 Good evening, everyone. Thank you for coming to the NeuroEthics Seminar Series for tonight. My name is Tos Cochran. I am the director of NeuroEthics at the Center for BioEthics. The NeuroEthics Seminar Series is funded by the Center and by the Harvard Mind, Brain, Behavior Initiative Interfaculty Group and the Harvard Brain Initiative. And the webcasting is made possible by the International NeuroEthics Society. So welcome to anybody who's following us via the web. You can tweet questions or comments to us at HMS BioEthics. And we have somebody who's monitoring the feed. And if we have time for your question, we'd love to have it. What's our hashtag for today? The hashtag. Oh, there is a hashtag, actually. It's NeuroEthics, N-E-U-R-O-E-T-H-X. So yeah, get your phones out and start tweeting, everyone. We want to see that on the trending page for Twitter. So our topic for tonight is one that I find exciting as somebody who practices clinical neurology. The title is Busting Clots on Your Behalve, the Ethics of Presumed Consent to IV-TPA Inecute Stroke. And as you'll see, we've got a great panel that represents the spectrum of knowledge of medical ethics, expertise in acute neurological care, expertise in emergency medical care. And I think we'll have a great panel. Our first speaker is from out of town. He's Winston Chung, who is a neurologist at UCSF, who specializes now in neurodegenerative disorders in the Memory and Aging Center. But he also holds a PhD in philosophy from NYU. And if you're a philosophy fan, that's a factoid that will make you raise your eyebrows and say, oh. And so he's long pursued issues that are at the intersection of neurology and ethics. And he has got some interesting empirical data to share with us, but he also has a well-founded understanding of the ethics of the issues. Our second panelist is Lee Schwam, who is the Vice Chairman of Neurology and the Director of the Acute Stroke Service at Massachusetts General Hospital. He has very prominent roles at the American Heart Association and is a world expert in stroke neurology. He's very recently published his own views on emergency consent in stroke. And so again, a perfect commentator for our topic tonight. He also taught me everything I know about stroke. So if I do know anything about stroke, it's thanks to Lee. If I don't know anything about stroke, it's because I didn't pay enough attention to Lee. And finally, we've got Julian McGreevy, who I'm very grateful could join us tonight. He's an emergency medicine physician at Boston Medical Center and also holds an appointment in the Department of Philosophy at BU. He's a director in the residency program and the medical student clerkship in the emergency medicine. He holds a large number of degrees. Most importantly for us, a master's in bioethics. And as a resident, he was selected to serve on the Society of Academic Emergency Medicine Ethics Committee. So as an emergency medicine specialist with training and interest in bioethics, he'll be yet another perfect panelist to discuss issues of consent in emergency care for incapacitated patients. So thank you for joining us. And I'll turn it over to Winston. So how do I start my slides? Okay, and I should keep my time microphone off and just run off of this microphone while I'm at the podium. Okay, all right. All right, so thank you for coming everyone. I'm very glad to be here and thanks to Toss and everyone for the welcome and the introduction. So yeah, I'm here to present kind of the basic challenge we'll be discussing today and present some data that we've recently published on the topic. Hopefully, yeah, not to end the discussion but as a start for further discussion. Bansing the slides. There, okay. So first I'll say that I don't have any conflicts of interest to disclose and in fact, I actually don't even have what you might even consider a clinical conflict of interest, which is as Toss mentioned, I don't see these patients acutely anymore. My clinical practice is in dementia but I'm also a research ethicist and I have a general interest in issues around consent and consent in capacity. I wanted to raise what I think is a somewhat underexplored area in bioethics which has to do with problems of consent in what we might call neuro-emergencies and I've included ischemic stroke, subarachnoid hemorrhage and other hemorrhagic strokes, malignant cerebral edema, status epilepticus and I think these emergencies have a few kind of challenging features that make consent especially problematic and so for instance one, unlike many of my patients who have kind of slowly progressive disorders, these neuro-emergencies often involve patients who are acutely incapacitated. So many of us have cared for middle-aged patients with subarachnoid hemorrhage or ischemic stroke who were in complete command of their cognitive abilities until really the moment of their sort of catastrophic neurological injury and that has a lot of implications for the kinds of realistic advanced planning that are available for these types of emergencies as opposed to more chronic or progressive conditions. A lot of the interventions involved have a relatively stark risk-benefit profile so oftentimes we're talking about interventions where there's really a choice between dying or living but surviving oftentimes with severe neurological impairments and finally the time window for decision-making in these emergencies is quite narrow. So a lot of this reflects the particular kind of exquisite sensitivity of the brain to metabolic and other homeostatic insults so when we're talking about a stroke where cerebral blood flow is disrupted, we're really talking about potentially irreversible neural injury and loss of function on a time scale of minutes and this contrast with a lot of the other conditions that we see in medicine where a lot of the other organs are just a lot more resilient to that sort of injury. So if you think about the equivalent question in cardiology in addition to the fact that most of patients who have MI can consent, you're also talking about a situation where even though cardiologists try to move as quickly as possible there's still lots of cases where they can benefit patients by opening blood vessel 24 hours or so after someone's had an MI and the time scale we're talking about in these sort of emergencies really requires decision-making again on the order of minutes. Finally, I wanted to discuss the application of empirical data to address an ethical question so today we'll be discussing the ethical question of essentially what emergency docs and stroke docs should do in this sort of situation and it's always interesting to think about to what extent an empirical finding can inform that question. And I also want to talk about a particular method of population-based survey experiment that I think so far has been somewhat underutilized in bioethics. So in particular, if anyone here or on the webcast is a fellow or resident or another trainee who's interested in asking analytic questions in relatively large numbers of participants like thousands of people and also potentially interested in getting data for free there is a mechanism for doing that. So please come talk to me afterward. So I want to talk about the particular problem of thrombolysis with illytic agents such as TPA so-called clot-busting agent in acute ischemic stroke. I'll say also that although there are newer therapies and endovascular therapies that are less widely available a lot of them present similar ethical dilemmas and in any case most of the protocols for using those call for patients to first receive TPA with thrombolysis thrombolysis with TPA before going on to endovascular procedures. So we'll begin with by just noting that most patients with acute stroke cannot give informed consent. Consent for research is a little bit different a question than consent to treatment but still I think it's telling that when you look at these larger acute stroke clinical trials that only about a third of patients were able to consent to enrollment in the study and in general these were the patients with less severe strokes. And so in general when surrogates are available clinicians will often ask surrogates to give their best guess as to what the patients would have wanted but we also know that surrogates knowledge and accuracy about patients wishes is quite poor and in many cases we're really left with ultimately the same situation which is trying to give treatment to somebody where we don't really know what their preferences are. There's a particular problem with these thrombolytic agents in acute ischemic stroke which is that these are agents that break up blood clots and the intent is to restore profusion to an affected area of brain but because they break up blood clots they also carry with them the risk of bleeding and really the most feared consequence of thrombolysis is interest cerebral hemorrhage and which can produce early death. Here's a figure basically from a recent meta-analysis that indicates that use of TPAs associated with increased mortality particularly in the first week after an ischemic stroke. So the benefit here is functional independence and improved neurological outcome and with a guess that if you treat about six people you'll provide this benefit to one patient but then there is this significant harm of intracranial bleeding and early death where again if you treat about 40 patients with this treatment the risk is that you'll seriously harm one of them and then finally as we mentioned before there's a time sensitivity to the treatment benefit where if the treatment's not given within the first three or further out the first four and a half hours after a stroke then actually there's a greater risk of harming somebody than benefiting them. So in general presumptions of consent are important because overall we know that informed consent is generally a rather strict ethical requirement that's rooted in our respect for patients' autonomy but there are recognized exceptions to the requirement for informed consent. So as I mentioned before surrogate consent if surrogates are available but in many situations in an emergency department a patient might be found down, they might be witnessed by other people without a stroke but there aren't people present who can provide evidence about what patients' wishes are and so the question is whether consent can be presumed for such patients if we don't know much at all about what their preferences are and there's a lot of other contexts in emergency medicine where presumed consent is routinely applied and these are typically cases where we think that delaying treatment to see consent would jeopardize the health or life of the patient and where we presume that a reasonable person would consent to treatment if he or she were able to do so. So typical example would be a trauma patient who's in shock, who's suspected of having internal bleeding would be common