My name is Erica Andrist, and as mentioned, I'm an intensivist at Mott Children's Hospital in Ann Arbor, Michigan. I'm super honored, and frankly, a little starstruck to be here at this session today, presenting with these folks. And I'm gonna be talking about, starting at the very end, how we make decisions to stop ECMO. I do have a couple of disclosures that I would like to make before we get started. The first is that I'm very grateful to have funding for some of the work that I'm gonna talk about today via the NHLBI. The second is more of a personal disclosure, and it's that I believe that it is not logistically feasible nor ethically permissible to continue ECMO support indefinitely. I make this disclosure for a couple of reasons. One is that, by design, my talk today is not gonna argue a particular side, and I would venture that many of us in the room have had the experience of consulting ethics and having them be like, well, you could do this one thing that's ethically permissible, or do this totally opposite thing that's also ethically permissible. Good night, good luck. And in order to partially mitigate what might be perceived as some of the wishy-washiness of my presentation today, I'm just gonna lay my cards down on the table. But the second, and I think more important reason that I make that disclosure today is that I am going into a project where I'm looking at some of these questions. Is it okay, is it not okay to stop ECMO? And I have this preconceived notion of whether it is or isn't, and I wanna be very careful to evaluate the data before us, the arguments before us, even if they don't necessarily agree with what I think. And as they say in ethics, good ethics starts with good facts, and there are certainly a lot of empirical data about this question that we do not currently have. So with those disclosures, the roadmap of what I will talk about today is gonna look at the relevant medical factors and ethical arguments that we need to think about when we're approaching a situation in which we might remove a child from ECMO support, especially when the expected result of removing the child from ECMO is death, sometimes, or what I refer to as terminal decannulation. After we look at some of those arguments, we are gonna look at some of the things that we don't know, our outstanding questions, our knowledge gaps regarding terminal decannulation, and then the last few minutes today, I'm gonna spend talking about an ongoing project that I'm a part of that is designed to answer some of these outstanding questions. A couple of current state assumptions that I would like to lay out before we get into the meat of the talk today are that one, ECMO is a bridge therapy, by which I mean ECMO is designed to support a child either to recovery from the underlying illness or to a more definitive therapy like lung transplantation. I certainly do not mean to imply that that statement isn't without some controversy. There is debate in the literature over ECMO as a destination therapy in and of itself, but today, I only mean to posit that ECMO in its current state is broadly considered to be a bridge to something else. The second current state assumption I'd like to put forward is that ECMO cannot be provided outside of an intensive care unit setting. Maybe someday it will be able to, but that is not today. Today, you need an ICU bed, you need ICU staff, you need ECMO equipment and personnel within the ICU setting in order to receive an ECMO therapy. And finally, why would we stop ECMO? I do think it's worth mentioning, given the depressingness, for lack of a better word, of what I'm gonna talk about today, that all of us hope when we cannulate a child to ECMO that the child is gonna get better. And so that is one reason and the goal reason that we might stop ECMO. But as we have likely all seen, sometimes that doesn't happen. Sometimes there's a catastrophic ECMO complication that requires separating the child from ECMO, and sometimes the child does not get better. And that in particular is what I would like to focus on in the rest of my talk today. So what do we know when we're approaching a situation like this? We know that there's medical uncertainty about the time over which a child may reasonably be expected to recover lung function. The current iteration of the ELSO guidelines for the management of venovenous ECMO say in section five on discontinuation that ECMO support may be considered futile if there is no lung recovery after 14 days of ECMO support. This is a data-driven recommendation. Some of you may have seen this figure before. This is from Tom Brogan and colleagues looking at prolonged venovenous ECMO in children via the ELSO registry. And I know this is a little bit hard to see, but you can see down here at day 14 is where ECMO survivorship drops off to below 40%. But this is where we start to see that this is a bit of a value-driven proposition as well. 40% is not 0%. And so to label ongoing ECMO support after the 14-day mark is futile certainly could be debated. And that's why that hasn't been a hard and fast rule in many centers even since this paper came out in 2012. And this question of how long is too long was really thrown into question with data from the COVID-19 pandemic. Multiple reviews out of China, Europe, the United States have demonstrated that COVID-19 patients had good survivorship after far longer ECMO runs than 14 days. This is just one that I chose from Russ and colleagues out of Germany which showed that ECMO survivors were supported