Latest / The JEMS Report / Cardiac Arrest Recognition: What 145 Real World Videos Reveal About CPR
Transcript
- Nat Kapustin: All right, hey everybody, welcome back to the Gems Podcast. Excited to be joined by a couple of new friends of mine to to discuss a recent paper. this is very relevant this month being National Sudden Cardiac Arrest Awareness Month. I'm excited to to talk about what came out of this paper and what it can really mean when we talk about cardiac arrest, resuscitation and education and where that can really change some things. a lot of cardiac arrest education begins after the diagnosis, if you will, is already made. The mannequins unresponsive and instructor announces, hey, there's no normal breathing and they start CPR. And it's been my experience and it has been the findings of this paper that actual cardiac arrest is a little less order orderly than that. So I'm gonna pass it off to my two guests to introduce themselves and then we'll talk about the Talk about the paper. Hello everyone. Thank you for the opportunity to join your beautiful, fantastic podcast. And I'm Tommaso Squizzato. I'm from Milano, Italy, and I'm a cardiac reserve surgery and I work in a cardiac intensive care unit. and I am Brandon Odo. I am a critical care PA nowadays in Connecticut, formerly an EMT on the ambulance, and I'm a member of the program advisory committee for the Citizen CPR. Heart Safe Community Programs where we work on the cardiac arrest chain of survival. Awesome. Well again, thank you to both of you. I really appreciate you spending spending time with me today to talk about this. So Tomaso kinda wanted to talk with you first, if you don't mind, because it it's really the the work that you and your colleagues did that that prompted a lot of this. So you you took a look at what publicly available video showed cardiac arrest looking like. And I'm curious what prompted you to to look at that. Sure. So there are two main reasons reasons why we undertook this work. So the first thing is that this work is part of a bigger project named Time Care Project that is a European Union and Italian Ministry of Health funded project that has the main aims of trying to develop new technologies to improve recognition and response to cardiac arrest. And years ago I worked with engineers to develop a prototype of an algorithm that was able to detect when someone suffer a cardiac arrest in front of a surveillance camera. So in order to detect when someone suffer an unwitnessed cardiac arrest and use surveillance camera to detect these patients. collapsing in an automatic way. And at the same time, we we thought that analyzing existing video could be helpful, yes, for developing the algorithm, but also to provide a rare window of the the first second of a cardiac rest that the it is typically not available because as clinicians we see patients in the pre-hospital setting when we arrive with an ambulance or even after in the emergency department or in the ICU. So we completely miss the immediate the the first minutes, the first seconds of of a cardiac arrest. Yeah, and it's those those first seconds that I think are are important and impactful, especially as we look at educating the public on their response to cardiac arrest. So your study looked at I think it was a hundred and twenty seven publicly available videos, with a hundred and forty-five actual out of hospital cardiac arrests. curious what what do you feel was really revealed from that that traditional cardiac arrest registries failed to capture? As you said, cardiac arrest registries miss the initial part. So we don't have the first second. Frequently, also the first minute is not available, it's not described. We know we collect when the caldecrest happened, what bystander does, but typically we know in a binary way if they perform or not CPR, if they use an AED or not. But we don't have any more specific data on that. And our study had a different perspective. So we had a window prior to cardiac rest, so we are able to see how the collapse happened. So what were the the characteristics of the collapse and what were the action of the bystander seeing the patient collapsing. And and theoretically, but it was not very easy, and probably we will discuss later, understanding. What are the barriers for bystanders to provide life saving interventions? Yeah, I I think you know there can be some fairly significant barriers there. you know, was it was it difficult during the research to truly parse out the cardiac arrest from like other like syncopies or see or the start of seizures or another cause of that collapse? Definitely was not very easy to perform such a study. It is clearly we immediately understood that performing a systematic review, for example, of a published article is clearly much more simple because we have a database that is PubMed and we are able to search with keywords. On the opposite, video cardiac arrest or people suffering a cardiac arrest in the real world. Are not easy to detect. They are published frequently on YouTube or other social media, and it took a year to identify more than a hundred of of videos. And and this was again not easy. And the other problem was exactly, as you mentioned, understanding if we were in front of a true cardiac arrest or not. clearly we verify where possible the case. So cross searching other sources, describing or commenting on a cardiac arrest happening on that day we on that location. we use the the description of the video and clearly if we saw in the video performance CPR using an AD