practice to take that patient for an exploratory laparotomy because that's sort of an emergent situation and we presume that anybody would want this emergency treatment if they could be asked. So the question has to do with the presumption of consent to thrombolysis and it's worth noting that there's some heterogeneity among professional societies about this and we may discuss this later on but recently the American Academy of Neurology and the American Heart Association Stroke Association have endorsed thrombolysis as a standard of care and with that have endorsed this presumption of consent for incapacitated patients without surrogates. It's worth saying that this presumption has been controversial and one of the main things is that unlike other situations like CPR or treatment of a patient with acute blood loss, again thrombolysis is not in the short term setting life saving. So we're not talking about saving people and in fact in many cases, I shouldn't say many cases but in some cases treatment with thrombolysis may produce the death of the patient. This decision has been thought to involve a balancing of very personal values. So we're really talking about on one hand a benefit that involves increased functional independence and good neurological outcome against a different thing which is mortality and so it's been thought that patients who have different weightings of the importance they place on living longer as opposed to living without disability might have very different views about this. And finally, there hadn't been a study of stroke treatment preferences in a representative sample. So laying my own cards on the table, I'll say that really the study that I'm gonna describe really issued from a ethical disagreement between me and my co-author Anthony Kim. We both teach didactics to our incoming residents at UCSF, I teach an ethics didactic, Anthony teaches a stroke didactic and as these things often go, I had prepared this vignette that was intended to encourage a lot of debate and discussion about what should you do if a patient comes in and they don't have surrogates? Is it ethical to give a TPR or not? And they said, you know, Anthony talked about this case yesterday and we know that the professional societies say that it's okay to presume consent so that's what we're gonna say. And I should say that at the time I was quite skeptical of the presumption and so really the paper came out of discussions between me and Anthony and our colleague Andy Josephson about how can we best kind of answer this sort of question. So early in our discussions we realized that there were really two different questions that could be asked. So the first question is just, would older adults, the primary risk factor is age, would people who are at risk for ischemic stroke want thrombolysis in a situation where they couldn't consent? And initial problem that we came up with here is that if you just designed a survey and you asked people would you want this treatment or not, the natural question for us is, well what's the appropriate threshold that would justify the presumption? So would we be satisfied if a bare majority of people said, yeah I would want that treatment, would that be sufficient to presume that any reasonable person would want the treatment if we don't know what they want? And that seems too stringent, that seems to lax. There's other context where we insist upon say 95% confidence but that seemed like it might be too stringent and anyway it's not clear if you look at the legal standards, what is the right empirical threshold for answering this question about what's reasonable? And so in this case we decided to compare treatment with thrombolysis for stroke to a situation that we all generally accept the presumption of consent is warranted and that's CPR for sudden cardiac arrest. The second question we had is, well even if we knew that people would want thrombolysis it might be a different question of whether people would still want to be given treatment in a situation where they're unable to consent. So you might think those are two different questions that if I were just asked what I want this treatment that I might want it but that there's some people might object still the idea of being given treatment that they didn't have an opportunity to consent to. So I've included a table here from a really important study by Wendy Levinson where they ask people essentially what are your preferences about being involved in care and then making the ultimate decision. And what we see here, I don't know if I can use the mouse, so basically what we can see here is that a lot of people want to rely upon their doctors to make the decision, but still even among people who want to leave the ultimate decision up to their doctor that almost everybody wants to be asked their opinion. So we might imagine that there's some people who could object to being treated without having an opportunity to say whether they would get the treatment or not. So to get at these questions we designed a population based survey experiment and this is a relatively new type of study that's been kind of a source of innovation recently in the social sciences. And there it's been recognized that there's kind of this traditional dichotomy of different types of studies that have different weaknesses. So traditionally people in these fields might conduct broad surveys that would tell you about a whole population and have a lot of external validity but would only be observational in nature. Really they could just tell you about the relationship between different variables that they sampled people on. Or alternatively they would bring people, often these would be undergraduates at the college where they do work into a psychological laboratory where they might present different experiments. They might vary the presentation of different stimuli. They might give slightly different scenarios to people. And in that case they would have a lot of internal validity. They'd be able to judge how different presentations or different modifications of an independent variable modify people's response. But usually this would be limited to generalization among the convenience sample that they drew from. Again typically this is undergraduates at the college where you teach. So a population based survey experiment is a hybrid methodology that's meant to capture the strengths of both. And so basically this is one where you use population based sampling methods to basically you get a sample group that represents the target population of interest. And then however you don't just survey them but instead you basically control their assignment to different values of the independent variable to measure its effects on the dependent variable. So this is basically an experiment you conduct within a survey. So what we did was we sampled a representative sample of U.S. adults age 50 and up and we had over 2,000 respondents and we randomized them according to this fully crossed two by two factorial design. So the first factor is basically the clinical scenario that they were gonna imagine themselves being in. So half the participants, they're not patients, are randomized to imagine that they're patients with an ischemic stroke. And half of them are randomized to imagine that they are patients who are presenting with a sudden cardiac arrest. And then we second randomized them according to this other factor of question type where half of them we called hypothetical choice which is basically we present kind of risks and benefits and we ask would you want this treatment? And then the other one is the specific unable to consent condition. So that's where we say the doctors can't ask you what you would want. Would it be okay if they were to go ahead? Actually the precise phrasing is later because the phrasing is important but essentially would you want them to go ahead and treat you even though you can't weigh in on the decision? So to get in more detail these are the scenarios we presented. So for the half of participants that were randomized to the stroke scenario we presented the situation. It's basically a left MCA stroke that results in hemiparalysis and aphasia. And we presented this clinical decision-making tool that's based on one that's used at our hospital that presents essentially risks and benefits of treatment that include the greater likelihood of benefiting from treatment. Actually the greatest likelihood is the treatment not influencing the clinical outcome. Next to that is the greater likelihood of benefiting and then the smaller likelihood of being harmed by the treatment. And then for the other half of subjects they were randomized to the cardiac arrest scenario where they had a serious heart attack, the heart stopped beating and they would die without resuscitation. And here we presented, now we don't have data on the risks and benefits of CPR because no one's ever conducted a randomized trial of CPR. It would be unethical to do so. So in this case we could not present information about risks and benefits and instead we presented a similar graphic that presented outcome data on CPR with the assumption that untreated CPR is uniformly fatal. Or I'm sorry, untreated cardiac arrest without CPR is uniformly fatal. So, and then the question type again. So the simple hypothetical choice was would you want treatment with either this medicine or CPR for your condition? And then the unable to consent condition was basically imagine that either the doctors or in the other case the paramedics can't communicate with you, can't find out from family and friends what you wanted. Would you want them to go ahead and treat you? I guess any questions before I get into the data themselves? So the findings were actually quite interesting. So we published the findings on the first factor as a research letter in JAMA and basically there was really no difference. There was actually a slight difference in the intensity. So people can answer definitely yes they'd want it, probably yes they'd want it, probably no they wouldn't want it, definitely no they wouldn't want it. And so there is a slight difference where the people who said yes or no to CPR were more likely to really be sure that they didn't want it or did want it. But in terms of whether people said yes or no, there was effectively no difference at all. We had a number of different predictors for refusing treatment, demographic predictors in the case of stroke. These are consistent with other studies in demographically non-representative samples. And then one thing that was interesting is that we queried people about a series of health attitudes. We asked about attitudes related to patient-directed or physician-directed medical decision-making. We asked about attitudes about the importance of longevity in itself regardless of functional status or fear of dependence. And none of those were really associated with wanting treatment for stroke. They were