity to see inside the body, to make a diagnosis, to identify potential problems before they exist, to help us be safer providing a second set of eyes or a backup system for when you may be fallible or tired at your most risk of having a mistake. And ultimately to do less harm when we are helping people that we have sworn to protect and have sworn not to hurt, and potentially have been doing that in places that they would otherwise not have the ability to provide a specific service. It can help us identify novel diseases or diseases in novel ways. And it can certainly help us provide care to critically ill or critically injured individuals that are in harm's way in very austere locations in the military and in very austere locations in the civilian world as well. I'm not sure how many people saw this article last year. It was actually one of my favorite articles of the past year. It was a report on Avera Telehealth Systems in Sioux City, South Dakota. Anybody hear from Avera by any chance? Okay. I'm amazed, I'm enthralled by what you're able to do. These individuals really provide expertise and support to people who are otherwise without capabilities in very remote places in the United States and enable novice providers or uncredentialed providers to do things they would not otherwise be able to do and make significant impacts on patient lives. So, just by a quick show of hands, it's always good for the presenters to know their audience a little bit. How many people are clinicians or take care of patients on a regular basis? Good. And how many people have used telemedicine in their work practice? Oh wow, this is like the most experienced audience I think I've ever done this in front of before. Okay, well, clearly all of you know that we can deliver telemedicine. We do it every single day by using one of these just to communicate with other people. And I'd be amiss if I wasn't the first lecturer to at least give us a definition of what telemedicine is. So, telemedicine is really the use of telecommunications technologies that allows medical or even non-medical individuals to connect over space and sometimes over time if you're using asynchronous technologies in order to provide patient care at the point of need. And it's really been demonstrated to do a variety of things in a variety of different contexts. It's been able to help us improve safety. It's been able to help us improve problem-solving skills. It's enabled us to improve even morbidity and mortality in certain circumstances. It's particularly in the safety aspect by adding that second set of eyes or providing another opinion, maybe a second opinion for the clinicians that allows the telemedicine or the technology solution to improve the outcomes. So, if that's true, why doesn't everyone use it today? And it's probably because of this. These are two articles that have been published that have been trying to identify the impact of telecritical care on outcomes in patients in a variety of different hospitals. And what you see is a tremendous amount of variability in outcomes between different hospitals. And that probably, not probably, but most definitely reflects the challenges and the differences with which people implement this technology or they use the technology. Some people actually implement or use the technology better than others. And it's not people. It's really the system. But if the system isn't intentional in understanding how it's implementing the tool into their environment, then it may not work effectively. So, clearly, one size does not fit all. The complexities that exist in an intensive care unit between teams, team members, the team member makeup, the day of the week, the problems that those individuals are having, the patient census, the environment that they're working in, the restrictions on the environment that they're working in, the limitations of resources in the environment they're working in, both in terms of different medication shortages, in terms of different technologies, in different terms of budgets, all of these contribute to an incredibly complex system that if you don't at least understand some of those problems and what's important to the individuals that are within them, then you're likely to implement a technology that doesn't fit their work needs and consequently may not have the ultimate outcomes on patients that you are interested in achieving. And it starts to beg the question, are these the right outcomes that we're interested in studying for, you know, for telemedicine in particular, but really for all technologies that we implement in medicine? Because I would venture to guess that if you try to study those types of outcomes in these types of environments, large-scale natural disaster, war, even emergency situations, you'll find it's very difficult to measure the individual impact on that individual patient related to the technology that provides a capability that would otherwise have been unable to be utilized or a service that was otherwise not available to the individuals at the point of need. And understanding the access to that type of capability is probably more important than actually the outcome for the average patient population. So what does the future hold for telemedicine and for really technology in general for critical care? Honestly, there is no limit, right? All things that are possible today were once impossible. So really what is necessary is the imagination, the creativity of the individuals that are in this room, that are in the hospitals back home, understanding how a tool may help humans do work. And if the tool helps the humans do their work effectively, more than likely the outcomes associated with that tool will be better than they would be if the tool was not being used. If it impacts or deleteriously impacts the work and performance of the humans, then it's likely to cause unintended consequences and may actually harm patients. In fact, as you can see here, we have already started to imagine novel ways of using particularly telecritical care to enhance patient care in many unexpected locations. So using telecritical care during rapid response team activations, using telecritical care to monitor ER patients who are waiting to find a bed somewhere in the hospital, ideally in your ICU. And certainly we'll hear more about, in a few moments, using telecritical care to standardize the way we screen and monitor for life-threatening illnesses and how that can impact those outcomes. These technologies are incredibly adaptable, right? They're incredibly flexible in how you want to apply them in use. And if you think about all the communications tools and the data that's generated at the point of need, goes to the remote expert and comes back to the point of need, all that technology component can really be leveraged to create new solutions that may reduce risk or harm even to the healthcare provider. And this was demonstrated in the 2014 Ebola kind of scare at the medical device plug-and-play laboratory up at Harvard, where they literally provided patient care for about 12 hours without ever having a human individual enter the