Hello, my name is Ryan Morgan. I'm a pediatric intensivist from the Children's Hospital of Philadelphia. Today I'll be speaking about pediatric cardiac arrest and cardiopulmonary resuscitation. I don't have any relevant financial conflicts. I do have grant support from the NIH, all of which is related to pediatric cardiac arrest and CPR, and I do serve on the American Heart Association's Emergency Cardiovascular Care Committee, which does make policy and guidelines related to this work. Our objectives today are to understand pediatric cardiac arrest epidemiology, to know the key components of high-quality CPR and the supporting evidence for them, to understand the essentials of resuscitation physiology and the rationale for monitoring and guiding according to physiology during pediatric CPR, and we'll finish off by discussing elements of post-cardiac arrest care in children. Let's dive in on pediatric cardiac arrest epidemiology. Starting with pediatric out-of-hospital cardiac arrest, this affects greater than 5,000 children annually in the United States. Rates of survival to discharge and survival-favorable neurologic outcome, as you can see, remain quite low. This disproportionately affects infants, which ranges from 30% to 60%, depending on the registry that we look at, really depending on whether or not patients with SIDS are included. These frequently occur in the home, are mostly unwitnessed, and have predominance of non-shockable rhythms, and all three of these characteristics, unfortunately, are associated with worse outcomes. Greater than 15,000 children receive CPR in the hospital each year in the United States. This includes 7,100 children who have pulseless cardiac arrests, and another 8,100 who receive CPR for bradycardia with poor perfusion, some of which, as we'll discuss later on, go on to have pulselessness during the course of their CPR. Rates of survival are substantially better than out-of-hospital cardiac arrest, with rates between 40% and 50% over the last several years. The mean age is around five years, and the median at about two, with slightly more males being affected than females. Event survival, being either return of spontaneous circulation or return of circulation through eCPR, occurs in about 80% or even 90% in some series of patients with in-hospital cardiac arrest. Rates of survival to discharge are about 45% to 50% in all cardiac arrests, though about 40% among those with an initial pulseless rhythm. These rates of survival improved dramatically from 2000 to about 2010, from about 20% to 40%, and have subsequently remained stable here. In the literature, 80% to 90% of survivors are reported to have favorable neurologic outcomes. It's important to note that most of these studies use gross indicators of neurologic outcomes, such as the Pediatric Cerebral Performance Category score, which does lack granularity in picking up on more subtle neurologic dysfunction. And outcomes from cardiac arrest are highly dependent on a number of factors, all of which we'll talk about in the moments ahead, including the underlying conditions of the patient, the etiology of the arrest, the initial CPR rhythm, duration of CPR, CPR quality, and multiple post-arrest factors. As opposed to adult cardiac arrests, which are split between unmonitored and monitored settings, roughly 90% of pediatric cardiac arrests take place in ICUs or operating rooms. This increased over time, as shown in this work from back in 2015, and is largely attributed to the widespread implementation of medical emergency teams and rapid response systems, as well potentially as early warning scores. It's also important to note that pediatric cardiac ICU patients are more likely to have cardiac arrests during the course of their hospitalization than general pediatric ICU patients, with rates that are about 3% to 6%, as opposed to 1% to 2%. Most pediatric cardiac arrests are attributable to respiratory failure or progressive shock, and sometimes a combination of the two. The interventions in place at the time of arrest reflect these, with the majority of patients being invasively mechanically ventilated at the time of arrest, many of them receiving vasoactive infusions, and nearly half having arterial catheters in place. It's important to consider these as the underlying etiology of the arrest is related to outcome. As you can see here, patients with sepsis, or patients who are on inotropes at the time of cardiac arrest, are more likely to die prior to hospital discharge, whereas patients with respiratory failure, as the etiology of the arrest, are relatively protected. Let's change gears with a clinical case. So a six-month-old girl is admitted to the PICU with influenza A and pneumococcal pneumonia with associated ARDS and septic shock. She is sedated and mechanically ventilated. Over the last four hours, she's had progressive hypotension requiring crystalloid boluses and escalating dosing of an epinephrine infusion. You're in the process of placing an arterial catheter when the bedside nurse alerts you to bradycardia on the patient's monitor. Her heart rate is 48. The noninvasive blood pressure cuff is cycling but isn't providing a numerical value. You note a weak but present central pulse. Pulse ox is 