associated with wanting treatment for cardiac arrest but not for stroke. And we did find these two other questions we asked. One had to do with whether people were confident in receiving care from the medical system and confidence was associated with wanting TPA. And then we had an interesting question we asked about whether you believe in using statistics to make important life decisions. And people who said that they didn't believe in using statistics were more likely to refuse TPA. And then we published the data on the second factor manipulation later that year. And essentially what you see here, so this is kind of the contrast of interest. This is the ischemic stroke hypothetical choice condition and then the condition where people are specifically unable to consent. And here there's actually really no difference. One curious thing we might discuss later is actually that CPR is more desirable in a situation where people are told they can't consent to it. And I have a few thoughts we could talk about later about why that is, but that would take us a little farther from our original purpose in meeting today. Okay, and then just the main thing to show here is just that this effect is pretty consistent across demographic groups. There's a little bit of association with marital status that could be a function of noise and just the number of different subgroups that we're looking at here. So we had to interpret it for the paper but I'm not confident in making grand claims about the meaning of those bars. So in conclusion, in our studies, about three quarters of older adults would want thrombolysis for acute ischemic stroke. This is comparable to the percentage of people who would want CPR for cardiac arrest and it's not diminished by asking people about a situation in which they're unable to give consent to treatment. And finally, it's not influenced by the sorts of personal medical values that have been hypothesized to be relevant to this sort of decision. So it's not influenced as far as we can tell by attitudes about the importance of living independently or living longer regardless of independence and it's not influenced by attitudes about patient or physician-directed decision-making. So our conclusion has been that this gives some qualified support to a presumption of consent in acute ischemic stroke. However, we also note that almost a quarter of people wouldn't want treatment and that for us, this is a source of caution, right? So in our case, it's sort of a reminder that the presumption of consent is really a pragmatic convention that we have for allowing physicians to give the treatment that most people would want, but at the same time that this is a defeasible presumption. So if we have other evidence about patient's wishes, then obviously, those should remain of primary importance. Potentially raising some further questions for discussion. I think it's interesting to think about consent and other neuro-emergencies. So here's some interesting data about another treatment, hemicraniectomy, for basically malignant MCA infarct with sort of malignant cerebral edema. So this is from a meta-analysis of studies in patients 60 years and under. This is from a study that was done in patients over the age of 60. And basically this purple bar is death. And so in younger people, we can see that hemicraniectomy, so basically taking off a piece of the skull and waiting for brain swelling to subside and then replacing that piece of skull is lifesaving. However, the greatest proportion of people who are saved are in this state modified ranking scale four, which I think Lee can say more about than I can, but basically these are people who are unable to walk and unable to participate in a lot of their self-care activities. So the question is, is this something that most people would want? Situation's even more dire in the case of older adults. So here again, the treatment is often lifesaving, but here there's no one in either group with a modified ranking scale score of two, which is still being able to carry out a lot of activities. And most people here again are really actually pretty much everyone who's saved by the treatment if you read these bars is somebody who's unable to walk and in many cases severely disabled. So I think there's been a lot of discussion within the neurocritical care community about the proper ethical approach to this question. I think another question that might be interesting to talk about would be about the potential time cost of the consent process, especially if we're talking about shared decision-making, which we know is a time-consuming process that involves exploration of the patient's preferences and the physician's interpretation of the choices available to the patient in light of what the physician knows about the patient's preferences, that this is a time-consuming process. And particularly in the case of endovascular therapies, it's been estimated that every five minute delay in restoring profusion to infected blood vessel is associated with one out of 100 patients having a worse functional outcome. So, so far we talked about presumed consent in the case where we just don't know and don't have the opportunity to ask patients what they would want. But there is an interesting and troubling question to ask here about is this a case where the consent process itself is potentially detrimental to patients? And what should we think about those cases? And a final topic that we might consider, so far only presented data about receiving treatment with standard accepted therapy for acute stroke. But given the really urgent need to improve our armamentarium of different clinical options in acute stroke, there is a great need for new research on different stroke therapies as well as different stroke systems of care. And the same problems emerge in terms of conducting stroke research. So we've already seen that most patients in acute stroke clinical trials are entered on the basis of surrogate consent rather than individual consent. And there's a general question that's been raised among investigators about whether a broad exception to informed consent would be appropriate in these cases. So, thank you. Should we just do individual presentations and then discussion at the end? Yeah, okay. Hi everybody. That was really a wonderful presentation Winston. I'm just jotting down a bunch of notes while you were talking and I'm gonna talk to you. I'm not gonna show you slides, so that'll be a change for most of you. So the first thing I wanted to comment on is that I think that with neurologic emergencies, particularly time dependent neurologic emergencies, it's really a different framework that we need to think about ethically. And we need to think about the harm of inaction or delay in a very specific way. And I just wanna remind you that during the time Winston was talking to you, if there had been a patient in the emergency room with an acute stroke, they would have lost 60 million neurons and 420 billion synapses during that 30 minute presentation, right? So, and 30 minutes may seem like a really long time, like oh, you could do whatever you need to do in two minutes, but if you have to track down a family member, if you want to explain this procedure with exquisite detail, the way you would with a patient who was gonna be undergoing brain surgery for brain tumor removal, you would spend 30 minutes going through the process, explaining the risks, the alternatives, how the procedure would be done, so that the patient had a fully informed and comprehensively educated experience after which their preferences and values could be incorporated into a decision about treatment. And I think part of the challenge with this field is that we've done a very good job in the last decade of really emphasizing the importance of shared decision making and preference-sensitive conditions. So if you have prostate cancer, it's really not clear what the best treatment is, because the outcomes are pretty similar, whether you have a radical prostatectomy, whether you have a radioactive therapy with seeds implanted, there are a couple of alternatives, or even potentially going on hormonal therapy. And to determine the best treatment for you, I need to know which of the four outcomes that I'm about to tell you about matter the most to you. Does preserved sexual function matter the most to you? Does longevity matter the most to you? Does incontinence matter the most to you? And I would argue that in stroke, in acute stroke, where we have very few efficacious therapies, and where we know from an abundance of data now, in real-world practice, that treatment is beneficial and that the faster we treat the more benefit there is probably to an exponential degree. Most of our data starts at 60 minutes after symptom onset, and we know it decays exponentially from there. What we don't know is whether the benefits in 30 minutes or 20 minutes after onset might be five, eight, 20-fold better than they are at 60 minutes, because there's very little patient data accumulated in that time window. So I would argue that we're really asking the wrong question. What person in their right mind would refuse thrombolytic therapy if it tripled their chances of being home and living independently three months from now? Right? So the real question in my mind to ask is, number one, is there a subgroup of patients for whom this treatment does not make sense? And in those patients, we should be investing a lot of time and energy in making sure they understand that the therapy may be of little benefit and does carry some risk, but that really we ought to be looking for informed refusal, not informed consent. Our goal should be, is there anything about your mother, your father, or you that I need to know that would make me think this wasn't a good treatment for you? Have you had recent bleeding? Have you had recent surgery that could put you at risk? You know, do you or your, does your father have advanced dementia? Does he have six months left to live? Things that would have a major impact on my decision about the appropriateness of this treatment. Because spending a lot of time trying to educate the patient on the process, the possible outcomes, the mechanism of the drug's action, the alternative treatments, there are no alternative treatments. So to talk a lot about the alternative treatments is to talk about standard care that we would provide anyway. So my feeling in this space is that we're doing the wrong thing. When somebody, when the fire alarm goes off, we don't have a long conversation about what your preferences are for smoke inhalation, right? We try to get you out of the building as quickly as we can, unless there's a really good reason why you wanna stay like, your child is still upstairs. And someone has to take that child out and you're not leaving till the child is rescued. So I would argue we're asking the wrong thing. We wanna be finding out, is there