Hi, my name is Wendy Wright. I'm a critical care neurologist at Emory University Hospital in Atlanta, Georgia. I'm happy to be here to participate in this debate about how to best care for patients who are boarding in intensive care units. I'm going to take the position that patients admitted to an off-service intensive care unit should be followed by the specialty service. I have one objective, and that's to present this side of the argument. If the question had been answered by data, we wouldn't be debating it. Rather, we're usually driven by either custom or sometimes even emotion. So hopefully we can bring a little bit of reason or rationality to this question. I don't have any financial disclosures. The opinions contained herein are not reflective of my employer. And I am a critical care specialist with a background in neurology, so I live and work in the world of the specialty ICU. I don't know if that's why I was assigned this position for the debate, but it certainly makes sense, and it does inform my opinions when we're making these arguments. Let's move past... Sorry, a little hair trigger here. There we go. Let's move past a few things that we're not going to be debating. We're not talking about whether units should be geographically based. They should. We're not talking about whether there's value in specialized compared to general intensive care units. There is. And we're not debating whether boarding negatively impacts the care of the patient, because it does. Rather, we're going to be talking about what is the better way to deliver care to a patient that cannot be placed in a specialty unit that best aligns with their admitting or most prioritized or most important diagnosis. We are going to move past those points, but we do need to first just touch on them a little bit to help inform our argument here. Geography is so fundamental to what it means to be an intensive care unit that it's basically in most of the definitions. It's where we come from historically, and it's really not easy to separate geography from the concept of an ICU. Equally fundamental and entrenched are the personnel that it takes to watch these patients closely and intervene if there's a problem, and the monitoring and life-sustaining equipment upon which they depend for these functions. Once you start concentrating critical care resources, you really can gain a lot of efficiencies based on economy of scale, and it's really convenient for the rest of the critical care team to have these things collected. What you'll see here is a picture from the 1950s in Boston of a polio intensive care unit. This is a very classic picture. You can see that the patients are closely lined together in their iron lungs, and you see the healthcare personnel and all their equipment. They're collected there so they can intervene for the better of the patient. Notice I did want to make a point of talking about the rest of the team because the nursing staff, which of course is a very important part of the intensive care team, is geographically based. Regardless of which side of this argument you take, the nurses are going to stay in the confined geography of the intensive care unit, so I have to make that special point to say the rest of the team because nurses are going to be in one specific place. Similarly, we passed over the fact that specialized critical care is good. There is value in having intensivists from diverse backgrounds. Diverse backgrounds bring a variety of diagnostic expertise. There's a lot of thought that can be shared. It's good to have these different clinical skills, and it's really helpful when we're trying to educate each other. The ICUs may improve outcomes by reducing diagnostic variability and focusing nursing experience within a unit. That way, the patient can go where there's a specialty unit team assigned, and that can be based, again, on their primary illness or their predominant health issue. Familiarity with these disease processes may also allow for better patient and family support. There has been shown to be reduced mortality in certain patient populations, especially in my patient population, the critically ill neurologic patient in diseases such as intracerebral hemorrhage. Certainly, there are arguments against specialty ICUs. There are common ICU syndromes that all ICUs can or certainly should be able to handle. If we start collecting patients aligned really tightly in these specialty ICUs, it can destabilize those care teams' abilities and their expertise and their familiarity with some of these common syndromes. That really can reduce some of the adaptability that we need to function when a different type of critical care patient comes into our unit because of resource shortages. That's where we get to boarding. Boarding is bad. Boarding is defined as caring for patients in an alternative location when a bed is not available in the appropriate specialty ICU. This could be the emergency department, and boarding critically ill patients in an emergency department is shown to lead to worse outcomes and increased mortality. For some surgical patients, this boarding could occur in a PACU, which has also been shown to increase mortality and increase in complications. Alternatively, the patient could go to a different intensive care unit. Let's just say as a critical care community, we can all agree that having a patient go to another intensive care unit rather