94% prior to the pulse oximeter no longer reading a numerical value. What is your next action? Increase the epinephrine infusion, begin CPR, administer epinephrine, prepare for transcutaneous pacing, or administer atropine? The answer is B, begin CPR. So this is a patient who has bradycardia and poor perfusion, for which CPR is recommended in the pediatric patient. First we see that the heart rate is 48, which is below that threshold of 60. We then look for signs of systemic poor perfusion. So we see that she's sedated and mechanically ventilated, so we don't have breathing or neurologic status to go by, but we do have weak central pulses and likely absent peripheral pulses as signified by the pulse ox no longer picking up. So we can assume that this patient has poor perfusion, has bradycardia, and therefore we should begin CPR. Increasing the epinephrine infusion would be a reasonable simultaneous step. Introducing a codose of epinephrine in choice C would also be very reasonable, though there should be no delay in actually beginning mechanical CPR. Transcutaneous pacing could be reasonable in refractory bradycardia, but first we should immediately provide temporary cardiac output with CPR. And atropine could also be reasonable if we thought that there was a pathologic increase in vagal tone. But according to PALS guidelines, CPR and epinephrine should be our first two interventions. And we now know, based on a few recent observational studies, that bradycardia and poor perfusion is the initial CPR rhythm in more than half of pediatric cardiac arrests. Unsurprisingly, these patients have better outcomes than those who are pulseless at the time of CPR onset, though about half of them actually become pulseless during the course of CPR. And those patients do worse than either those who were never pulseless and just remained in bradycardia and poor perfusion, or those who were initially pulseless with a rhythm of PEA or asystole. Among patients who are pulseless at the onset of CPR, the vast majority have a non-shockable rhythm of pulseless electrical activity, or asystole, while a minority have shockable rhythms of VF or pulseless VT. This is registry data here on the right. It's about 15 years old now. But what I want to point out is that of 272 patients with VF or VT, only about 40% actually have VF or VT as their initial rhythm, while the majority of shockable rhythms were actually detected after the onset of cardiac arrest and CPR and occurred as a secondary rhythm. And the proportions here, though this is older data in terms of survival outcomes among these different groups, really hold steady. So patients with initial ventricular fibrillation or ventricular tachycardia have better outcomes than those with initial shockable rhythms of PEA or asystole. But those who develop subsequent VF or pulseless VT after PEA or asystole tend to do the worst. That wraps up epidemiology, so now we'll actually get into the fun stuff and talk about what we do for our patients during CPR. I'll briefly touch on basic life support as it serves as the foundation of the pediatric advanced life support that we all provide in our ICUs, and because we care for patients who have received basic life support in the field or prior to their admission to us. I've highlighted a few considerations here that tend to be the things that intensivists infrequently do ourselves and that we tend to get hung up on when we renew our BLS or PALS certifications. So these generally relate to the provision of ventilations and compressions in patients without an advanced airway in place. So any time there's one rescuer, the time to switch between providing ventilations to providing compressions and back takes a few moments. We want to minimize this time and thus minimize the number of switches, and so in these cases we use a 30 to 2 compressions to ventilation ratio. Any time there's an adolescent or adult patient, we similarly provide a 30 to 2 ratio as cardiac etiologies of rest predominate in these populations, and we therefore prioritize compressions. The 15 to 2 ratio therefore only applies to prepubescent children with at least two rescuers. That's because respiratory etiologies are frequently seen in this population and because when you have two rescuers, you can easily complete both of those tasks in tandem. We also see hands-only CPR being highlighted for adults as an alternative to conventional CPR with both compressions and ventilations. We do have data in children with out-of-hospital cardiac arrest that shows us that hands-only CPR is not as efficacious as conventional CPR, and we therefore do not advise this, though guidelines do allow it as a substitute for no CPR at all in the case of bystanders who are unwilling or unable to deliver rescue breaths. So as we begin to discuss the various elements of resuscitation care, I think it's important to consider the overarching goals of CPR. The first is to provide temporary cardiac output through the provision of high-quality CPR, and this is really what chest compressions do. So we have adequate chest compression depth, rate, minimization of interruptions, full chest recoil, and an appropriate ventilation rate. We're also, at the same time, trying to optimize conditions to