any reason I shouldn't be treating you? Not would you like treatment? The presumption should be that you would want treatment to improve your disability. Unless you started a position of extreme disability and have already made it clear that any further disability renders your quality of life undesirable. Now we could make a separate argument as a society about whether it's worth the $6,000 that Genentech now charges for this drug. Is it worth that to marginally improve some patients' life? That's a very different conversation and that's not a individual patient consent conversation. That might be a societal conversation about who we offer organ transplants to who we offer certain expensive therapies. As expenses go, it's actually a pretty cheap therapy. So I think the bigger question here is if that's the case, then why are we having this debate? Why are we having this conversation? And in this editorial that I, that was just published that Toss referred to, which is in the journal, one of circulation's subjournals, it's called Cardiovascular Quality of Care and Outcomes. It was a stroke-themed issue. I basically posit the following. Either the effect size is too small. So the frontline physicians who are predominantly emergency physicians don't believe the drug is effective and therefore it's not worth offering. And I think that argument doesn't hold a lot of water because number one, the effect size for acute MI with intravenous thrombolytics is much lower than the effect size for stroke. So that really, I don't think it really holds a lot of water. The second possibility would be that you really wanna incorporate the patient's preferences and that really this is a preference-sensitive condition. And again, I would argue, I don't think that if you look at the, what's actually happening that that really is relevant. Probably the most important one would be the risk of harm. And estimates of the risk of harm and stroke in many times are double counted. The figure that Winston showed was probably the best summary of the evidence in which you see of a hundred patients treated. Only three are red or pink color. That means there are only three patients who have increased harm attributable to the drug. People with really severe strokes do poorly. They have hemorrhages, they die. There was no difference in all of the randomized trials that TPA increased mortality at the expense of increasing mortality increased good outcomes. It didn't divide people in half like the Red Sea. It was everybody got better. Mortality was no different. So we're not in a situation where harm is being occurring as part of the drug. Predominantly, harm is occurring because of the disease process. So of a hundred patients treated, three will get worse from the drug and one will die from the drug. That's a 1% risk of mortality triple to the drug. Many physicians will cite a 6.3% risk of hemorrhage and mortality. Well, that is overemphasizing the risk of harm because there's a equivalent risk of mortality without treatment. So the idea that somehow the drug is causing the harm, I think is a direct reflection and not an unreasonable one of a individual physician's fear of giving a drug to an individual patient who will have a hemorrhage do poorly and they would be at risk for having committed malpractice, particularly if they're not confident about the diagnosis and the management of that patient. And so we had a unique situation here where I think we had a lifesaving therapy that had to be given immediately and the only person who's around at six o'clock at night or two in the morning is the emergency physician. And so in many centers across the country, they were picking up a telephone call of telephone saying, hey, I have a patient here. The radiologist tells me the CT looks okay, the guy can't speak and he's not moving that well, should I treat him with TPA? And the neurologist would say from the comfort of their home, yes. And then the emergency physician would feel okay, I'm giving a treatment based on a set of data that I don't really understand. I don't have the support of the institution, but I'm the one given the drug, I'm gonna get sued. And I think that in defense of emergency physicians, they didn't have the expertise and the support they needed. That was 20 years ago. TPA's been on the market now since 1996. We have telemedicine where we can provide acute stroke expertise within minutes to any hospital in the United States. And it's sweeping across the country. 25% of all TPA cases are treated at one hospital and transferred to another, often the hospital that supervises the treatment. So the idea that the emergency physician is out there swimming on their own is no longer a credible argument because hospitals can fix that problem without spending a minor amount of money. They can fix it cheaper than they can buy a new ultrasound machine for the emergency. So I think the concept that this is really about the overall risk of harm is also overstated. It's really the attribution of harm. And the irony is if you look in the medical malpractice literature of all the cases we know of that have actually been published, there may be one or two where someone gave TPA and there was a bad outcome. The overwhelming majority, 99%, are for not offering TPA, not evaluating for TPA, not giving TPA. So the idea that you're actually protecting your own self-interest by not giving TPA is also spurious. And I would argue with the increasing data that we have that delays in treatment relate to harm, the next wave of lawsuits could be you took so long waiting to talk to my father's, my sister, when you could have treated him the minute he arrived, the fact that he didn't recover is because you delayed the treatment unnecessarily. So I think we're really, we're in a place where certain ethical principles are being applied under a presumption that time is not important and that certainty and preference are the most important. When really I would argue it's the opposite. And now I think, as Winston pointed out, people have a different attitude toward judging the relative preference value of disability versus death. They would say, well, for that guy who I don't know, if it's not life saving, there's no imperative for me to provide presumed consent. I should only provide it if it's the guy's gonna die without the treatment. But if you actually ask older Americans what they fear most, they fear stroke disability more than they fear death. So again, I would argue the fact that death is the TPA is not life saving or has not been proven to be life saving again is not a relevant argument because if you're looking at preferences that matter to the patient, not to the doctor but to the patient, patients don't wanna remain disabled and TPA increases their chances of recovering from that disability. And then the last thing I would say is there's this concept, what about exception from informed consent? So we've had presumed consent. What about the idea of exception from informed consent? So we do this in the context usually of clinical trials where there's an expected benefit that would accrue in a scenario where consent is unlikely to be able to be obtained. Great example is intramuscular or rectal or sublingual benzodiazepines for patients in status epilepticus. So trials that have been done to try to determine which is the best agent to terminate status epilepticus. There's no opportunity for consent in most of these cases. The patient often can't provide the consent. And so by definition, the patient can't and often there's no one else available. And so because there's an expectation of significant benefit from receiving a medication in this scenario, the community can be solicited for their input and a clinical trial can be conducted under exception from informed consent or ethic is sometimes abbreviated. Now I would argue that's worth thinking about not because I'm proposing that we move to an exception from informed consent. But if you look at the statutes that govern when exception from informed consent would be reasonable, I would argue that TPA treatment would meet all of the requirements. So in a research context, exception from informed consent would be a very reasonable way to conduct a trial of thrombolytic therapy. So I think I just, I'll summarize now. The harm that's due to delay, I think greatly outweighs the benefits that come from a much more fully informed conversation about consent. If you look at data from Get With The Guidelines, which is the largest U.S. registry of TPA treated patients and has thousands and thousands of patients, many more than the clinical trials, patient refusal or family refusal as the reason for no TPA is a very rare event. So in centers where patients are refusing, those rates are incredibly low. So the argument again that during these fully informed conversations, lots of people are refusing, also doesn't hold water. So really the risk here is that by decreasing the rates of early treatment, you're affecting everybody. So you're harming everybody by having a lengthy consent process. Not in order to save a very small sliver of patients from the lost opportunity to refuse TPA if they had only fully understood the risks and benefits. So I am reminded of the Salem Witch Trials where the way you figure out who's a witch is very simple. You tie the woman up with a bag of stones and you drop her to the bottom of the river. And if she floats, she's a witch. And if she dies, well, that's unfortunate, but that's what you have to do to find the witches. So you're essentially packing everybody with a bag of stones and dropping them in the river while you have this prolonged consent discussion for the one patient who's gonna float on the surface and refuse their TPA. So that's maybe a little bit of an exaggerated, but still Halloween relevant. I figure out how I could work Halloween into this. So I think what I will do there is to summarize by saying, if we really believe that delay is the major driver of harm here, not the risk of false consent where the patient would have chosen differently or the family would have chosen differently had they been fully informed of the risks, then our job ought to be to start with a presumption of consent and do everything we can as quickly as we can to find out if there's a legitimate grounds for refusal. And in the absence of that, give the drug. And I would argue you ought to be getting ready to give the drug while you're doing this so that for the 90 plus percent of patients in whom you're going to end up giving the drug, no delay has been introduced by this activity. And in the rare case where there is a decision not to treat, at least in the United States in 2016, the company that provides the drug will replenish your supply if in fact you mixed drug but were unable to give it because of a reason for non-treatment. So I think our job here is really to formulate an ethically consistent position that