clearly we we were mostly sure that this was a correct case of a cardiac arrest. Awesome. Yeah, I can only imagine how difficult it would be just looking at the looking at the video and assuming did the did you also look at did it include audio or since it was just surveillance camera footage, was it just video? They were mostly without audio. Yeah. So what types of things did you notice in those first those first moments after after collapse that that are notable as we kind of Through what this what this truly looks like and where this may impact how we train. So, first of all, it is important to remember to all the listeners that we performed a qualitative analysis of this video. So we were not able to quantify except for timings to quantify exactly what happened. And this is important because it was mostly a subjective adjudication. of our team to describe what what happened. And our key our key aspect, key finding was that there are a short interval of time, I mean seconds, where appears to be people appears to feel that something is happening. You see that they immediately stop what they are doing. and they try to grab f something. and this clearly lasts three seconds. So there is it is we we typically teach that's it is sudden and people are not able to understand what what is happening. But probably there is a three seconds interval where people understand that something is happening but they don't have the time to act. Clearly this is not important for training because I don't see think that this could become a warning sign, but this is important for research, for cardiac rest research, in order to identify people that are just think automatic recognition that is becoming very, very common in today's world thanks to technology. Just think the the smartwatch that is able the Google smartwatch that is able to identify someone suffering a cardiac arrest. And the second part is the first second after collapse. There is frequently what now we call barriers that we know are seizure, for example, or movement. So we typically teach people during basic life support courses, especially to lay people that cardiac arrest is something sudden that happens suddenly and there is no movement. But rarely I participate to a lot of basically I support CPR courses but very rarely I heard instructor tell to participant that they may witness seizure-like movement or other type of movement probably we only teach to know that gasping agonal briefing is possible. But we probably underestimate the the possibility for bystander to see seizure or other movement. And this is a very important aspect for in my point of view. Yeah, I agree completely and I think that's where one of those barriers comes in is because it can be especially for a lay person who doesn't who doesn't, you experience this with any type of regularity that could make them question the actions they're about to take and you know the the initiation of that resuscitation. so Tomasa will give you give you a break for a little bit. and I'm gonna kinda pop over to Brandon for for just a few questions and then we'll continue to to unpack this. But Brandon, just wondering if you could discuss from a clinical perspective what is happening physiologically during those f those few seconds that that Tommaso just just described in the in that immediate cardiac arrest. Sure. You know, I think this is a an interesting question to ask. We're really describing the loss of blood flow to the brain happening very abruptly, and of course the brain is is everything that you are. So what that will look like is gonna depend on the person. Of course we know what the eventual outcome will be, but what is the what does the transition look like? And you know, you may see abnormal movements, you may see abnormal breathing, a loss of consciousness, which eventually will be complete, but there may be some kind of an on-ramp to that. And what I find interesting is that this this brief window of time is one that is almost invisible. To most people most of the time, because it's so brief and so abrupt. How do you study it? Which obviously is the the challenge that Tommaso faced. But really, you know, when we teach CPR classes, when we discuss this clinically, it may be a phenomenon that most people have never seen. You know, we might take care of these patients later. But it may very well be that if you take the public as a bucket of people, they see more. cardiac arrest occur than we do, unless we work in a very acute setting like the ICU or the emergency department. But of course, some arrests are not witnessed by anyone, but when they are witnessed, it's often by bystanders. So when we try to teach people how to recognize these things and what to expect, which of course is what we're looking at today, we're not very well equipped to do that. I could tell them all about what this patient will look like the next day in the ICU. But this this is a a phenomenon of the field in many ways. It's a phenomenon of the public, of the the public environment. So finding evidence to you know, know what it looks like so we can then share it with people when we educate is very tough. And I think, you know, thirty years ago we we couldn't have. Now there's a a lot of cameras in the world and other ways of of monitoring that are always out there which has made this kind of conversation possible. Yeah, absolutely. And I technology has come a long ways, right? And you know, for for us to start to see some of these signs on video and understand what's what's happening. something else that I kind