for a median of 43 days, more than three times that 14-day support with good outcomes. And one thing I'd also like to highlight in this particular abstract as well is that even the non-COVID patients were supported for a median of longer than 14 days. So clearly this is something that is happening that we're supporting children for longer than that time mark. There are multiple different case series and case reports of adults who have been supported with good survivorship for longer than 14 days. In these case series, the longest patient that recovered from, that came off of ECMO without a lung transplant actually had an ECMO run of 151 days before native lung recovery was seen. So it literally an order of magnitude longer than that initially positive marker. And this is a study or a review, I should say, of patients at UT Southwestern, which actually showed in what they called their ECMO long haulers, there was better survivorship in that cohort than patients who required ECMO for a shorter period of time. So certainly some question over what duration of ECMO support may be considered reasonable and what duration of ECMO support may be considered futile. We know that because there is medical uncertainty over how long is too long or long enough, that there's almost certainly practice variation in between centers. And when we have practice variation, that leads to potential issues of justice and equity. If we have a person who is living near center A and cannulated there, where they will stop ECMO or start to discuss stopping ECMO after 14 days, they have different access to the potential benefits of ECMO than a patient who is cannulated at center B, who may let someone go in almost indefinitely on ECMO support. So the benefits may be unequally distributed, but I also wish to note here that the harms may be unequally distributed as well. We all know that while there are many potential benefits of ECMO, there are also many harms as well. So the direct harms to the patient of potential morbidity, of a potentially unnecessarily prolonged death, those are unequally distributed between center A and center B. And there are also an unequal distribution of the potential indirect harms. Indefinite ECMO is something that will contribute to ICU capacity. And we know that ICU capacity negatively impacts critical care access and critical care delivery. So if we have many patients on ECMO for weeks, months, perhaps even years at a time, that may impact care delivery to other patients and their ability to access potentially life-saving services, or potentially require them to take additional risk with longer transports to other areas where there are ICU beds open. Moving on to another principle of medical ethics, respect for persons, respect for autonomy and parental authority. We need to think about how we respect persons and respect the autonomy of children in the ICU. Children don't make their own medical decisions, right? It's one of the major differences between adults and pediatrics. So someone must make decisions for children for them. One of the ways in which we do that is called the harm principle, the idea that parents may make decisions for their children without interference from the state or outside entities so long as some reasonable threshold of harm is not surpassed. If that harm is surpassed, then the state or another entity may intervene to remove parental authority to make those decisions. And as we talked about, ECMO causes harm. We know this, but is that harm potentially worse than death that may result from removing a child from ECMO is more questionable. The best interest standard is also a consideration when we're looking at how we make decisions for children. This one I think probably most of us are more familiar with in the sense that we have an obligation to support the best interests of the child in terms of their safety, health, and wellbeing. Traditionally, parents are responsible for making those decisions, and we give parents a pretty wide latitude in making decisions for their child. But we as pediatricians have duties to the child as well. And what duties and what authority do we have to intervene and protect them from potential harm or to promote their wellbeing? These are also considerations that must be balanced. And this is particularly interesting when we think about how we make decisions when it comes to ECMO. I think many of us have probably come to have a more shared decision-making model for many decisions that we're making with parents and their critically ill children in the ICU. But most providers don't actually see ECMO as a decision that should be shared, particularly when it comes to initiation of ECMO, which we're gonna hear about in a little bit, or discontinuation of ECMO. Multinational surveys of ECMO clinicians have indicated that large majorities, more than three quarters of clinicians, believe that those decisions belong with medical providers. And I'm not here to argue whether that's right or wrong, only that it would require a really substantial culture shift within medicine, within critical care, if we were to adopt a model of more shared decision-making when it comes to these considerations about when to stop ECMO. And finally, when we're thinking about respect for persons, we also need to think about respect for the healthcare workers who are providing this service. In deontologic ethics, and Kantian ethics in particular, we are taught that we are not to use another person as a means. We are all ends unto ourselves, and