supports the practitioner in operating under a model for presumed consent and that if we can solidly articulate that, accompany that with a sort of standard language that is something that physicians feel comfortable using as a single standardized discussion format, we would have made a tremendous amount of progress. We did for a very brief period of time before the American College of Emergency Physicians changed their position statement based on another re-analysis of the same data, not any new data. For a very brief time, the American Academy of Neurology, the American Heart Association and ASAP came together on a single risk benefit statement for TPA that could be used as a tool for conversation, but unfortunately that was short-lived. So my hope would be that if this accomplishes something, it would accomplish, it would be the generative seed for trying to come up with a uniform statement that all practitioners would be comfortable. So those are my thoughts. Thank you very much. So it's really a pleasure to be here. Thanks for the collaboration. So in thinking about this talk, so I have a few notes kind of prepared in advance and then I took a few notes as you both spoke. And in our conversations in advance, we sort of thinking about what my perspective is gonna be, kind of what my role is here. So I'm not a stroke specialist among emergency physicians. There are some people who are and who come to this, I think as Dr. Schwum sort of alluded to with very strong views one way or the other about TPA as a drug and whether it's something that we should or shouldn't do. I'm not one of those folks. What I'm trying to represent really is sort of your average emergency physician who's trying to figure out in this kind of mess of data, what do you do with that? And then beyond that, as somebody who is a residency program director, trying to teach emergency medicine residents, well how do you sort through this through this data and what do you do with that? And then I think as the third perspective is really sort of somebody who does clinical bioethics. So I think like many of you, I do sort of carrying the pager clinical ethics and get involved in a lot of the ethical issues that physicians throughout the hospital run into. And I find it in general, we're not very good at the common ethical issues. We just struggle. We don't teach medical students very well. We don't teach residents very well. And I find that very frustrating. So that's kind of my perspective on this. The one thing I will do is just sort of introduce a little bit of my understanding of the controversy within emergency medicine, only because I think it's morally relevant. I think it's hard not to bring that up. And the caveat is that I'm not an expert in this in one particular side or the other, but have I think like your average emergency physician read a lot of this stuff in various journals of emergency medicine such as Annals of Emergency Medicine, Academic Emergency Medicine with the Big Two and then plus all those sort of throwaway journals that we get where this is like the hot topic, the big bad controversy, as it's sometimes said. So I think it's important to bring that up. So I think to start, emergency physicians face many situations in which implied consent is necessary. We intubate patients in the middle of the night. We take patients to the aura as mentioned when they're a victim of a gunshot. We do a chest tube in the middle of the night. And there's assumption that it's really our more responsibility to do that. So it's not that we're somehow overriding informed consent in some simple pragmatic way, but we're sort of looking at a modified version of it, which is to say that it is in fact our duty to not let this patient be excluded by virtue of the fact that they can't tell us that this is what they want. So it's really, I think, a positive duty. So what's so unusual about it is why is TPA then singled out as a separate kind of thing? Why is it not like all the rest of them? And Dr. Schwum sort of alluded to the kind of decades history of this, which he's certainly much more familiar with than I, but it's so odd. And I find it, so tonight after this, I work 11 p.m. to 7 a.m. in the emergency department. That'd be two hours, well, I'll be the only attending in one of the busiest emergency departments in New England. And the question is why is it that I now need a struggle to sort out if a patient comes in, whether this is something that they should or shouldn't get. And obviously I'm not in the win by myself. I have neurology colleagues to help, but still. It's the emergency physician who recognizes the stroke or takes the call from EMS and is ready in the room when the patient arrives, gets the IV in two minutes, gets into the scanner in three more minutes and makes it so that everything happens quickly so there's no delay, so someone's not harmed by the failure to give this thing. So why then in the back of my head do I have to wonder, am I doing harm or am I doing the right thing? And that's what I find I think the most frustrating about this whole thing. So I agree with the point that time is the biggest factor. And I think all emergency physicians would agree with this as really the central issue. And what then becomes a challenge is how you incorporate patient's values, how you incorporate uncertainties of the data to the extent that they're real and that's not entirely clear to me. And that's why I think that Dr. Chung's work is so fascinating. I think more broadly all of us who do ethics and emergency physicians specifically would love more data to sort out some of these bioethics issues. Empirical bioethics has a lot of power. As it is, we borrow most of our data from the critical care literature because we don't have all that much of our own. And we try to use that in those moments where we just have a few minutes to make a decision. In this case, as you had mentioned where millions of neurons are going away or years of disability, the disabled life are creeping in. I think to start with kind of looking at some of the data, we had these slides in advance and I kind of thought about what is the meaning of the work that you've done and in the middle of the night tonight, if somebody comes in, what am I gonna do with that information? Knowing that 75% of people would want TPA, 25% wouldn't. It's the same as what would want CPR and what wouldn't. What exactly is it that I do with that? And some of the things that struck me about it and these really are open questions. I don't know the answer to these and maybe to some extent you can answer it having kind of understand that the data a little bit deeper. But one is why is it that only 75% of people want CPR? But you're using that as the standard that everybody would want that, barring them having made some informed decision advance that they wouldn't want to be resuscitated. That basically everybody would want that. So why is it only that 75% of patients would want that and that makes it difficult then to use that as the benchmark against which you're kind of judging TPA? So that's something I'm struggling with a little bit. The other thing is although at the end of the day 25% in both groups didn't want whatever the intervention was 75% did the reasons for wanting it or not wanting it were different. And I find that personally more morally relevant than the number itself. And again I don't know what to do with that information. But so for example you had mentioned that in TPA that lower education which presumably correlates in some way with lower socioeconomics which as you know in my hospital is a big factor. I think you have very vulnerable patients all over the place. And it becomes an extra challenge about values and the values of physicians creeping into some of those decisions. So why is it that lower socioeconomic status makes somebody as I understand it sort of less likely to want TPA? And why is it that it's a matter of their confidence in the statistics as well? Which I find a little bit unusual. And how you overcome that? One thing that I find, I'm sure a lot of you have experienced this too when you do an ethics council. Whenever the team tries to give more data it usually doesn't resolve the issue. You know, it just doesn't. We'll show you the scan. We'll give you the results of the test. We'll show you the study. Doesn't resolve it. So this I think goes to a deeper issue about lack of confidence in some of these numbers that is more broad. Whereas the contrary of that, you know your counter example which was a comparison example, sorry, of the cardiac arrest where people feared poor health and dependence. You know what Dr. Schwam as you mentioned is true in the setting of stroke as well that people fear dependence. Why didn't that come out in these data? Where that was the case in both TPA and CPR. I think these are really important questions to try to sort through. One guess that I have is that there's a whole lot of experience with CPR. We've been around for a while. Lots of conversations about whether or not this is something that we want or we don't want. So people have a way of sorting that out into their own values. Whereas TPA, maybe they do, maybe they don't. The other question is whether or not it has something to do with presenting data in terms of specific risk benefits such as your disability might get better, you might bleed, or outcomes. As in your study you had divided it a little bit in different ways. So I think that's an open question that I think people would be somewhat interested and probably have some thoughts on. So those are I think some of the initial things that struck me about that. Now I wanna talk a little bit about the controversy of TPA in emergency medicine. And as I prefaced, I'm not an expert in this and I also am making no attempt to represent one side or the other perfectly well. But one of the things I think that your average emergency physician struggles with is as Dr. Schwalman alluded to, these guidelines have changed. So in 2012 the American College of Emergency Physicians had some guidelines. So these guidelines were put together by a group of emergency physicians and a group of neurologists together. And they basically were in line with the American Heart Association guidelines which said within three hours the TPA is the standard of care you should give it. The new guidelines, 2015, say TPA, maybe you should give it. Now this is a big deal. This is a big difference. Because I think on the first hand in 2012 your emergency physician say the folks who thought there was a problem with TPA or anyone who maybe had no thought about it whatsoever felt like they had to give it. And to not do it was wrong, was harm, right? Which very well may be true. Whereas now it's sort of up to them. What exactly does your average emergency physician do with that in the middle of the night? I think that's a real problem. And how you solve that I think has a couple