room, right? And the technologies, the automation tools, the solutions that come around that are incredibly adaptable. And if you think about it some day, these types of tools, when added to some of the other technologies that are kind of emerging right now, like additive manufacturing, drone delivery systems, robotics and automation, and ultimately artificial intelligence, are going to have a huge impact. Artificial intelligence, as we heard a significant amount this morning, is likely to be able to bring fairly advanced capabilities, not just to your community, but to your home and to places that we never imagined before, maybe even under a bush. As we heard this morning, artificial intelligence is really a fascinating technology that is emerging for not just critical care, but all of healthcare. And a lot of the data that we have is curated and accessible data sets. But telemedicine really offers an opportunity to us, because all the data and all the information necessary for a remote expert to understand how to make a good decision or how to help the local caregiver make a good decision and provide good care to a patient is the same data that a computer might need in order to understand how to make a similar decision, or at least promote some of the type of information or decision to a human to make a better one. And this is already happening. If you think about automation and artificial intelligence, it's not like snap, it's all of a sudden there. It takes time to develop and it grows over these small iterative processes, one element at a time, until all of a sudden you have a self-driving car. And it's already happening in medicine. You can think of multiple areas where medicine has been automated and intelligence has been applied to the technology itself. Think about the IV pump. How many individuals in the room still remember how to actually set up an IV pump or an IV system without a pump and deliver medications? It's few and far between. The technology does that for us. The fully autonomous implantable cardiac defibrillator. No human tells that defibrillator when to start or when to go off. It just does it. The artificial pancreas. These tools are coming to an ICU near you. But what's important to understand is that when you are implementing or developing these types of technologies, what the differences are between humans and machines. Because there are some things the machines are exceedingly better at than the humans, and there are some things the humans will likely always be better at than the machine. Value judgments. The individual preferences of another human being. Religious beliefs. Things of these natures are very difficult to teach a computer somewhat because there's no data, but also because it's just hard to work that out in a logical fashion. It's not logical. So what you need to understand is that as you're developing a technology, the most important component to focus on is actually how to develop the human-computer trust. And it's not always about replacing a human with a computer. Really it's about how the computer or how the machine may be able to augment the human to be able to do better work, improve their performance to enhance patient care and ultimately patient outcomes. And while Altuogluwande said men are fallible, but maybe men are less so, I think Missy Cummings, Dr. Cummings out of Duke, would probably say that yeah, and perhaps men and machines are even less so. And I think what we will pause, we'll do questions at a later time. Thank you for your time. Good morning, everyone. Thank you all for being here. Today my topic will focus on bringing the multi-professional team to the core. I have nothing to disclose. By the end of this presentation, I'd like you to be able to discuss how a multi-professional approach to staffing your tele-ICU can have multiple advantages and significant patient care impact, and also to list some considerations for implementing these services into your tele-ICU. As a tele-ICU pharmacist, I will primarily focus on my experience at Atrium Health, also discuss our virtual respiratory therapy program, and touch on advanced practice providers in the tele-ICU. Tele-ICU pharmacist services were first described in 2006 with Aurora Healthcare, who implemented a 24-7 order verification and clinical monitoring service for 246 ICU beds at 13 facilities. In a three-month period, they demonstrated over 1,000 interventions, cost savings of over $121,000, and a significant reduction in order processing time. Another model described is in 2009, UMass Memorial placed overnight clinical pharmacy services in the tele-ICU with order entry, drug information, medication reconciliation, and a particular focus on sedation management for 80 ICU beds at a single center. Compared to just the day shift clinical pharmacy services, the addition of tele-pharmacy resulted in a significant increase in overall therapeutic and sedation-related interventions, with an increase in sedation interruptions from 45% to 54%. Finally, a hospital in Germany in 2015 implemented daytime audio-visual rounding at three centers with the bedside teams and the tele-ICU teams, and were able to identify over 200 medication-related problems in 100 patients. This center described their tele-ICU services as a way to leverage services to remote hospitals in Germany without tele-pharmacy services. So, Atrium Health, our tele-ICU services were implemented in 2013, with tele-pharmacy services added in 2015. Our initial tele-pharmacy services covered 3 p.m. to 11 p.m. And then in 2017, we added a daytime critical care pharmacist Monday, Wednesday, and Friday from 7 a.m. to 12 p.m., who is a faculty member at a local school of pharmacy. We currently cover nine facilities and 176 ICU beds across North Carolina, although this has been as many as over 330 beds at 13 facilities across North and South Carolina. Tele-pharmacists review alerts for abnormal electrolytes, blood glucoses, lactates, serum creatinines, and triglyceride levels on propofol. And also during second shift review all new admissions during their shift. Our daytime clinical pharmacist has implemented real-time audio-visual rounds at two facilities, soon to be three. And our recommendations on second shift are made to the tele-intensivist via an electronic triage board. During first shift, our pharmacist makes the recommendations directly to the bedside. In a four-year period since implementation of services, over 21,000 interventions were performed, with 88% of these being performed on second shift. Our tele-pharmacy services have been able to leverage limited resources to provide critical care pharmacy services across our system, as many of our facilities have limited or no critical care pharmacists implemented into their daytime teams. This slide breaks down the number of activities per shift between first and second shift operating concurrently. Our first shift pharmacist reviews all ICU status patients at six facilities. So, her primary percentage of intervention are in medication management at 45%. And this is addition, discontinuation, dose adjustments, ordering labs, followed