of wanted to to discuss at the same time. What what type of competing diagnoses do we need to be considering so if we're an EMS clinician that's that's witnessing this or or perhaps a layperson who's witnessing it, what else what what are we potentially looking at other than cardiac arrest? You know, I I if you are let's say a paramedic who's with a patient and they suddenly lose a pulse. And of course that can happen. Maybe you're called for chest pain and then they arrest in your presence, I there's not a huge differential for this specific phenomenon. And By specific, what I mean more than anything is the tempo, how quickly it occurs. A lot of people now have seen videos of some athletes who have sudden cardiac arrest. That's become better known. And I think the one thing that everyone would agree is that this is something that happens instantly. You can be walking along and then you completely collapse bonelessly. Of course, there's always a possibility for something a little more gradual, but that's the classic phenomenon where you suddenly go into V fib or something like that. Not that many things cause that. So if you're with a patient and that is literally what you see, there might not be a huge differential other than cardiovascular collapse, let's say. And that doesn't have to mean a pulseless rhythm. It could mean something more temporary or benign, like another cause of syncope or Or some severe insult which is not quite pulseless, you know, they threw a massive PE, something like that. They they lost consciousness, but they may still have a pulse. But if you kind of lump those things together, people will often talk about seizure. Sure. Seizure is not usually that abrupt. You know, there is a little bit of a prodrome. Sometimes the patients know them well. There may be some change in consciousness. of course there can be psychogenic effects, people feeling faint for whatever reason. but really I think you have to think cardiac when there's something that abrupt. Yeah, I agree. And i it's it's interesting that that we're having this conversation now, because just just a few months ago I had a case with a with an engine company crew that that I was supervising and we made essentially a What ended up being a cardiac arrest that presented very similar to what what you're describing. A patient with crushing substernal chest pain was awake and alert, but but diaphoretic on arrival, we start start to gather the s gather some vital signs. He's you know, short of breath and I he he has that that short few seconds of shaking that I've seen before in a witnessed cardiac arrest. and then immediately unconscious. so I start to start to lay him lay him down onto the floor and one of the other one of the other responders is is is asking, you know, should I it looks like he's not breathing as well. Should I put a non or breather mask on him? I'm like, no, he's in cardiac arrest. We need to let's put the pads on him and and we did that and confirmed that there was a cardiac arrest. He was in a V fib cardiovascular arrest. So yeah, I think there's a real case to be made for working with probabilities here and about risk, right? You can talk about a differential in theory, but in practice the greatest harm will be not intervening on pulselessness rapidly. And then in a case like that, what is the probability that this patient who looks cardiac just happened to have a neurologic event or some other th why are we trying to turn this into, a a world of abstracts. Work with what's obvious. And that's why, we'll talk about layperson education here, but we've really moved towards don't sweat so much proving this person has no pulse. sweat proving that they the opposite. Treat what you can. What's the worst case? Maybe they had a pulse. But that's generally not so harmful, whereas not intervening is clearly harmful. Yeah, absolutely. Well I I want to unpack how how truly important this is for us when it when it comes to training and what we really need to we need to be doing to to help in cardiac arrest resuscitation. So you know I think a lot of this starts with the layperson potentially and the extension of community CPR programs is you know dispatcher guided CPR programs which are which can be very beneficial. but Tomaso I wanted to kinda pop it back over to you because there it was frequent in the study that there would be agonal breathing that's occasionally mistaken for effective breathing. How do you feel that dispatchers should be helping the layperson understand the difference and know when to react. Yeah, sure. So this is a very important topic because I think everyone in their career had cases of cardiac arrest misrecognized so where bystanders didn't recognize the cardiac arrest and for example they put the patient on recovery position and when you arrive with the ambulance you find the patient in recovery position and I was actually in cardiac arrest. to me it happened a lot of time, unfortunately, and clearly the outcomes are very, very poor. And I don't think that so your question was how how the dispatcher can help the the the the bystand. I think it is not I mean it is the dispatcher that should be able to understand based on the on the questions provided to the to the to the caller if there is true case of cardiac arrest or not. And probably these we should really reduce the threshold for cardiac arrest recognition even even because there is very low risk of performing harm to patient if we start CPR in a patient who is not in cardiac arrest. So very low risk of CPR in patient