factors. So one is getting a little bit more clear about the data although given that we have randomized trial and we have pooled data in the real world that seem to support and we kind of decided for this talk we go on the assumption that these data are perfectly fine and demonstrate that the benefit far outweighs the harm then we're probably doing okay in that respect. So I think it's a little bit of getting people, everyone who's involved in this, emergency physicians, neurologists, EMS providers to kind of decide what it is that our system is gonna be and make sure that we make that system work. I think that that's a really critical factor. To go back to those 2012 numbers just to show you that I'm not randomly citing concerns of just a few people. There was a poll right after those 2012 guidelines of emergency physicians. I don't know the details of kind of this poll but 60% of emergency physicians thought there was a problem with those 2012 guidelines that needed to be rescinded. So a big number. Which I think is part of what led to those 2015 guidelines. You could argue on the one hand that they're sort of reactionary to what may or may not be an actual representative concern or a valid concern. There are equally concerns about the 2015 guidelines which is that there's only emergency physicians, there are no neurologists on this. And some would argue that none of the emergency physicians on here are people that really specialize in stroke within the realm of emergency medicine. So in both cases I think people have grounds that a lot of people would find relevant and valid to be concerned about this 2000 and 2015, or sorry, 2012 and 2015 guidelines. The other question that I have is something about the way this stuff is discussed in public. To some extent I think there are some examples of discord even in the lay press major newspapers involving neurologists and emergency physicians about sort of care in the emergency department. And I think that has a real problematic impact too on how people interpret these data. So it's not just a disagreement on mug emergency physicians or between emergency physicians and neurologists in their own literature, it's actually in the lay press. And so whereas when you look at CPR there's not a whole lot of disagreement about whether or not people should or should not be doing CPR as a clinical matter in the lay press but there is in this case. And I think that's one of the things that sets it apart from other things that emergency physicians are quite comfortable using on the basis of implied consent. So one of the things that I did sort of in preparation for this was sort of talk to some of my colleagues, both who do EMS who do stroke related stuff and sort of everybody else sort of in my department to think about what are some other areas of implied consent that we might compare and contrast this to. So as I said, there's so many emergencies that you face. What is it about this that's a little bit different? And one thing that comes to mind that's of interest is therapeutic hypothermia which is another thing another instance where the guidelines have somewhat changed a little bit. Now this is a case where the data aren't amazing. So it seems as if the data for this are far better than that. That therapeutic hypothermia, there's a small studies a couple hundred people, found for people who have shockable rhythms, the FBT and seem to show benefit, kind of try it with a sisterly, it doesn't really work that well, try it with different temperatures to find out what works and doesn't. And seems like as my understanding is that the most recent guidelines now in the pre-hospital is you may not cool people say with cooled saline. This is a change but there's not so much of an outcry about the fact that there's this change. So there's something unusual and different about TPA that I think needs to be explored a little bit deeper before we can come to the kind of consensus that will keep us from second guessing ourselves in the middle of the night. And the last thing that I wanted to mention is the issue of vulnerability among patients. This is something that is kind of my main area of interest because it's kind of what I struggle with the most. And another of your bioethics that I find very fascinating is finding out what physicians' values are and how they have an influence on the decision. And it's fairly clear that nationally the things that physicians believe affect the choices that patients ultimately make. Physicians' knowledge about ethical issues affects the choices that patients also make. So a lot of times physicians will have misconceptions about things like trials of therapy. They'll say things like, well, once we start this, it can never be taken away. And that affects a patient's decision. Well, that's in fact not true. So we know that physicians' values and knowledge have an impact. We know from the data that Dr. Chung presented that patients' concerns about trust in the system, trust in the data have an impact. And what do you do with I think those two pieces of information? The other thing that I find very challenging about the case of TPA, let's say you do in fact have a family member who's present. We're not talking about applied consent anymore. We're talking about surrogate decision-making. You have a family member who's there watching their loved one with some severe deficit and you're saying, hey, I got this drug, do you want it? Well, they're basically watching their loved one on a ledge falling off of this ledge. And in the back of, so you're trying to give them something which I think by virtue of the discussion they're gonna, for the most part, go for. Because there's really no alternative and they're seeing their loved one just have with a future in front of them that they wouldn't want. And it is very different from what they've had. And in the back of your mind as an emergency physician you're saying, okay, I'm recommending this, but I know there's a controversy somewhere lurking in the back. Should I tell the patient about this? Do I have an obligation to disclose this? Would that be good to do? Would that be harmful to do? Can people do anything with that information? Is that irresponsible? Like giving them a menu that's somehow sort of clouded with water, you know? And I think that that's something that people really struggle with and it's really bothersome to me, certainly. So my interpretation of this stuff at the end, I've already kind of said that I think the waters are quite muddy, but personally I think implied consent in the setting of TPA makes sense and I agree with the view that you gotta come up with a pretty compelling argument within three hours to not do it. Because time really is the issue and I think you're depriving patients of that and if we're gonna go through the effort to be in the room, make door to physician time like minus five, make door to CT scan time five minutes then what we're saying is that this thing really just needs to be done now. And so I personally agree with the implied consent and the one thing to note to me at least at this moment, although I found this study, the data that was presented here very interesting, those numbers actually don't sway me in any way. The 75%, I really just don't know what to do with that because we use clinical decision rules like 95%. 75% is like, oh my gosh, it's to me not so much better than a coin flip that it really helps me very much. It makes me feel like, and especially because there's a deeper sense in there that among the two populations, there was a difference in their values and a difference in their reasons. So the 75% are not actually equivalent. So I don't really do a whole lot with that. I do more with the sense that our best guess is that this is really good for people and time is really of the essence. So I think in summary, the two areas where I think a lot can come together is one better collaboration among everybody. When physicians just absolutely have to collaborate regardless of their specialty. And that's really the only way that you can give the reliable emergency care when you're really talking about seconds and minutes to patients. That's number one. And then two, sort of especially the public conversation. I think personally there should be a lot more conversation about stroke and TPA, just like there was for decades about cardiovascular care. And that, especially in the setting of collaboration, I think could make a pretty positive difference for patients. Thank you all. That was really excellent. In a moment, I'm gonna open up questions to the audience, but I'd like to take the opportunity to take the first. So let's presume for a moment that we, let's presume that we agree that presumed consent in this context makes sense, right? That there's enough professional consensus that the patient who comes in without a surrogate who needs TPA should just be given it with a very brief attempt to find a surrogate. I think once we accept that, then that does have implications for the patient who comes in with a surrogate already there. And it maybe it tells us something about what informed consent should consist of when there is a surrogate available. And specifically, I think maybe it makes it so that the amount of discussion that's required is very minimal, right? If you would come in without your surrogate, we would have just given you the drug. Now you do have a surrogate, should we spend five minutes having a conversation about whether or not to give the drug? So I'm gonna float a standard for what that consent discussion might be in the emergency room. And then I'd like to hear what you all think of this. So when I was trained, oh wow, 15 years ago, 15 years ago now, I haven't thought about it. We have fairly long discussions. It would take 10 minutes of discussion, I think, to figure out whether we should give TPA to this patient. Instead, if we accept that consent should be presumed, maybe the conversation should be, rather than a seeking a yes or no, the conversation should go something like, I'm going to give your loved one this drug that breaks up clots. The benefits of this drug are better than the burdens. Rarely patients have bleeding and occasionally they die from the drug, but more often than not, it provides a benefit. Are you okay with that? So it's very brief. It's giving a recommendation. It's not asking you to make the decision, it's telling you what the decision is and asking you to at least assent to it, if not carefully ratified. Do you think I would be violating any ethical standards in doing it that way, right? I'm stacking the deck in favor of a yes. Can we kind of make a friendly amendment to your statement, could the conversation be everything you said up to the last point and then can you think of any reason why this wouldn't make sense for your loved one? Something they might have said or a preference they've expressed that would make us think