by stress ulcer prophylaxis at 18%. And code sepsis antimicrobial review for timing, appropriateness, and dosing in 15%. On second shift, our interventions tend to be more alert-driven due to larger number of patients covered. And limited opportunity to review all of those. Glycemic control accounts for 36% of all interventions, followed by blood culture identification in 21%, and medication management in 19%. This slide displays our Best Practices Award impact on glycemic control across the healthcare system. Glycemic control was chosen as an initial telepharmacist target, as it was one of the most commonly performed interventions by the teleintensivist prior to implementation of telepharmacy services. And was also a system-wide goal in the Quality and Safety Operations Council. As you can see, the bars in blue are the one-year pre-telepharmacy implementation period of the percentage of average daily blood glucose measurements greater than 180 milligrams per deciliter. And the red is the year post-implementation of telepharmacy services. And as you can see, there's a significant reduction in this number in the majority of facilities. With a total reduction of 18.8% to 16.1%. Another area in which telepharmacists have had a significant impact is on medication safety and the avoidance of adverse drug events. From October 2018 to February, October 2015 to February 2018, a total of 536 adverse drug events were avoided. The most common activities included discontinuation of medication and hypoglycemia-related interventions. And the most common medication classes included glycemic control agents, antibiotics, and anticoagulants. High alert medications, as classified by the Institute for Safe Medication Practices, accounted for 44.1% of these interventions. Cost savings reported in the literature vary widely. However, we have estimated to have saved over $890,000 to over $3 million with these interventions alone. A more recent scoping review by Hammond and colleagues puts this estimate as over $1,800,000. While we have used the scoping review by Hammond and colleagues, to assign cost data to our interventions, we have not performed a formal return on investment on our program. However, we do feel that we've demonstrated value on investment with our program, not only with the medication safety and number of interventions, but also with physician time savings. So, our initial service proposal incorporated physician time savings, as there were many tasks being performed that could be delegated to a pharmacist with physician oversight. In a two-year period, approximately 9.5 interventions were performed per shift, with a median time of 10 minutes per intervention. This represents 95 minutes of physician time that was reallocated during peak admission hours, minimizing or reducing the need for physician electronic medical record review, and freeing up their time for new patient evaluations and higher priority issues. Additionally, the telepharmacist had identified clinical concerns beyond just the bedside requests and best practice alerts, imparting a greater impact on patient care. Some of the successes of our program have been the provision of a system-level drug information and critical care pharmacy expertise in a resource-limited environment, as many of these facilities do not have critical care pharmacists embedded. In addition to some of the other successes I have mentioned, we have identified many time-sensitive opportunities for intervention in hyponatremia to prevent and mitigate overcorrection and osmotic demyelination syndrome, which have led to some system-level protocol changes. We had a rapid blood culture identification partnership for approximately two years with Antimicrobial Stewardship, which allowed us to employ timely antibiotic therapy in inpatients, emergency department patients, and our ICU patients. Finally, our day shift rounding has allowed us to increase our tele-ICU visibility and communication with the bedside teams. And we've been able to customize our technology to build several alerts and custom interventions that make data extraction easier. Some of the challenges have been justifying the cost of additional pharmacy staff. While we would like to ultimately be 24-7 to provide continuity of care, some of the fluctuations in the number of beds covered have made that somewhat of a challenge. We work with over 80 tele-intensivists located globally and various facilities and with three separate EMRs at one point. So providing consistent care has been a little bit of a learning curve. Additionally, we provide our data back to the facilities, but empowering them to take control of some of these interventions, such as glycemic control, something that we are looking forward to in the future to free up our time for focuses on things like sedation management and fluid management. We've undergone multiple technology changes, including migration from one clinical decision support system to another, and soon to be a third, and multiple EMRs. On day shift, it's been somewhat of a challenge to optimize avenues for intervention. However, our day shift pharmacist has identified a point person at each site who may be an intensivist, an advanced practice provider, or a pharmacist to get these interventions taken care of. Second shift can also follow up on these. Finally, we know that in the traditional ICU setting, telepharmacists reduce costs and improve morbidity and mortality. We have yet to evaluate whether this is the case in the telepharmacy setting. I'd now like to move on to our virtual respiratory therapists, or VRTs. At Atrium Health, this was developed through A through F bundle initiation with a focus on ICU liberation, and their current hours are 530 a.m. to 530 p.m. With this early start for a focus on spontaneous breathing trial optimization and assessment for SBT retrials, of which there's a very intense focus at four facilities. Patients are identified for virtual respiratory therapy review by tele-ICU nurses or respiratory therapists based on predefined criteria, including increasing oxygen requirements, new mechanical ventilation, and advanced modalities. Other responsibilities of the virtual respiratory therapists include ensuring protocol adherence and documentation, particularly with title volumes, assisting the bedside clinicians, serving as protocol superusers, mentoring new graduates, and also rounding currently with one facility but soon to expand to two. At the American Association for Respiratory Care, our tele-respiratory therapists presented this last fall. This was a collaboration between the tele-ICU and bedside nurses and respiratory therapists to establish a daily time for spontaneous awakening and breathing trials at a 14-bed intervention ICU over a three-month period. In that three-month period, 330 encounters were evaluated for optimization and implementation of SBT retrials. Of those, 57 SBT retrials were recommended, 48 were performed, and 32 patients passed, resulting in 31 patients being liberated from the ventilator over half a day earlier than they previously would have. Ventilator days in 2019 