achieve ROSC. So we're looking for that shockable rhythm that we can defibrillate. We are trying to maximize coronary perfusion pressure and myocardial blood flow to make the heart want to come back. We're going to talk more about that later on. And we're looking for those reversible causes and underlying physiology that we can reverse to end the arrest. We'll begin our discussion of CPR mechanics by discussing chest compression depths. These are the recommended depths by the American Heart Association as well as by other international resuscitation organizations. So for children prior to puberty, we provide depths of at least one-third of the anterior posterior chest diameter. That's about four centimeters in infants and five in older children. In adolescents, we follow the adult guidelines of five to six centimeters. Data supporting the establishment of these goals is pretty lacking, but observational data does demonstrate that when we meet these goals, our patients do better. This is single-centered data that was published back in 2014, looking at compliance with the 2010 guidelines, which are pretty consistent with what we use now, demonstrating that when depth goals were met, patients had higher rates of 24-hour survival as well as ROSC. What about chest compression rates? So as you might expect, there's limited pediatric evidence for ideal chest compression rates during CPR. So in order to simplify training, we match the adult guidelines of 100 to 120. Patients that are too slow will compromise cardiac output during CPR, while those that are too fast will not allow enough time for diastolic filling between chest compressions. And we know from observational studies that when we push faster, we tend to push more shallow and we tend to lean. There's inconsistent associations between chest compression rates during CPR that meet these guideline recommendations with outcomes from CPR, though right now, these guidelines of 100 to 120 are the best that we have. Chest compression fraction refers to the percent of time during a cardiac arrest event during which chest compressions are actually being provided. We know that optimizing chest compression fraction and minimizing both the duration and frequency of interruptions during CPR can lead to more optimal outcomes. But to convey the importance of this, I want you to take a look at these two pictures and think about what they have in common. So on the left, we have a VA ECMO circuit, and on the right, we have someone actively providing chest compressions to a patient with cardiac arrest. And both of these, VA ECMO and chest compressions, are the source of cardiac output for these two patients in question. So I would argue that we shouldn't be any more tolerant of significant interruptions in chest compressions than we would of significant interruptions in ECMO flow during a VA ECMO run. Chest compression interruptions are harmful, firstly, because they are an interruption in cardiac output and therefore no flow time to the brain and other organs. But they're also harmful because their effects potentially last longer than the interruption themselves. So here we can see laboratory data from about 20 years ago in swine, where both the aortic and right atrial pressures, seen in pink and yellow respectively, are being measured during CPR with a 15 to 2 ratio. And these are the interruptions in chest compressions to facilitate the two rescue breaths. The difference between aortic and right atrial pressure, as we'll talk about in more detail later on, is the coronary perfusion pressure, seen there in the white brackets. And this value is important for CPR quality, as it's directly related to the chances of return of spontaneous circulation. So as you can see, during these sequential interruptions, aortic pressure declines and coronary perfusion pressure declines with it, which you would expect during a period of no flow. However, what I want you to note is that it takes time after resuming chest compressions for that coronary perfusion pressure to get back up to the level that it was previously at. So can we actually do anything about it? And here's some really practical strategies for minimizing interruptions, both in terms of frequency and duration, during CPR. The first really gets to multitasking. So time your compressor switches with other things that need to happen during a chest compression interruption, like pulse checks, rhythm checks, perhaps a point of care ultrasound on the chest or something like that. Second, use a coach and coach through the compressor switches. This is where the idea of a dedicated CPR coach independent of the code team leader comes in, with some recent simulation data showing that it offloads the leader in terms of what they have to focus on during cardiac arrest and really can allow them to think more about the broader picture, and it also enhances the actual quality of CPR. When possible, having your next compressor on the opposite side of the bed as the current compressor may decrease the amount of time that you have to have your hands off the chest. If you have an arterial catheter in place, you can tell pretty quickly whether or not your patient is still pulseless during a pause, and you don't have to wait that whole 