twice about doing our standard treatment. So that's an even stronger way to frame it in favor of treatment, I think. And so I accept that amendment. And I guess Lee and I probably are okay with how the way I just framed it. I wonder if the others have any concerns about it. I think that, so I think one thing that I found really interesting in Lee's framing has to do with the potential harms associated with informed consent. And I think, again, I think one thing that is really interesting to me about these neurological emergencies is that really the history of bioethics for certainly for the last half century now has been about really the preeminent role of respect for patient autonomy in bioethics. And with informed consent really being in some ways the central focus of early work in bioethics. And to the point that it still retains this I think kind of central role in the way that we conceive of ethical patient care. So as a countercurrent to paternalism. Yes, certainly. And then to really immoral treatment of certain classes of patients. Right, so yeah, and actually a couple different roles. So there is the autonomy preserving role. There's also the protection against certain abuses. And so I think that, we've talked a lot about, I think a lot of the important kind of pathophysiological data for why early treatments necessary. But I think it's important to take a step back and kind of remember how different this discussion is of most discussions we have around informed consent and autonomy, which we hammer the med students on. So I guess, this is a question where, again, I think if we think that there's a harm associated with a five minute discussion or a 10 minute discussion in terms of a reduced probability of a good outcome. I think another thing that is interesting in the data and I want to get maybe later on to some of Julian's concerns about the data I presented because I have some similar concerns. But some of the data suggests that consent itself is important to patients. So it wound up not playing the role we expected it to have when we presented two different versions. But it's a consistent finding, we found it as well, that there are lots of patients for whom they are gonna go with what the doctor says they should do. And still, like everybody wants to be asked. So there is something about being presented with options and being asked their opinion that people seem to value that doesn't really depend on choosing differently. Then, yeah, and I can't quantify that for you. I can't weigh that against, you know, Jeffrey Saver's data about a five minute delay being associated with a 1% reduction in good outcome. But there is a value there that is a little bit somewhat independent of actually choosing differently. Well, this is why I proposed having a conversation at all as opposed to what I could have done, which is say, hang the bag, start the drip, and then have a brief conversation to explain what you've already done. And then let them raise objections, right? So the reason I proposed it as a pre-treatment discussion is I think precisely because of that, because right from your data, people want to be asked. Right. I wouldn't ask about CPR, right? Patient comes in at cardiac arrest. I'm not gonna ask if it's okay, I'm going to start. And then if they raise objections, then I might stop. You might start by, you might before you put your hand on the chest say, is your father DNR, right? I mean, if he's 85 and he's wearing and he's wearing like a clothing from a nursing home, you might ask that question. So I would argue that you test that presumption first and you quickly assess it and make a decision. Right, very, very briefly. Could you introduce yourself? Hi, I'm Michael Young. I'm a fourth year medical student here. Dr. Uchung, I was particularly intrigued and surprised by the piece of data that you presented that patients' preferences were not influenced by medical values around longevity versus independence, which seemed to be the salient features in the risk-benefit analysis. And so given that, I was wondering what is driving patient decisions in these situations and patient preferences and if they're not responsive to what perhaps ought to be the key deciding factors, what are we to make of preferences in these cases? And this kind of transitions to a second question I had about how are we distinguishing between preference sensitive conditions and preference insensitive conditions? Yeah, so I think the, so I'll make the comment here that I was suggesting before having to do with, I think some of Julian's concerns about the data. So I, first of all, would be, and part of the reason the study is set up the way it is as a comparison is that I think that the specific number and percentage I think is, I would warn against, I think reading too much into the specific proportion, partly for the reason that it can be so sensitive to the way the data are presented. And there's actually, I think it's interesting to point out that there is a discrepancy between the number of people who say that they would want this presented in this very blank format. And then I think our clinical experience in practice, which is that patients rarely refuse proposed interventions, right? So there is this discrepancy between this bare presentation of would you want this treatment to, even the way that a typical informed consent discussion occurs, which is a little bit more leading, right? And it is often a proposal for treatment. And, you know, and we had data basically that, you know, in practice patients don't, you know, it's not like 25% of patients refuse DPA. Yeah, I mean, I think there's also another, there's another piece to this discussion that you alluded to briefly when you're talking about sort of the history of informed consent, which drove to a large extent out of research, right? The idea that you could conduct research on people without their consent was an accepted practice until, I guess, Nuremberg or shortly thereafter, right? That just was why not? Do it on prisoners, do it on children, like you did it on slaves, like you did it on kind of people who were either a lower class of citizenship or just what was, you did what you did because it would be better for society. So I think part of the challenge here is that the rigor and the application of informed, written informed consent in the context of clinical research is different than informed consent in the context of discussion about a proposed procedure. And I think when we talk about it, a lot of times people mix those two together. For a long time, people were convinced they needed written informed consent for TPA, like literally they printed it on a procedure consent form and they made the patient sign the form and put it in the chart as a way of acknowledging that somehow they would be protected if there was a bad outcome. When from a medical legal standpoint, all that's required is a note that describes the conversation. The signature is less valuable than a note that documents that there was a conversation that took place. I think to get back to your question about preference-sensitive conditions, those are typically conditions where the outcomes, where the treatment does not have a clearly favorable outcome compared to the alternative treatments. And you're choosing between one of several efficacious treatments whose outcomes differ slightly in terms of different preference-sensitive conditions like sexual dysfunction versus incontinence versus not longevity as an outcome measure per se. So, and typically the differences between the treatments are extremely difficult to measure. Unlike something like TPA in the first 60 minutes, the treatment is 18 times more likely to benefit you than harm you. So in circumstances like that, we generally don't think about shared decision-making as a hallmark of the process because a reasonable person when confronted with those sorts of choices would choose the efficacious therapy. We don't do informed consent for antibiotics. We might, if we were trying to choose between two antibiotics that were both required, six weeks of administration and one was more likely to make you nauseated and anorexic and the other might make you more likely to have a gallstone, but both were needed, both were reasonable alternatives for treating your co-existence. Hi, I'm Bob Trug. I'm the director of our Center for Biowethics here and I'm also an intensive care physician at Children's Hospital. I want to first of all say this one of the most interesting conversations I've been a part of in a very long time. You guys did a terrific job and it's a very rich discussion. And I've got a long list of things here but I'm just going to focus on one now which really builds on TOS's question. If we accept that this is just so obviously the right thing to do, then I just think you're not taking the notion of presumed consent seriously enough, okay? Now, I might go back and question whether it is obviously the right thing to do but let's take that as a given that it is obviously the right thing to do. Then certainly that justifies initiating treatment in somebody who can't consent. But why, Lee, would we care about a uniform statement to talk to a surrogate about if it is so clear to us that this is the right thing to do? Why are we asking surrogates? The data shows that surrogate decision making is little better than chance compared to what the patient would decide for themself. So if the surrogate says no, is that really a reason to stop? So I don't think the surrogate should be making the decision. I think the surrogate should be providing us with the information that will allow us to make the correct decision in the patient's perspective. And I think that's really smart but you're sort of lumping that into somehow informed consent, that's not. That's getting information as to whether there's contraindications to the procedure. Exactly, which otherwise the absence of that presumed consent. Do you have a DNR order? Have you had recent surgery on your brain? Do you... They're demented. Right, okay. So there'd be various rule out criteria but that's not getting informed consent. That's figuring out if this is the appropriate treatment. I totally agree with where you're going but I think we don't have an agreement in the medical community around that point. Yeah, well this is where I'm kind of going. And Winston, to your point that patients value choice for the sake of choice itself, I think that that's true but I'm not sure that that necessarily extends to surrogate choice. I mean if you could ask this person directly do you want the TPA and if somebody says no I don't want it that's a pretty high bar to me. I mean that's battery. We use surrogate consent as a substitute but it's a poor substitute and in this case I'm not sure it's a substitute that should necessarily change the decision to treat. All that being said, I think if this is really how you view it