during the intervention period were 3.31, reduced from 3.75 in January of 2018. When compared to an 18-bed ICU of similar acuity, this was 4.39 days. Increased spontaneous awakening and spontaneous breathing trial documentation were observed, and there was also a trend towards increased ICU and mechanical ventilation length of stay. This is a graphical representation of the observed-to-expected ventilator length of stay. As you can see in the teal bars, in January, February, and March in 2019, there was a reduction from 2018 data. I'd finally like to touch on advanced practice providers before discussing some considerations for implementation. Advanced practice providers serve as physician partners for new admissions, documentation, order entry, family meetings, and procedural oversight. Some of the more common interventions described in the literature include evidence-based practice adherence, identification of sepsis, which Teresa will discuss further in the next section, hospital-acquired infection prevention, and serving as the first point of contact with the bedside in addition to code response. Some additional roles include e-mobile carts taking care to the patients outside of the ICUs, rounding, rapid response teams, skilled nursing facility and neonatal ICU care, education, and patient triage. Roles and staffing models are dependent on scope of practice laws, with some teleICUs having APPs directly embedded in the core and others using APPs as the hands of the teleICU at the bedside. Some considerations for implementation of multiprofessional services within your teleICU team include obtaining physician buy-in and setting the stage for acceptance at the local sites through site visits and discussions with those leaderships. Regulatory and legal issues are important to consider, so looking at your governing organizations and state boards in order to determine what's recommended. When assessing the facilities that you are going to cover, what sort of coverage do they have, what sort of staffing model, where are there gaps, and how can the teleICU personnel add to that? I know for our pharmacy services, the facilities that we cover with daytime clinical care pharmacy services, we have no daytime involvement in and limited second shift involvement. Are you covering a single healthcare system or are you looking across systems? Across systems may be a little bit more difficult to implement from a protocol and EMR standpoint. What technological capabilities do you have in regards to clinical decision support systems for alerts and identification of patients as well as communication? What are the goals of your additional personnel and what can they add? Cost reduction, length of stay reduction, decreased ventilator time. Pilot programs may be a great way to explore what these services can do in the teleICU setting. Our virtual respiratory therapy program grew out of two separate eight-week pilot programs, and I know that some teleICUs have actually used pharmacy residents to assess the role of the telepharmacist within their teleICU. Communication and collaboration are key, not only with the bedside but with the teleICU staff to ensure optimal patient care and ongoing revision to the program as needed. And assessing your return on investment and value on investment is important for justifying these positions and possibly expanding in the future. So where do we go from here? I envision the teleICU of the future to be similar to the bedside teams with incorporation of advanced practice providers, pharmacy, nutrition, palliative care, physical therapy, working with the preexisting system of bedside teams, critical care registered nurses, and intensivists. Additionally, I see roles expanding to include services such as more mobile and disaster care, system-level triage, and possibly even a role in post-ICU clinics to provide care to patients in remote locations. And with that, I'd like to challenge you. If you have multiprofessional services in your tele-ICU, please share what you're doing. We'd all love to learn from your experience. If you don't have multiprofessional services in your tele-ICU, which ones could your tele-ICU benefit from, and how could you consider implementing those in the future? Thank you. And I'll turn it over to Teresa. Good morning. So I'm going to kind of narrow us down a little bit on one topic. I'm going to talk about sepsis. And we know it's a hot topic. It's been a hot topic for years now. And unfortunately, we've gotten a little bit better, but we're still struggling. And many of that has to do with the fact that we don't have good, strong ability indicators and definitions to help us identify these patients early. I always like to give my disclosures. I do every once in a while get asked to speak places with professional organizations, and they pay my way. I didn't get my way paid today, but I always want to make sure that I disclose that. Today's objectives, I really want to talk about a conceptual framework around knowledge management, because I think that's what's really what we have to focus on here. There's just way too much to know, and I'm going to talk about that for any one individual to take care of all of the patients that they have to take care of in a given shift. And so I want to talk about how we can assess knowledge deficits, and then how we can create systems around that. And I'll do a little bit of defining the problem. And then I'll focus on how a virtual team can actually aid in that detection of sepsis, and even optimize not only detection of sepsis and outcomes, but focus on improving reimbursement. So you can kind of pay for your own program that way. And then I'll describe other health systems that provided me with some information that I can share with you today. We know, if you know anything about sepsis, you know that globally about 5 billion deaths, or 5 million deaths occur, and 31 million sepsis cases occur every year globally. So this is a huge problem in health care across the world. And we know it's very expensive. In the United States alone, there's been estimates of more than $24 billion annually. That's how much it costs. So we've got to get better at this. And we know that it's difficult to detect these patients. And we have different ideas of how to define it, and even how to treat it at times. And we know that every time we delay in treatment or identification, we increase the risk of death. So some of the things I'm going to talk about specifically around nursing, but I think these principles can apply to any discipline in health care, is around factors that contribute to errors that harm patients. So nurses have difficulty articulating the rationale for their practice patterns at times. Sometimes it's because that's the way they've always done it. It may not actually be something that truly is within the scope of practice. And I don't mean they're practicing outside their scope in a negative way. I mean, sometimes they're taking on tasks that others could probably do. We also have inadequate knowledge and lack of experience issues, just like every discipline does. We have new people