than staying in the ED or staying in the PACU is much more desirable. We are critical care specialists. We believe in the critical care management of patients by intensive care teams in intensive care units. When we are driving these patients into alternative ICUs, it does bring up this question. Should that patient be managed by the local geographically assigned ICU team who may have less relevant expertise, or should the patient be managed by the diagnosis appropriate ICU team who may be at a geographic distance? Data are conflicting about where to best house the patients and how to best care for the patients, but there are enough reports of increased ICU mortality and in-hospital mortality and complications amongst patients who are boarding to raise concern. Even though boarding is bad, unfortunately, it is necessary. That's because we need to be flexible. One of the many lessons from the COVID-19 pandemic is that we really need to remain flexible. Here we see an article from the New York Times in early December when we thought things were really bad and we had no idea how much worse they were about to get, but essentially the headline says, there's no place for them to go. It talks about how we ran out of ICU beds. This really did drive critically ill patients into different care arenas and just brought this problem to the forefront. I will say that boarding doesn't eliminate our fundamental principles of collaborative co-management. In fact, it probably makes collaborative co-management even more important. It might change the teams with whom we are collaboratively co-managing. We can understand why there would be communication barriers if we're bringing in different team members or caring for patients in different spaces. That's the problem that we're going to try to overcome. This is what this issue looks like. This is an article that I have referenced and will reference several times in this talk. It just so happens the figure matches up with my side of the argument, so I'm going to go ahead and use it here. What you can see geographically here is a diagnosis appropriate ICU team will be in their appropriate specialty unit. They'll be taking care of the patient who is assigned there because that aligns with their most relevant diagnosis. My argument is that this same diagnosis appropriate ICU team should go to an alternative specialty unit if they have a patient boarding there. This diagram should go on to explain that it's not just the appropriate specialty ICU and the non-boarder versus the boarder. There are a lot of other things collected in the intensive care unit that aren't going to travel with the rest of the ICU team. We already touched on that a little bit. We have our nursing staff who are critical members of the team. There are protocols in place in these geographic units and equipment and people know how to use them and they're familiar with the disease processes. That's illustrated here in the green. Over here in the purple, I've put the alternative specialty unit where you do still have critical care RNs and protocols and specialized equipment, but it might not be the RNs and protocols and specialized equipment that aligns most with this patient's diagnosis. Why would someone think that's a good idea? I have three reasons that the specialty ICU team should follow the boarders. One, I think it's more patient-centered. Two, I think it is more consistent with the SCCM slogan of Right Care Right Now. And three, because I think it best identifies patients who really should be repatriated to their specialty unit because placement in the geographic confines of their specialty unit with the ICU team available and all of the nursing expertise and protocols and equipment is still the best approach. Going through those reasons, let's get back to our true north here. I think that it is more patient-centered to have a patient in their unit that would be ideal based on their disease process because a specialty unit implies some adaptation of personnel or procedures or equipment or all of the above that a patient would find desirable, that the patient's family would find desirable, that the care teams who are involved in that patient's care would find desirable. And let's bring a little bit of emotionality back into it. If you were the patient, if this patient was your family member, would you rather them be cared for by the neurology critical care specialist, even though their primary problem may be pulmonary or maybe an MI? Would you rather them be cared for a pulmonary ICU specialist if their primary problem is subarachnoid hemorrhage? So you see there, there is a reason why these specialty ICUs were built. Additionally, it is a little difficult to explain to a patient why they're in the wrong intensive care unit. I don't know if you've had to have that conversation, but I have. You might ask, why would that come up? It comes up. Patients and family members especially are tuned into what's going on around them. And it doesn't take them too long to figure out why they're in a specialty unit and they wonder why it's not lining up with the reason they're there. And that's not usually a comfortable conversation. It's a little bit difficult to talk about, not to take the argument to the point of absurdity, but we don't change out surgeons, specialty surgeons, just because a different operating room is available. And it really does raise the question, should