and I'm pretty convinced that it is then this really needs a major carve out because if this is just sort of one way of applying our principles of informed consent and surrogate decision making it doesn't fit. And I would worry about extension to other things where like I was trying to think of examples I couldn't think of good ones but I'm thinking of like a stat cesarean section for somebody. We would say well this is obviously the right thing for this woman to do so we don't really need anybody's consent. We should just go ahead and do it and that's wrong. So I'd worry about creep with this kind of reasoning but it does sound like in this particular case there might be a very good justification for making a carve out and such a firm carve out should go a long way towards handling I think the very legitimate concerns of emergency medical physicians about feeling like they're doing the right thing. That they've really got the institution and there's various societies firmly in their court when they go ahead and do this. So I just I'm very intrigued by your comments I wanna make two observations. The first one is about wrong surrogate decision making and the second one is which I'll get back to in a minute my observation that in academic hospitals where there are neurologists in the emergency room 24 seven because there's training programs this issue is a non-issue. I really think the issue is I don't think most emergency physicians I'm speaking on behalf of them I know a lot of them would intercept the neurologist and prevent them from giving the TPA. It's more that they don't feel that the data is sufficient to justify their role in executing the presumed consent because they don't have the confidence that the patient in front of them actually is a person for whom benefit is substantiated beyond any reasonable doubt. So I think if you asked me to decide who goes to the OR for appendicitis and I have to decide in three minutes and I don't get to have a conversation with the patient I would be really uncomfortable with that because that's not my field of expertise. The challenge with the surrogate decision maker we had the circumstance just a few months ago when I was on service patient with mental illness came in had a severe stroke was unable to speak his surrogate decision maker also had severe mental illness and her response when we mentioned about thrombectomy was I don't think he would want something like that I'll just take him home, right? So that is clearly not informed refusal, right? That's mental illness not appreciating the severity of his disease and his likely outcome. Circumstances like that are worse than no, right? Having a surrogate whose judgment and context you have to interpret is almost worse than no surrogate at all. Yeah. So do you think my comments were too inflammatory? No, I don't find very much inflammatory but I think that there's enough ambivalence among emergency physicians that I think that a good number would in fact intercept and say and especially in the emergency department where it's the emergency physician that puts in the order or gives the drugs in most cases an neurologist can't push this in some places the emergency physician and the nurses will only take orders from the emergency physician that is the case in at least a number of institutions that I'm familiar with. So I think in fact they probably would and I think with these new guidelines these new ASAP guidelines that number of people who would in fact intercept is gonna go up. And one of the responses to those 2015 guidelines from folks who are in the emergency medicine community but feel very strongly that the data about TPA are perfectly fine and indicate that you should basically give it based on implied consent is that the problem with there's no guidelines is they're gonna make emergency physicians block this, delay this, deny this and that's gonna be a huge problem. I think that's the reality. To go back to as well related to this toss this question of if you present to them hey there's this drug, you know this is the benefit there's a small risk of harm we're gonna give it is there anything that you know about your loved one that makes it that they wouldn't want it? At least in my experience I think that some physicians or many emergency physicians would feel a little like they're hiding something and I feel like a good number of patients especially very vulnerable patients would wonder why at the end are you asking me if I have any objection what objection are you talking about? Nothing comes to mind but you must know something. I had a mental image of somebody saying wait what go back to the thing you said quickly about you know the risk of harm. I mean I think that there's a couple different conversations I mean it's one reason this is a rich topic and so I wanna be thoughtful about where the conversation is going right because I think there is one conversation to have about why do we do surrogates consent at all right so we all know that the data on the accuracy of surrogates is quite poor right so and I think that that gets to some other questions about you know what is actual function of informed consent and in what ways does surrogate consent play that function? Who's benefiting us or the patients? Well I think also you know I mean I think that there is something about we talked before about paternalism and wanting another voice and there is something about in addition to the autonomy preserving function and the protective function of informed consent and surrogate consent does play that role imperfectly and then there's a separate conversation we could be having about which we have been dancing around a little bit like we said you know when we had the call that we were gonna try to bracket as much as possible the controversies over the interpretation of the data that's becoming harder and harder to do. You know I mean I think there is something about the clinical circumstances of acute stroke that make clinical decision making difficult and that also I think have contributed to this you know kind of peculiar history that we've been referencing right. So in an emergency setting acute stroke can be difficult to diagnose obviously you know we know that on clinical grounds it's very difficult to distinguish ischemic from hemorrhagic stroke you know without a scan and in some cases you know people don't feel comfortable reading scans you know and then I think just the peculiar history of acute neurological interventions which you know I think for a long time neurology has had the reputation as being a therapeutically nihilistic sort of field and so I think that it's been kind of a real change in identity for a lot of people around you know the idea that all of a sudden the neurologists are the ones you know originally the last people you would ever run into an emergency department but you know are the ones you know in many cases running into ERs and barking orders of people. So I don't need to be sensitive to like I think there are particular different directions as conversation could go but I think some of them are actually you know in some ways a little bit more kind of historical and almost sociological. I think we've got time for one last question. Okay I'll try to make this quick. I just want to say oh my name is Ben Tolchan I'm a epilepsy fellow and so recently was a neurology resident in an academic center pushing TPA in the emergency department quite frequently and in that setting I'm very dismayed to hear about this change in the policy of the emergency medicine physicians. I want to just throw out a perception and I think it's a pretty extreme perception but it's one that I find myself having and I'm curious to hear the response of the panel. You know it seems to me that you know the bioethics really came out of a very specific time and place when the major problem was paternalism and that the medical establishment was very much ignoring patient and family input in decision making and that situation has changed a lot since the birth of bioethics we're in a very different position now it's a point where I think those wrongs and those harms still can be found but in my perception overwhelmingly the much more common harm is where autonomy is allowed to trump other concerns other values like patient welfare and beneficence and in my perception this is one of those cases where the welfare of many patients who very much would want to be treated and are able to express that given time are not treated or have delays to their treatment because of concerns about consent and I think there are other cases in medicine on treatment of some psychiatric patients who are sort of on the border lands of capacity and the border lands of consent where patient welfare is sacrificed to autonomy and so I'm wondering if we might as a bioethical community need to think more about the balance between autonomy and patient welfare. Very well said I think I think you've really crystallized the issue here is that autonomy is being valued at a level that is so extreme that it is interfering with other bioethical principles in this very time sensitive critical condition. Where I thought you might be going is also that we've almost abandoned our response of the paternalism pendulum has swung so far to the other side that we now treat the patient surrogate or the patient like the consultants were like well there's a 30% chance of this a 40% chance of that this is unknown that's unknown what would you like to do? Like we're giving them a menu and they're gonna order dinner as opposed to saying here's how I put together the situation here's what I recommend but I'm interested in your thoughts about this and your concerns. I mean I think there is a literature in bioethics about precisely this problem right and so in some ways shared decision making has been promoted in a certain way as sort of a response to this worry about kind of abandoning the patient with the data leaving the patient with the data and that's kind of the intended role for the physician in the context of shared decision making is basically to help the patient interpret their values and goals in the setting of making this decision. I think what is interesting about this particular case is one where the tools of shared decision making may not be well suited to the clinical situation. So I mean I think there is a concern among people in bioethics about what's sometimes called a consumerist model of autonomy where you're picking out what you want on a menu and I think we recognize that that's not what most patients want and I think the particular question I think is kind of interesting here is that this is a particular case where the tools that have been promoted to get maybe a better handle on medical decision making because of the time sensitivity might not be appropriate to the situation. So I want to thank our panelists one last time. This has really been a great conversation.