sometimes. And usually the night shifts are saturated with our less experienced nurses. And that has been cited as a major contributor of high errors that harm patients, right? You just don't have the situational awareness and knowledge to make the right decision. And novice nurses, as I said, they just don't have situational awareness. They miss some of the cues, right? They don't have the big picture thinking yet. They get there at some point, but if we think about it as, oh, we'll wait till they get there, then we're actually putting patients at harm, then aren't we? We have to have a system in which they have an expert available to them to help them and make sure we mitigate harm. And then current evidence is not always used or even available when making patient care decisions. A big one I'm talking about this week as well on Tuesday in the Nursing Year in Review is the whole Joint Commission requirements around titratable medications, right? There's literally no evidence. We've been titrating medications at the bedside for decades and there is no randomized controlled trials telling us exactly how we should do it in incremental dosing. So again, sometimes the evidence just isn't there. Knowledge management is not something that nurses struggle with alone. It's across all industries. And this happens to be a paper that came out recently from the Mayo Clinic and how they've identified that the pace of medical discovery, the amount of research that's coming out has put caregivers to a point where they just can't keep up, right? There's too much coming out. In fact, I know one statistic, and this is old, I think it's about five years old, said for an internal medicine physician to keep up on the evidence, they'd have to read like 20 hours a day. They cannot keep up. There's absolutely no way to keep up on this. And on top of that, we've had more technologies that are more complex to use and electronic health records that hide the information that we need to see. So what do we do? Well, we use this conceptual framework around knowledge management. And knowledge management, it focus on acquisition, sharing, translation, and application of knowledge. Okay, I've already kind of outlined there's too much to know, right? So then you've got to have someone who's focused on that or an expert, and they can become your knowledge translators. Knowledge translation is used to close that gap, right, between knowledge and practice in order to improve those outcomes and improve adherence to best practices and clinician effectiveness. So again, I've said this before, there's just way too much to know and there's way too much to do. We don't have time to stop and think when we're at the bedside caring for patients. We're too busy running from task to task. This is a paper I published recently and focused on the idea of surveillance and human factor science and looking at what were some of the issues when we try to identify patients for sepsis. So if you take a patient who hasn't been diagnosed or undiagnosed, then you have everything from external factors related to political, government policy, and public awareness, right? Before CMS said you have to do something with sepsis, many people weren't doing anything with sepsis, right? And so we have these things that are external latent contributors. We have social environment. Is there good communication? Is there empowerment at the bedside? If I'm a nurse, am I empowered to screen patients for sepsis? Do I have good strong leadership teamwork? What's the group norm like? When I started talking about sepsis in 2004 at hospitals at Sutter Health, I'll talk about that in a minute, many had never heard of it. They thought it was some EICU thing. So really that whole social norm that accepts acceptance. Then you have the whole physical environment. Is it too cold tonight? I'm too hot, she's too cold. We're fighting over where the temperature gauge is gonna be, lighting, noise, other distractions, interruptions. And then you have the nature of the work. How complex is that work, right? We put people in situations to do complex level work and we do it in environments that do not allow them to think, right? We saturate them with tasks to the point that they can't even stop and think about what they're doing. And there's competing tasks, sensory overload. And then what tools and resources do we have? Alerting and alarming can be a good thing, but it also can be a bad thing as we all know. Are we doing paper documentation, electronic charting? Do we have screening tools? And then even at the individual level, how tired am I today? Did I get enough sleep last night? How am I at problem solving skills in general? Am I a novice nurse, am I an expert? Situational awareness and experience competency and all those things that contribute to our best working knowledge. All that is happening just to get from, I don't know if this patient has sepsis to this patient has sepsis, right? So this is not an easy thing. And this is why we need to focus on using knowledge translators and surveillance techniques. So increased complexity demands increased expertise. And so I'm gonna talk a little bit about my early years at Sutter and the work that I did there. And we as nurses actually identified sepsis is what we call in nursing a phenomena of concern in 2004. That's a long time ago. And we said, wow, in our health system, we had a 26 acute care hospital health system with two EICU hubs. And we were looking across 13 hospitals at the time. And we said, boy, there is disparate screening processes. Some people don't screen at all. Some people have never heard of sepsis. And no one's given fluids on time. No one's given antibiotics on time. We have a problem. And so we initially tried to go out and do education. I brought in, I actually got an educational grant and we brought in Kathleen Volman from the nursing side. We brought in Wes Ely, Mitchell Levy and Bryant Nguyen. And they came in and did education for physicians and nurses. Hundreds of people we tried to train and we still didn't get much traction. So then we thought, well, what if we use the tele-ICU nurses, right? We take this small group of experienced critical care nurses, most of them had their CCRN. Most of them had been nurses for about 10 to 15 years that were working in there. What if we trained them? What if we taught them and made them experts in sepsis? Could they close that knowledge gap, right? And could they help out? And so we did. We started doing manual screening every 12 hours and on admission. So when a patient arrived, we first started on paper. We literally downloaded the surviving sepsis campaign tool for screening, 161 beds and we had four nurses working and every new patient we were screening on a piece of paper. And we then worked with some of our partners in IS and IT and they developed a smart form for us in SharePoint that actually allowed us to collect data on the backend using Cognos. And then from that, we worked with our partners at VisiQ at the time, now Philips, but VisiQ at the time and we developed a sepsis alert. So we took the data we did on that