patients have some form of consent to going to an off-service unit or should they at least have a disclosure that they're going to an off-service unit? I think that, as I had mentioned briefly earlier, I think that there is also a degree of patient and family support that is best backed up by the expertise and confidence of the specialty critical care team. So that's why I think this model would be more patient-centered. And let's, again, expand on that just a little bit. So here we have our diagram. Part of the ICU team, again, not the nursing staff, but MD, APP, PharmD, can follow the patient into the unit that they've alternatively been assigned. And at the very least, that would bring in the expertise of the other critical care team. Because if you keep it geographically based, what you do then is you've got, again, the team's familiar MD, PharmD, and APP, but they're all from the same specialty and it's not the specialty that lines up with the patient's diagnosis. Whereas if the specialty team follows, then you bring the right expertise, at least on the part of the physician, the pharmacist, the APP, other team members, you bring that to the patient. And then you don't have an entire mismatch of a team that that isn't their wheelhouse, that isn't their primary form of specialty. And again, I think that most aligns with the Society of Critical Care Medicine's trademarked logo, Right Care Right Now. And I got this right off the website, I'm going to read it to you. The society maintains that the right care right now is best provided by an integrated team of dedicated experts directed by a trained and present physician credentialed in critical care, an intensivist. Right care right now means the right care is delivered at exactly the right moment to achieve optimal patient outcomes. Now, if you're going to prioritize expertise, that would favor the care being provided by the specialty ICU service. Certainly if you are going to favor the presence or you're going to favor the timeliness, that would argue for care to be provided by the geographic service. But I'm going to give you, again, one more reason why I think that the specialty service is the better model. And that is because the specialty ICU service bringing in their expertise to a unit that may not be as familiar with the disease process is in the best position to determine which patients most need that integrated team and most need the timely access that can only be delivered in the appropriate specialty ICU and, therefore, are candidates for repatriation, meaning they're going to be transferred to their specialty unit when there is a bed available. I'm sorry. Okay. There we go. So repatriation will reestablish that coordination of care. There are some unfavorable aspects of it. This will increase the cost of clean rooms because rooms will be turned over more frequently and it will add to that downtime while the room is being cleaned. But I think that this can add flexibility when resources are stretched. And it makes sense to put a critically ill patient in another intensive care unit when there aren't any beds available. But when a bed opens up in that specialty ICU, I think that the specialty ICU team is, again, in the best position to decide who should get that bed. And that way, they're already following the patient. They've already established the relationship. They've already implemented the care plans. And the relationship and care plans won't be negatively impacted by the repatriation. And it'll also reduce the number of handoffs, which we all know are a weak point in communication. I also think that this provides the best opportunity to right-size the specialty ICU. We are not in a state where our ICUs are nimble enough where I can look at my 12-bed neurocritical care unit and say, you know, I'm really only using eight beds at this time. I'm just going to convert four of those into a cardiac intensive care unit or a COVID intensive care unit. Similarly, if I'm often running a census of 16 patients, I can't just flex up four beds. And I think having this information readily available to the specialty ICU team really helps plan for the future and how to get that right size. I think all too often, patients will be boarded in other units. And just to keep everything sort of flowing evenly, I feel like a lot of specialty services don't even necessarily know that there's a stroke patient in the medical ICU or a severe asthmatic patient in the surgical ICU. I think that the patients are really off of our radar. And having a specialty ICU service follow would help bring that connection back into what is the right size of this ICU. So based on those discussions, we have a lot of opportunities for improvement. I think that the starting point should always be rigorous, conscientious bed allocation strategies. Yes, we do want the patients out of the emergency department. We do want the patients in the specialty ICU if they can get there. But what are we going to do, again, in those situations when we just don't have the resources available? That's where I think aggressive repatriation comes in. I've already made my argument for that. I think that that little bit of discomfort, that extra cost and that extra time, even though it is causing some little pain and suffering, I think that's necessary to drive change. I think that otherwise, again, just whisking a patient out of the emergency department