10 seconds, which is really an eternity. And of course, go with the classic teaching that if you do see a shockable rhythm, go ahead and continue chest compressions while the defibrillator is charging, clear and stay safe as you shock, but then immediately resume compressions post-shock and continue them for up to that full two minutes that's recommended. Institutions that practice doing just these things, which are relatively simple interventions, can really consistently have chest compression fractions for most of their events over 90%. Allowing full chest recoil and avoidance of leaning is important as it facilitates venous return during CPR. We have some pretty convincing evidence from a piglet model of pediatric CPR in which they measured myocardial blood flow as well as cardiac index during CPR, and then applied a controlled amount of residual lean to the chest of 10 and 20%, and saw that the myocardial blood flow went down significantly and substantially, as did the cardiac index. Really just a 20% lean, decreased both of those by half and nearly half. So that covers chest compressions. Now we'll discuss ventilations during CPR. So remember without an advanced airway in place, we interrupt chest compressions in order to deliver ventilations, and the ratios of chest compressions to ventilations is 15 to 2 in prepubescent children and 30 to 2 in older children and adolescents. For patients with an advanced airway in place, there's a new recommendation from the 2020 AHA guidelines that tells us to give 20 to 30 breaths per minute. And this is a pretty substantial change from a previous recommendation of 10 breaths per minute. So let's dive into where that came from a little bit. Prior to discussing these new ventilation rate recommendations, it's important that we have a sense of the logic behind the previous ventilation rate recommendations that persisted through the 2015 guidelines. So why were we told to ventilate at what is such a low ventilation rate relative to the native respiratory rate of a child? Well, the physiologic principle behind this is that we were trying to avoid the potential adverse hemodynamic effects of overventilation, which would lead to higher mean intrathoracic pressure during CPR. As was seen in this animal study by Tomoffer-Heide in 2004, higher ventilation rates led to higher intrathoracic pressure, higher right atrial pressure, and lower coronary perfusion pressure. This approach is further justified by the fact that during CPR, there is relatively low pulmonary blood flow, and thus a low minute ventilation matches this. Further, these recommendations were extrapolated from adult guidelines. Adults have a lower native respiratory rate. And in the absence of pediatric data to support a different approach, these physiologic data and the principles behind them won out. This lack of pediatric data regarding ventilation rates during CPR began to be addressed by this 2019 multicenter observational study of 47 patients with 52 unique CPR events. Strikingly, even though the recommended rate at which we were told to ventilate patients during this time was 10 breaths per minute, the average ventilation rate among these experienced ICUs was 30 breaths per minute. This ranged from 14 to 62. So none of these 52 CPR events actually came close to pediatric guidelines at the time. Furthermore, looking at these spline curves that demonstrate the association between these ventilation rates and the predicted probability of survival, we see that the highest rates of survival are closer to rates in the 20s to 30s, even approaching 40 in children under one year. So the principal finding of this study, this association of relatively higher ventilation rates during CPR with better outcomes, is really the predominant factor that's influencing these higher ventilation rate recommendations by the American Heart Association. I think it's really important to note that this is coming from just 52 CPR events in 47 patients. So this is highly limited, but at this time, it is the best we have. But what else is factoring in here? Well, the second finding that I think is even more impressive from that study is the fact that none of these clinicians at these leading ICUs were actually providing the ventilation rates that were recommended by the guidelines at the time. And some of this suggests that clinicians at these sites may not have actually been okay with ventilating at such a low rate. Why might that be? Well, as we mentioned earlier, children frequently have significant respiratory pathology prior to CPR. And so ventilating a child who has significant respiratory failure as the etiology of their CPR, or at least as something that's happening at the time that they require CPR for another reason, is counterintuitive to clinicians. So I would say that the likely right answer for how we should ventilate children during CPR honestly depends on the specific characteristics of that patient and the cardiac arrest events itself. However, the likely right test answer, at least in the last couple of years, is 20 to 30 breaths per minute in infants and children with an advanced airway in place at the time of CPR. So I've mentioned that the majority of patients are already intubated at the time