smart form and that Cognos backend data collection ability and on over 5,000 patients, we were able to develop and train and test an alert that we could use electronically that is now used in many of the tele-ICUs across the country and then the tele-ICU nurses were influenced by this repetitive, continual and routine diffusion of evidence-based practice at the hospitals because they would call and talk to them about their patients. And this led to clinical decision support systems and other tools that allowed for decision-making and real-time auditing and feedback. And this is what our data looked like at the time. So you can see that in 2004, this is using Apache, I think it was four data. Yeah, this is Apache four data. And we were above one, almost at 1.2 in our ODE ratios for severe sepsis patients, admitted to the ICU with sepsis. And you can see that we came down underneath one and stayed down, got down to about 0.6 in our mortality. That's 40,000 patients in that particular graph. And then we said, well, how are we doing on our ICD-9 and 10 codes? Because we were finding we had more patients we were identifying in the Apache data set than we were identifying in the all hospital data for the ICD-9 codes at that time. So we started not only, you can see where we were screening. We still did see some reductions in mortality. And then at about 2008, 2009, we developed a tool to not only screen, but to have our providers in the EIC, our intensivist document all of the criteria needed to code this patient for severe sepsis so that we could get all patients who had severe sepsis coded appropriately. And that turned out to improve our ODE ratios on that. And then when you look at it here from this, we did an abstract that we looked at pre and post that particular intervention. And we demonstrated that we increased the number of times that we documented and coded a patient as having severe sepsis, which improved the bottom line. And I think we estimated at that time it was at about $2 million for about a 650 bed hospital. So I'm gonna talk a little bit real quick about a couple other facilities and how they targeted sepsis similar using experts in a surveillance model. So I was fortunate to have Michael Reese and Cindy Welsh send me some information from their organization. Michael's here in the audience. And they had a similar problem that we had at Sutter Health. They had clinicians who were not recognizing sepsis early enough. Most, as you know, the electronic health records, you don't see the alert until you log into the patient's record. How helpful is that if you haven't logged in in a couple hours and now you see that the alert went off a couple hours ago, not very helpful. And then the inpatient teams vary and the ED teams in their composition training and focus. Again, that whole knowledge management concept. So they wanted to find a better way of doing things. And so what they created was a process in which the sepsis alert fires. Clinicians don't have to be in the chart. And the central EICU tele-ICU team receives an alert, a sepsis alert, that allows them to go and focus on that particular patient. Nurses could grab that alert or the intensivists. And then it's reviewed by the EICU physician and they review the chart. And if it's appropriate, they agree that there's a problem of either deterioration or sepsis, they call the unit charge nurse. And then there's collaboration between the charge nurse and the primary care nurse on what are we gonna do next. They do huddle in real time when it's appropriate and they have a planned workflow for managing patients on the floor with these deterioration and sepsis alerts. And this is some of their outcomes data. So you can see that they had a reduction. So they had a hospital acquired sepsis, went from 36% down to 20% in the post-pilot. And you can see their average response time to the alert decreased. So they had alerts in the electronic health record that were going to everybody, right? I'm gonna interrupt you as soon as you open up the chart and make you do something that stops you from doing something else that you actually logged into the chart to do, you know what that feels like. So that was happening. And in this case, they were able to reduce the time that somebody actually, a clinician adhere to that alert, which is very important because we know that that's how we're gonna reduce mortality is early detection. They also looked at transfer, percent of patients transferred to a SNF and ICU patients for patients with sepsis. And you can see there that they had some changes over time there. In their stewardship for average direct variable costs also went down and their average ICU days and length of stay, they saw a reduction there. So again, this can not only help with your outcomes, it can help you from a financial standpoint. So some of the takeaways from the group there was to deliver the same science to every patient, right? And that's hard to do when you're trying to train hundreds or even thousands of people, right? In this case, you can have a small team of experts doing this. Consistent screening, rapid response times, appropriate treatment. Alerts outside the chart are a much better design for something that needs to be attended to now and not later. Huddles and really that collaboration and developing workflows around who's doing what. And of course, you always have to get that buy-in and engagement. Oshner, I have a friend here, Fiona Winterbottom. Thank you, Fiona, for providing some slides about the program they developed. They developed a deterioration alert in which the alert went to the APP advanced provider team. Now, they were doing this Monday through Friday at 9 a.m. to 4 p.m. for their initial. And their alert threshold, they sent an average of about three alerts a day. They focused on deterioration, not just sepsis, but they caught a lot of sepsis in that. And it was quite predictive and they felt that it was helping them in identifying patients. You can see that as the alert fires in the system, it can send it through Haiku, which is an Epic product. And then it can actually alert and give you a little vibration or tone on your watch so you don't miss it. So they were able to do some pretty cool stuff there. Really important, you have to have good standard operating procedures and you must define who's doing what, right? It's key. And this is what the APPs were doing for their proactive rounding. This was their role and they would review the list of medications and overnight emergencies. Because again, they were doing this as a daytime thing only for the pilot. They reviewed interventions in Epic. They reviewed IV access, oxygen, suction orders and did daily rounds on patients. And really focused on building the relationships with the units. And they became the liaisons for patient flow coordination. They're also the liaisons between the ICUs and the floors and the telemedicine intensive care unit. They have a very prominent role in making sure that there's a good connection and collaboration between the remote team and the originating site or team closest to the patient. They also did a lot of training. They helped to lead some of the interprofessional training. So you can see here, they've got simulation going on for deterioration of patients. And that helps to really solidify that. I can't say enough. You really should actually do simulation modeling before you turn a new workflow loose on patients, right? We don't do that a lot. We just come up with these workflows in a room and we think they're gonna be great and then we implement them on patients. Well, it's kind of like doing research on somebody. So doing the simulation modeling can help. And you can see with their data, they're not exactly where they wanna be yet. The green line shows some variation, but overall you can see that they have come down in their ODE. They wanna get underneath one, but for the pilot in this period of time after, you can see that they've improved over time. So again, Fiona promises me she has new data coming and that they're doing even better. So their program, they saw an 80% reduction in codes outside the ICU. That's important. 