because you'll do anything to get a patient out of an emergency department, you're not really seeing the back end of that if you have the patient in the wrong unit unless you're really working to put that patient in the right intensive care unit. And so I really do advocate for that. Regardless of the solution, how you feel about this debate, I think that widespread education is really important to help stress the particular vulnerabilities of the boarding patients. And I think that we also need to promote a sense of responsibility for boarding patients. I think it's far too often that people feel out of sight, out of mind, or, oh, this patient isn't really in my wheelhouse, so I'm going to concentrate more on the patients who are assigned in my right specialty because that's what I'm comfortable with. And we need to, however this debate falls out, the team that is assigned the patient, whether geographic or specialty, needs to really understand that this is my patient too. And we probably actually should even move away from the term boarders, but that's what we centered this talk around, so that's why I've used it throughout. And I think the other improvement opportunity is to establish clear and readily available communication channels with a lot of options available between the boarding unit staff and, again, the rest of the ICU team, because remember the nurse is going to be with the patient regardless. So in conclusion, boarding critically ill patients in non-ideal or alternative units has been shown to be detrimental to patient care, but in our current state it's a necessary evil driven by resource limitations. Patients who can't be accommodated in their diagnosis appropriate ICU should be followed by the specialty service that can best address the diagnosis driving the ICU admission. I think we can work to ameliorate risks of boarding with educational efforts, and again, I think with aggressive repatriation, if that's possible in your hospital. In the future state really should include more reliance on telemedicine or other forms of enhanced communication to help minimize the impact of the geographic barrier. I want to thank you for joining the session today and I look forward to the rest of the discussion. Hello, my name is Victoria McCready and I'm from the University of Toronto. And I'd like to thank the organizers for inviting me to speak in the neuroscience crosstalk session about whether patients who are boarded in another ICU should go to the accepting ICU team or stay on the specialty service. I'll be debating from the geographic focus side where I posit that it may be better for the assigned ICU team to take care of the patient. I have no relevant conflicts of interest for this presentation. So the topic of today's discussion includes whether the accepting critical care team should manage outlier patients from the neurocritical care unit, or should the neuro ICU team extend themselves to cover extra beds and manage patients in a distant location. To do so, we will review different models of ICU care for boarders and evaluate the potential benefits of specialty ICUs and thus the problems with boarding outside of these specialty ICUs. And then the potential disadvantages of neuro ICU teams extending themselves to cover these patients. And finally examine specific neuro critical care interventions that may help to standardize care and help improve outcomes in boarders in non-specialty or generalized ICUs. So hospitals use a variety of strategies to maximize availability of limited ICU beds. Boarding, which involves assigning patients to an open bed in a different ICU is one such practice. Patients may be admitted to alternative specialty or generalized ICU. For example, a stroke patient may be assigned to a medical ICU and managed by either the diagnosis appropriate neuro ICU team, albeit at a distance or by the local ICU team who may possess or potentially possess less relevant expertise. Overall boarding disrupts the normal geographic co-location of patients and care teams and exposes patients to nursing staff and interprofessional team members with different training and expertise to those caring for non-boarders. And for full disclosure for this talk, I work in Canada and a rather different healthcare system to the US where the majority of critical care is delivered in general ICUs. The Canadian Critical Care Trials Group actually surveyed all acute care hospitals to assess ICU capacity after the H1N1 pandemic. And we can see that the vast majority of ICUs are mixed medical surgical, while only three ICUs in Canada care exclusively for neurologically injured or neurosurgical patients. So I work in a different system than the US where the majority of us work in general ICU. So this debate fits into a much larger discussion on the topic of self-specialization expertise in critical care medicine, whereby patients, nurses and physicians are cohorted together in an organ system specific ICU to optimize outcomes. Examples of self-specialization are seen in neurocritical care as well as coronary care, cardiovascular, trauma and burn units. And by pooling patients with similar diagnoses and procedures, specialty ICUs may benefit from economies of scale to improve efficiency. And these ICUs may also improve outcomes by reducing diagnostic variability and focusing nursing expertise within the unit. Within these specialty ICUs, we're focusing really on