without cardiac arrest. And probably this would be very important and would change a lot the outcomes because very frequently I see hesitation even in the dispatcher. I mean there are very different protocols around the world on dispatcher recognition or cardiac arrest. Probably there are protocols that works better in cardiac arrest recognition and other have lower performance. Clearly the difference is that probably some protocols have higher false positive and other protocols lower false positive. But in the setting of a cardiac arrest, I would totally prefer having every week some CPR performed over people without a cardiac arrest compared to delaying or even not performing in CPR. So I think that to summarize this this concept, I think that it We we can't put the decision to to bystanders to the callers, but it is a matter of dispatcher the decision whether we are in front or not of a cardiac arrest. And finally it's their decision based on the information clearly that they receive. But I would suggest to reduce the threshold to To start giving bystander inst instruction to bystanders. Yeah, to me, you know, telephone guided CPR has been one of the most powerful tools for increasing the rates of bystander CPR because training everyone in the world to do the right thing quickly is tough. There's a lot of people. But as long as they call and they tap into that system, then now you've centralized it. All you have to do is train your dispatchers. And Tell them to tell the people what to do. So systems that we've seen do this well, they just boil it down to no no-go. Is the patient conscious? No. Are they breathing normally? Not agonal breathing? No. Do CPR. Yes, but no, no, just do CPR. They can teach them, they can reassure them, they could do whatever is necessary in real time. But then you're making sure that happens right away. And I think that distills it down to making it very simple. Tomasa, did you have something to add? Personally, I like a lot the non-go. I think this is the most simple algorithm I ever read about, and I would probably love to see implemented more widely across different dispatch centers. I think it is very used in the US. my understanding is that. And I would probably I love to see that every call for a sudden loss of conscious a sudden collapse, that the first thing that you come to your mind should be cardiac arrest in a dispatcher, in my opinion, until proven otherwise. I think that this mindset would would reduce a lot of missed cardiac arrest, and the other issue I would really prefer less use of lateral of recovery position. At least in Italy it is still suggests recovery position to patients who are unconscious and judge breathing by bystanders. And this is a very huge problem because I think that bystanders overestimate when someone is breathing. And so this leads the bice the dispatcher to put patient on to suggest to put patient on recovery position and this is deadly. Yeah, absolutely. something else that I'd like to kind of address next is how we're teaching CPR. you know most CPR courses teach recognition. they don't really teach all that well the mechanics and and what follows. You know, like I said at the beginning, walk in, you have a mannequin that's that's not doing anything and there's really not a lot of feedback there. So yeah, I think there's there's a significant disconnect in how we're teaching CPR and this should this this really needs to change. We need to better depict for folks what they're going to see and experience when someone has a cardiac arrest. I'm wondering what you guys think of that. I think it's been a missing element of how we We teach. As you said, usually they're already clearly this mannequin is in cardiac arrest. It's tough to teach. I mean, how do you how do you play that out? I think it would make sense to to show some of the things we've been talking about. Can you show a video or or some kind of footage? Maybe reenacted, maybe real, but something that makes it a little more real. Can you act it out? I mean, maybe, but now you're really asking a lot of of individuals. certainly we can emphasize the things that it's not. So you can teach. agonal breathing, that's that could still be cardiac arrest, abnormal movements, that could still be cardiac arrest and and so on. but I would like to see, a higher fidelity way to more reliably teach that, whether that's some sort of multimedia or something else. Yeah, I have an unpopular opinion with CPR teaching and frequently I discuss with a CPR instructor and they disagree with me. My opinion is that frequently for lay people we When we teach CPR we create some level of overcomplexity in the things and I would love to see more simple CPR courses. I mean many, many times we get in a lot of technical aspects that I think they are not important for for lay people and and frequently they create they create we create some barriers to performing CPR. I would like really a very much simple way of providing providing teaching teaching cpr and the first thing probably is around recognition you have someone with who sudden fall down in front of you sudden collapse in front of you and and I mean never give trying to teach them that there is no risk to perform CPR. So they they their first choice should be doing CPR to to to bystanders and not to think if it is correct or not to do CPR. They should be trained to to do always CPR and in some cases probably is not correct because I mean we from this video we clearly see that these is these are not syncope you immediately realize that people