32% reduction in patients transferred to ICU from inpatient floors. 27% reduction in resuscitation for patients. And they saw a positive cultural shift in the hospital survey of patient safety, improving in areas where events were reported, actions promoting patient safety and continuous improvement. So on their culture of surveys for their patients' safety, they saw improvements. So overall, I think screening for substance is a resource intensive process, and it requires high level cognitive processing using good working memory. And you just can't expect everybody to do that well. It's silly to think that you can train all of the nurses in the hospital or all of any one team to be effective in screening and identifying sepsis. Until we have good biomarkers, we have to really focus on this cognitive processing and working memory. Expert teams working in a controlled environment with a specific role to provide surveillance. They're gonna respond to those alerts. Don't interrupt people taking care of patients. Stop interrupting people taking care of patients. Give it to somebody who can focus on that and provide expert knowledge and assessment. And then you need to really look at system designs to control for, enhance the latent contributors, that big graph I showed you of all the latent contributors. So really using human factor science and usability to make sure that these complex tasks are done appropriately. All right, and now we have time for questions. We've got about six minutes. Sorry I couldn't give you more than that. Please come up to the mic if you have a question, and we would be happy to answer those questions. Hello, I'm Rich Waltz from Columbia, Missouri, VA. At the VA, we incorporated telemedicine the last two years. The first year, it was free under a grant, and then subsequently changed to about $50,000 a bed. The VA system really is out of monies. But overall, have you shown that the monies saved regarding patient care, catching sepsis early, and so forth, pretty much makes up for the difference? In the private sector, I assume that this costs facilities also. Now, one can't take away from what is appropriate good patient care, but administrators look at the bottom line, too. Can you comment on that? Yeah, absolutely. So Dr. Craig Lilly and I from UMass, we published a paper, I believe it was in 2017, that if you would like, I can get your information and send you a link to that paper. It's publicly available. That paper explained in chest financial implications of doing experiments with tele-ICU over a 10-year period at UMass, and we looked at it at our medical center is where we started out, and we did a patient-by-patient cost analysis, and we found that we were about $50 million to the good over a 10-year period. That's after paying for everything, the install of the equipment and the bodies. I think if you also look at, there's a couple meta-analyses that have also been published on this topic, and what you'll find is that on the average, the cost is a positive return on investment. But again, it really will depend upon how you implement your system and the impact that that implementation has on their care providers, your work processes. Teresa mentioned the imperativeness of standard operating procedures and actually thinking about how those play out, and if you have the ability to do so, which is a great comment that you need to, as best as you can, model that before you actually put that out to the native environment or the natural world, would be better to do than to just implement a technology. The technology doesn't make money, right? The technology helps the humans do their work, and if the technology doesn't help the humans do their work, then the money doesn't get made. Just to piggyback, I apologize, piggyback on Teresa's presentation, that data from Applied Aurora, what we did to answer your specific question, per patient reduced cost by $878. So when we transferred that back to looking at the entire system of 27 hospitals, we now are initiating that in early warning systems centralized to APCs, planning to save about $4.6 million a year besides the fact that you're reducing septic shock, you're reducing mortality, and what's the cost of a life? Hi, those talks were great, so I just wanted to give you all that feedback. My name's Tina Shaw. I'm from Wellstar Health System in Georgia. We're actually on our journey to implement tele-ICU, especially in our rural hospitals where we have trouble attracting doctors and APPs. So I think my question is on the real implementation side. Do you have any advice on where are you locating your physical people, ratios for the APPs, the RNs, and the docs, and we're not quite at the level of thinking about pharmacy yet, but just any practical tips. I think there seems to be quite a bit of variation of where you're physically housing people and kind of what's the optimal ratio. Yeah, so there's many different ways people are doing it today. We started out with a very similar model where everyone had everybody sitting in the core or the EICU tele-ICU center, and that's now morphing to different ways. Some people have satellite hubs where the physicians, because they have trouble attracting physicians in their area, so they may have using a service in Israel. Phoenix was using, at one point, physicians also in Santa Monica, California to help staff their units. Generally, the nurses seem to be centralized in one core, but to me, I don't even think that should be a must-have. I think that even if you had a shortage of experienced critical care nurses, as long as they have a place where the technology supports them to do the work, I believe that a good, qualified, professional nurse could do this remotely from anywhere in the world. The Emory EICU has a satellite hub in Perth, Australia, so their folks go out there, and that was helping them to staff the night shift because they couldn't get enough people who wanted to, but yeah, I'll go to Australia and staff your night shift because it's daytime there. So many ways, I think, that