the physician team's expertise and whether the specialized neuro ICU team should care for borders outside of their ICU rather than the local ICU team. So with this in mind, we'll take a closer look at the evidence supporting specialized neurocritical care. This is a meta-analysis by Andreas Kramer and David Ziegen, including 19 observational studies and over 40,000 patients comparing neurocritical care models of care versus alternative models of care. And in their mortality analysis, meta-analysis including 18 studies here, we can see that overall care in a neurocritical care area or unit was associated with improved mortality. Similarly, neurocritical care favored a better functional outcome. It's worth though discussing some of the limitations of this existing literature. There was significant clinical and statistical heterogeneity and the majority of studies included were single center with historical controls, including introducing Hawthorne effect, as well as the secular trends that outcomes are just improving in general over time in ICUs. And the ones that were comparing to general ICUs, the general ICUs were an open model, whereas the majority of ICUs now work as either a closed or more appropriately named high intensity or intensivist-led ICUs. And then finally, publication bias. You're more likely to get published if you have a positive result showing an intervention improves outcome. But I would agree that overall, there's a large body of literature suggests outcomes are improved when care is provided in a neurocritical care unit or environment. But we have to acknowledge actually that what does the neurocritical care intervention include? And only about a third of, just over a third of these studies, it was actually a dedicated physical neurocritical care unit. It was also about a third included the introduction of a neurocritical care interventional team. And then another, just over a third, also included a standardized management protocol. And then just over 60%, it was the addition of a neurointensivist. Remembering again that many of these used historical controls from many years prior, and then moving forward, the introduction of one neurointensivist, which again, there may be some confounding with secular trends over time. But since most of the literature for specialized neurocritical care includes the introduction of a neurointensivist, doesn't this provide indirect evidence that a neurointensivist team, the service line, should look after these patients that are boarded in other ICUs? And although there isn't much data on this, we're gonna now have a look at the available studies. So the impact of boarding on neurocritically ill patients has not been completely explored. And the majority of studies have focused on emergency department boarding. And so in those studies, we can see that there's worse outcomes for critically ill patients that are boarded in the eMERGE, which includes increased duration of mechanical ventilation, longer length of stay, and higher mortality. And it seems the longer that you're boarded in the eMERGE department, the worse the organ dysfunction and the higher probability of a poor neurological recovery, especially in stroke patients. So Stretch and colleagues analyzed whether medical ICU patients boarded in an alternative specialty ICU were at increased risk for mortality, even when they were cared for by the diagnosis-appropriate ICU team. And they conducted a retrospective single center study using the MIMIC3 database. And here they used instrumental variable analysis to account for the non-random assignment of boarders. Because, for example, two patients are vying for the last available bed, the decision about whether to assign a patient to a boarder status is likely influenced by the unmeasured patient factors such as perceived severity of illness. They also used other modeling techniques like logistic regression and propensity score matching. And they found that in all three models that boarding and also looked after by the service line had an increased ICU mortality, a risk-adjusted mortality, and similarly increased risk-adjusted in-hospital mortality. So overall, we can see here that it seems to be the deleterious effects of boarding may extend to patients also cared specifically for diagnosis-appropriate ICU teams. So it doesn't seem that we can reduce the risk of having a worse outcome by taking the service line ICU team from the specialty ICU to come over and look after these patients in the outlier ICU. But I think this excess risk here seen here may also relate to the geographic barriers between physicians and patients and the difference in training and specialization of nurses between these specialty ICUs. This also brings up the concern about intensivist to bed or patient ratio. And this is a study done by Dara and Efesa about 15 years ago, looking at four periods within an ICU where they actually went through different staffing models. And overall, they could see that one to 15 staffing ratio was associated with a prolonged length of ICU stay, but did not affect the ICU and hospital mortality compared to one physician to 7.5 patients, one to 9.5 and one to 12. A more recent study here is by Gershhorn and colleagues who conducted a retrospective cohort study using data from the UK from 94 ICUs and just under 50,000 adult patients to look at what the optimal patient to intensivist ratio was and its