who who have a syncopy it it is a completely different way of of of falling down it is difficult to objectivate and to put in word but everyone we we we work with cardiac spatial and sometimes we see cardiac space in front of us witnessed by clinicians and we rarely have doubt of what what what is happening i mean if someone Have a sudden collapse like a cardiac rest. And so I would like to have the the same way of teaching to bystanders that the right thing to do is most of the time is CPR. But it's not easy. I mean, and currently how we currently teach is probably sometime we put a lot of questions that remains in the in their way of thinking cardiac rest to in in bystanders. I I don't have a clear suggestion, but everyone bystanders and dispatcher should reduce their threshold to start CPR. This is my my suggestion. Yeah, I and I think those those last two comments from both of you really really are saying some very similar things and are are speaking to exactly what you know could start to emerge as a potential best practice is Tomaso, you said it's difficult to put into words what what they're seeing, so that that reinforces as Brandon had alluded to, you know, perhaps the use of video clips and things like that in the training so that people can see what a a patient freshly entering cardiac arrest looks like. so that they can see some of those those visual cues and feel more empowered to act. I think that's gonna now lower the barrier to them feeling like, hey, yes, doing these chest compressions is the right thing for me to do. I've seen this before. I can connect it, I know I'm doing I know I'm doing the right thing. I I always show a clip of of a cardiac patient at the beginning of every CPR course I I do. And I think it is very important for two reasons. One for the the reason you just summarized But also to catch their attention. they really understand immediately what we are talking about. That would be an attention getter. Yeah. So we're c we're kinda getting towards around thirty minutes or so. So you know, kinda getting ready to start wrapping this up. But Tomaso, one thing that I wanted to ask you about, because I think that when it comes to evidence and looking at studies, oftentimes when you study something it it gives you more questions, right? What what do you feel is next in this research? As coming out of it as a researcher, what do you w what's next? What are you what are you excited about? Yeah, so as discussed at the beginning, we collected these one hundred and something videos just initially just to develop an algorithm that is able to automatically recognize the patterns, for example, of someone falling down due to a cardiac arrest in order to probably at some point creating a surveillance camera that is able to trigger an alert. And while doing this, we we thought that could be useful to to also describe in a in a descriptive paper how cardiac rest looks like and I think we we had it was really important to do that also you demonstrate considering your attention to the topic. And what happens next? I think that our dream, I I say dream because it is not very easy. We are we have a lot of technical aspects that are very complex for developing these algorithms so it is not easy to develop an algorithm that is able to recognize the movement and separate someone who is falling down due to cardiac arrest compared to someone who is falling down for other reasons. Because when we started initially developing the project we thought that yes, someone who falls down due to a cardiac arrest falls down and has no movement. But after seeing the video we we couldn't use these this pattern to recognize cardiac rest and so it is not easy. And our data scientists are working hard, but this would be our main aim after this publication. other other research perspective should be to analyze better what other barriers and why in some cases people start CPR or not, why the delay or not, the calling, and try to understand what which are the factors. Because if we are able to identify the factors, then we could also tailor training and change how we deliver CPR courses to lay people. But it is not easy because the quality of the videos are very heterogeneous. There are high quality one, low quality, other are of low quality. And we don't have the audio, that is an important comment That you mentioned previously. And yeah, this could be an opportunity if anyone has access, because very frequently there are I think 40 percent of 30-40 percent of cardiac rest happen in public places actually. And so for example, in Milano we have eight cases of out of hospital cardiac arrest per day, and I think three or four are happening Public places and so theoretically in front of surveillance camera, but we don't have access. we tried, but for privacy reasons we are not able to get access to those videos. and this would be very important, but it is not easy because first of all, surveillance camera frequently overwrites videos every 24 hours or so, and so we don't have the time to get in touch and acquire the video. And when we get in touch, we receive Mm we o always receive a not as an answer. I I'm sure. I think we would experience I think we would experience similar barriers over here on our s on our side of the pond, if you will. Brandon, I'll kind of pop it over to you. if I'm if I'm listening to this, consuming it as a as an EMS agency, as a trainer, educator, or a community training program, or maybe I'm The American Heart Association or American Red Cross listening to this. what's one