you can do this, and you have to look at what are your resources, where you're at, in order to make those decisions. And I would encourage you not to rush as much as you can, right? There's a community out there, right? We have the telecritical care or tele-ICU committee here at this institution. The ATA has a similar one. Go get some experience with those individuals. Talk to them. If you can go, many of those places will let you sit a shift with somebody just to kind of get some experience, understand how that works, how their workflow plays out, and don't feel bad about asking for some help, and then, as best you can, I would say start a little slower, start smaller, and then get bigger. Don't try to go for the full thing as quick as you can. That's at least been some of our experiences. We've also done very distributed models of care, right, with nurses and doctors in various different places, which we experiment with things like virtual windows to allow you to maintain the team aspect and continue the relationship even though you're not actually physically next to each other. Multi-point conversations are a great tool, and the technology just gets cheaper and smaller. So our data was what Teresa presented at the end for the rapid response team, but I would say for anybody starting, use what you have. We actually intended to put APPs in our, we call it the bunker, but we started with that, but then our clinical needs at the bedside was so much that we actually ended up using those APPs that we had designed for one thing for something else, and so now we're going back to that idea. So I think it's what Jeremy said at the beginning. You go, you start with this. You can do anything you want, but you just have to take what's available to you. We've actually used some of our graduating fellows who are from other countries, from our Pulmonary Critical Care Medicine Program. They are now doing remote, they're doing remote services at night because they're in Manila, and it's cheaper, right, to do it for them at night for an odd, and we couldn't get, we couldn't hire doctors to work at night. They didn't want to, but they all know our system because they trained with us, so you can kind of make it whatever you need it to be, but as they said, we've got tons of people who have worked the system, so we can figure it out. Probably an important pitfall, though, is if you force it upon your environment, if they're not a part of the development of your system, it will not work. Thank you for the talk. I totally couldn't agree more that it has to be context-specific and modified based on your contextual need. So I have two questions. One, with the current technology in telemedicine, is there a context, especially in large academic centers or some other place like that, where you feel that it's not needed to go to tele-ICU right now? And second, how did you go about with the change psychology of telemedicine, especially in physicians? It seems like wherever you don't have physicians, you have tele-physicians, but in nursing, what you're suggesting is that you already have staff on ground, and then someone is watching over and giving you pointers or feedback, which is all good, but how does the autonomy and everything, how do you balance the change psychology? Yeah, and I've done this in two different health systems, so I have two different experiences to talk from. So the Sutter Health experience was a top-down decision by the CEO and president of the entire company, and so I was basically asked to make people happy with the service that they were receiving and paying for that they never asked for. So you can see how easy that task was. It was very difficult. And then about six and a half years ago, I was recruited to come out and work with Craig Lilly at UMass, and their approach was very different, much more methodical. They had quality issues in their seven ICUs at the medical center. The academic medical center is where they started. That's what they wanted to do, was solve problems there. They created a critical care operations committee that had representation from all disciplines within those ICUs, and even nutritional services and everybody. They would come to those meetings every other week on a Wednesday at 12 noon, and they still meet today. And they come up with the standard operating procedures, how we're gonna do business, who's going to do what, what's the role of the EICU intensivist versus the intensivist on-site, and on and on. And so that was a much more methodical approach. And their first paper demonstrated improvements in adherence to best practices, mortality, and length of stay. They did a subsequent one on the two smaller affiliated hospitals, which was independently reviewed by Price Waterhouse, all of their data, everything that they collected, and they demonstrated an improvement in outcomes at those hospitals, and they kept more patients there. So those patients then are getting more money, because they can reimburse for those patients they otherwise would have transferred to the academic center. So different ways of doing it, and again, at Sutter Health, we didn't have academic medical centers. We had large tertiary care centers all the way down to critical access hospitals, and it was implemented as a best practice and a leveraging of critical care expertise across the entire health system. So your question also related to not just how to implement, or if there's a place that it shouldn't be implemented, or it shouldn't be used. And I think one of the things that I learned from Craig early in my journey here, and he doesn't freely admit to this, per se, but the technology wasn't the magic sauce, right? It wasn't tele-ICU that was the magic sauce, it was the committee, right? And it was the methodical approach with a team of motivated individuals led by a very motivated individual himself that allowed that to be a successful tool. So it's really about the problems that you want to solve, right, with the tools that you have available to you, and then you implement the tools. So I don't think there is a space where technology will not help us do better work, right? I mean, we make better cars because we like to go places faster and easier. Eventually we're gonna like to go there without having to actually physically drive because then we can stay busy when we're in our car or pick an app or not waste that time of driving. And the reason that's growing is because people want that, because it is going to make their lives better. Good question. All right. We're a little over on time, but there's no other sessions coming in, so if anybody has any additional questions to ask, we take off, we appreciate everybody joining us here today. We'll stay up here for a few minutes afterwards. Yes, thank you. Thank you.

telecritical care

implementation

telemedicine

technology

healthcare professionals

patient outcomes

cost savings

collaboration

staffing models

standard operating procedures

Administration

Sepsis

Telemedicine eICU

Informatics

Presentation

Select

Professional

Associate

2020

Intermediate

Advanced

Sepsis Resources