association with hospital and ICU mortality. And they limited their analysis to ICU staff by one intensivist during daytime hours. And they find a U-shaped curve with the optimal patient to intensivist ratio was 7.5 with a significantly increased ICU and hospital mortality above and below this ratio. So the number of patients cared for by one intensivist may impact patient outcomes. So this leaves me with concerns about the service line of care extending out to look after these boarded patients due to the geographical barriers, the increase in intensivist to bed ratio, and then the differences in training and specialization of the nurses and interprofessional team within these different specialty ICUs. So we now look at the problems with potentially boarding patients in different types of ICUs, so non-ideal specialty ICUs and general ICUs. This is a paper by Lott and colleagues, and they were looking actually at the concept of whether care by the geographical team in the local ICU is associated with worse outcomes for actually six different admitting diagnoses. But I have focused on the two neurodiagnoses of ischemic stroke and intracranial hemorrhage. And they actually class patients on admission to either diagnosis appropriate ideal specialty ICUs, diagnose inappropriate, so non-ideal ICUs, or general ICUs. And after adjusting for important confounders, they found that there was no significant difference in risk adjusted mortality between general versus ideal specialty ICUs for stroke and ICH. However, there was a significant increase in mortality associated with patients admitted to non-ideal specialty ICUs. And so overall, I would say here that this study shows that there doesn't seem to be a difference for patients with stroke and intracranial hemorrhage who are admitted to specialized ICUs, so ideal specialty ICU versus general ICUs. But yet there seems to be an increased mortality when patients are actually admitted to non-ideal specialty ICUs. And so it does make me concerned about the investment in ICU admissions and my concern about the investment in ICU specialization may not be associated with improvements in mortality. So here, if we just look at, perhaps there's a specific neurocritical care interventions that can be brought to these general and non-ideal specialty ICUs to help improve outcomes. And these are some of the possible mechanisms for outcome improvement associated with neurocritical care interventions. We talked a little bit about dedicated neurocritical care units, but also the higher patient volume that they're seeing, the use of advanced neuromonitoring, perhaps a conservative approach to neuroprognostication and then adherence to management protocols. And so we actually focused on two of these looking at ICU structure versus processes of care to understand whether either of these was associated with improved outcomes. And here we did a retrospective cohort study linked with a survey that was sent out to each of these centers in TQUIP. And there was 134 ICUs in North America with just under 10,000 severe traumatic brain injured patients. And we looked at the delivery of neurocritical care either through a dedicated neurocritical care ICU structure or through the presence of a standardized management protocol. And each of these centers were very similar apart from one was care was delivered in a dedicated neurocritical care unit or care was in a general intensive care unit and the local ICU team cared for these patients. And overall, we find that dedicated neurocritical care units were not associated with lower risk-adjusted mortality, while a standardized management protocols were associated with an improvement in hospital mortality. And how might standardized management protocols improve care in generalized or non-ideal specialty ICUs? Well, they provide consistent care by helping to facilitate communication, reducing cognitive load, helping to coordinate the interdisciplinary team and increase the adoption of evidence or best practices and improve adherence to guidelines. So in summary, admission to a non-ideal specialized ICU appears to be associated with worse outcomes. However, mortality doesn't appear to differ between specialized and generalized ICUs. And the harmful effects of boarding may extend to patients cared for by the diagnosis appropriate specialty ICU teams coming out and looking after these boarders in distant locations. And overall though, this really doesn't address the important issue of the training differences and specialization of nurses and the interprofessional healthcare team between specialty ICUs really have focused on the physicians. And so future studies and next steps, I would say is to consider additional relevant organizational dimensions that may help to lessen the impact of boarding. So to really focus on the care provided rather than the label or name of ICU specialization to inform improvements and outcome in these boarders and other non-ideal or generalized ICUs and evaluate and standardize physician to patient staffing in ICUs would also be a helpful next step. And then understand if standardized management protocols really is low cost, but process targeted strategies may help to improve outcomes in these non-ideal specialty ICUs. So I'll say thanks very much and I will end there. Thank you.

specialty service

patient-centered

boarding

geographic focus

repatriation

protocols

ICU structure