it is wrong to use another person solely as a means to an end. And if healthcare workers are being used as a means to provide services like ECMO against their professional medical judgment, against what they feel is best, that may be a violation of respect for persons as applied to the healthcare worker. So there's those considerations when we're thinking about respect for persons as well. So even though I titled this slide, What Do We Know? I asked a lot of questions and talked about a lot of things that we don't know on that slide. And I'd like to elaborate on some of the questions that I think are really relevant when we're thinking about what is the right thing to do if we're looking at a child who's been on ECMO for a prolonged period of time. Under the bullet point of medical uncertainty and practice variation, we don't know how much practice variation exists between centers. I can only imagine that there is some, right? And anecdotally, many of us have probably heard different things that happen in different centers. But how can we quantify that? How can we measure that? And perhaps more significantly, does this practice variation actually translate into differences in outcomes for patients? Because that's a critical component of that question as well. We also need to ask, can practices as complex as ECMO decision-making be standardized or concretized within hospital policy? And if that is possible, what are the best ways to do that? From a justice standpoint, do prolonged ECMO runs alter ICU utilization patterns for other children? And if so, how? What is the appropriate lens for thinking about the harms and which harms matter the most? Are we thinking about the harms to the child? Are we thinking about the harms to the family, to the providers, the surrounding community, all of the above? How do we quantify and balance the many harms that could potentially result from prolonged ECMO support? And finally, what do providers think about this? What does the public think about this? And how do those opinions play out into questions of ECMO discontinuation and other kinds of resource allocation? Which brings me to the last part of my talk today where I'm gonna talk a little bit about the project I mentioned that is designed to answer some of these questions. It's called Terminal Decannulation from ECLS and Pediatrics, a Normative and Empiric Analysis of Current Practices. What this study aims to do is assess factors which inform decision making surrounding ECMO decannulation. It is going to delineate current practice patterns and that practice variation between centers and approaches to terminal decannulation. And it's gonna use that data to try to develop a set of best practice recommendations for people who are facing this certainly low frequency but high stakes and high acuity situation. What the study does is it is a bioethics supplemental research award to ASCEND, which if you were in this room about this time yesterday you heard Ryan Barbaro talking about the ARDS in Children ECMO Initiation Strategies Impact on Neurodevelopment Study. I'm pretty sure those are all the right words in the right order, but ASCEND for short. And ASCEND looks at neurodevelopmental outcomes in children who are cannulated to ECMO for respiratory failure and compares those outcomes dovetailing with the prospect study to patients who received conventional mechanical ventilation or high frequency oscillatory ventilation. Our study is gonna look at the smaller subset of patients who died while enrolled in ASCEND and particularly the patients who died following a terminal decannulation. It's gonna take a qualitative assessment of the clinicians who took care of those patients and the families whose children died after a terminal decannulation to figure out how the process went, what went well, what didn't go well, what support was needed, what support was ineffective to try to figure out how these practices actually look in the clinical setting, not just in hypothetical scenarios, but how they actually play out in the intensive care unit. We're also going to do a broader survey looking at current practices at centers, which will allow us to gather more data, although more superficial data from more clinicians who work in these spaces. And we're gonna integrate those data to develop that set of best practice recommendations for clinicians to use if they find themselves facing this decision. This is the study team. I've mentioned Dr. Barbaro already, who's the PI for ASCEND. We have multiple intensivists in our study team. ASCEND looks only at pediatric ICU patients with respiratory failure, but we have some neonatal intensivists and some cardiac critical care intensivists as well in order to try to make the data more, in order to consider the data from a broader lens. And we have some folks who are experts in ethics, qualitative research, and also research with bereaved families on the study team as well. This is where we currently are. Our IRB is currently under review, and I hope in the next couple of months we'll be able to start moving forward with the data collection procedures and hopefully present the results of this study perhaps at a conference like SCCM next year. I'm gonna stop there. I do have a questions slide, but it sounds like we're gonna wait until the very end to take questions, but I look forward to talking with you and hearing your questions at the end of the session. Thank you.

ECMO

pediatric care

ethical considerations

healthcare access

terminal decannulation