change based on this paper that you feel like could be advocated for? Well, I think this is a great illustration of the the age old challenges of sudden cardiac arrest. The reason that this is a tough problem is because it is so multifaceted. This whole chain of events really is a chain that all has to be strong. Any one point where it's weak and it fails. And we've identified One weak point here, recognition. And that's one that we haven't given a lot of attention to. It doesn't matter if you teach people to do great CPR and anything downstream of there if they don't know when to do it. So if you are an educator and you're walking into a class tomorrow of lay people and you want to teach them to respond correctly, ask the question: how can I better prepare them to recognize this event quickly? without doubting or questioning whether this is the thing we talked about in class, whether this is cardiac arrest. and is my curriculum, the way we're teaching, preparing them for that, or is there is there a gap? Because that's always what we don't want. We don't want there to be a difference between how we train in reality. And I think there is. There's not enough of whatever that would look like for people to immediately see an arrest and say, This is what we learned about This is that thing. Right now they're still gonna say, that's weird. What's going on? Is that guy okay? Is he having a seizure? Is he sleepy? is that is this cardiac arrest? So, ask that question, how can you bridge that? And whether it's just talking about it, emphasizing that it's better to, intervene than not, whether it's calling 911 early and having them help out with that, whether it's showing some kind of videos or other acting it out, something like that. Maybe the answer is all those, but there's probably more yield to drilling down on this question than, drilling down on a lot of the other elements here. Because someone who doesn't do CPR for five minutes and then does great C PR is probably gonna be a worse outcome than if you did it quickly and did it so so. Yeah, I agree. I think closing that recognition gap and that reactionary gap is what is is where we can really where we can really make a difference. So if that recognition is coming much faster, I'm I'm trained, I'm empowered, I know that I need to react and that I need to react quickly, I think that's what's what can what can truly can truly move the needle. so again, you know, we've gotta we've gotta remember that cardiac arrest doesn't always look the way we've been trained, right? It's not the motionless mannequin that is is very obvious unrest unresponsive and in cardiac arrest and that you know we have to we have to help shift training programs so that we can truly teach people how to recognize what cardiac arrest actually looks like. And you know, I think this this work that you've done, Tommaso, you know, hats off to I really appreciate it. I can only imagine e even for the the numbers that you have in this study, the amount of work that that went into it. So I thank you for what you've done and you know we appreciate you taking the time to to share that work with us today, share some of your additional thoughts. i if either of you have anything else that you want to share in closing, more than happy to, but and as we kind of wrap this up. Yeah, sure. I think that we already know theoretically the chain of survival recognition first. CPR and then the fibrillation and theoretically we should remember also to teach in that order and with that order of importance. The other thing I would I would add is that if you are a healthcare provider, you're in EMS, you're in the hospital, I think you should ask some of these same questions about yourself and in that environment, 'cause that's not really what we we've been talking about, but a lot of the same questions apply. And ask, if s cardiac arrest occurs in front of me, the kind of patients that we see, what would that look like? And is there a potential gap to our intervening there as well? And if so, how can we close that? Because it's gonna differ. You probably have more resources. Of course, you probably have more training, but there are still gaps too. We see patients in the ER, on the floors, in the field, where, you know, instead of CPRs being started immediately, there's a delay. So, you know, overlapping questions, but I think still a really important one. Yeah, and I agree because it it is not all that often, even for the EMS provider. that we literally see someone go into cardiac arrest in front of us. all too often and the vast majority of the time, and I'll say from my career experience, the lion's share of the cardiac arrest I've responded to, I found that person in cardiac arrest. It's been very few that that actually arrested in front of me. so, you know, I I agree completely. There's a there's a reactionary gap, be it lay Lay provider or train provider, either way. there's there's an ability to close. When I was training someone I I had a an ICU director tell me, listen, if you see a an arterial line in one of these patients and it's flat, please don't make your first reaction to go and play with the monitor. Make sure the patient has a pulse. Right. That should always be step number one. you can look at the transducer later, figure out what's going on. Yeah. Awesome. Well, hey, thank you both for your time. I greatly appreciate it. Thank you for everything you're doing for the for the profession. and you know, we we really appreciate it. So thanks a lot. Thank you. Thank you. Have a great day.