of an in-hospital cardiac arrest. But what about those who are not? So intubation during CPR is actually associated with worse outcomes among patients who don't already have a breathing tube in place. This may be due to interruptions to facilitate intubation, leading to lower chest compression fractions, due to diverted attention from other components of the resuscitation and toward airway management, due to unrecognized esophageal intubation, due to technical difficulties with intubating during CPR, and difficulties with interpreting enzytal CO2 values after intubation during CPR. But we also know that the scheduled interruptions that happen as a component of CPR when an advanced airway is not in place are also harmful in terms of the effects on chest compression fraction and CPR hemodynamics. With this data in mind, the AHA is actually not able to make a recommendation about what we should do in patients without a protected airway in place at the time of CPR. The best answer is that we should consider cardiac arrest etiology, the likelihood that we can successfully resuscitate a patient without intubating them, and our available resources, both in terms of personnel and equipment. While the minority of pediatric patients actually require defibrillation during CPR events as an important component of resuscitation care, the time to defibrillation, the first defibrillation attempt has been studied in both adults and pediatrics in terms of its association with outcomes. In adults, there's classic literature to show that there's a linear increase in mortality by about 7% to 10% per minute of delay of the first defibrillation attempt. An observational study in pediatric and hospital cardiac arrest failed to demonstrate this association, though it's important to note that these patients had substantially shorter intervals from CPR commencement to the first defibrillation attempt of about one minute, so we were probably underpowered to actually find those differences. Also, the application of high-quality CPR while preparing for defibrillation increases the chances of first shock success. Some practical points. First, get the pads on the chest and be vigilant throughout the arrest, as many patients can have secondary VF, or pulseless ventricular tachycardia. We should minimize our perishock interruptions by providing chest compressions through defibrillator charging and immediately after the shock is delivered. And I just listed here the actual recommended energy for the first and subsequent defibrillation attempts, as this is easy test material. We'll shift gears towards CPR pharmacology with the following question. So which of the following describe the primary purpose of epinephrine during cardiac arrest? So chronotropy, inotropy, afterload reduction, or afterload augmentation? You can take a minute. And the answer is D. The principal job of epinephrine during cardiac arrest is to increase systemic vascular resistance, thereby driving up your diastolic blood pressure, coronary perfusion pressure, and increasing myocardial blood flow. For any of the other mechanisms to really work, it invokes the need to have native cardiac output, which remember you don't have during cardiac arrest. Now for a child with bradycardia and poor perfusion, chronotropy and inotropy might make sense for why epinephrine early on may help. But the right answer here in terms of the main mechanism by which epinephrine works is D. Epinephrine appears in all of our pediatric advanced life support algorithms and really is the mainstay of pharmacotherapy for cardiac arrest and CPR. We know that delayed first doses of epinephrine are associated with worse outcomes. So in the most recent guidelines, there's really a renewed emphasis on giving early epinephrine for patients, particularly those in non-shockable rhythms, whereas for those patients in VF or pulseless VT, the emphasis is on delivering that first defibrillation. It's recommended that we give epinephrine every three to five minutes during CPR. However, it is important to know that there is conflicting data on the optimal frequency at which epinephrine should be administered, with some observational studies suggesting that we should give it more frequently, while others suggest that we should give it less frequently than is currently recommended. We do have data from a trial in the early 2000s that demonstrated that high dose epinephrine, i.e. that is 10 times our current recommended dose, was harmful. And then what about vasopressin as a vasoactive during cardiac arrest? So vasopressin previously appeared in ACLS guidelines as an optional therapy after initial dose of epinephrine. It has been removed from ACLS guidelines really to simplify those guidelines. So head-to-head studies of vasopressin and epinephrine typically did not show substantial differences between the two. So in order to simplify the teaching surrounding this, it has been removed. However, it may have a role in specific disease states, perhaps in patients with catecholamine refractory shock, who may not be expected to respond to more catecholamines during CPR, or in patients who just fail to respond to their initial doses of epinephrine, among those who have some sort of physiologic monitoring in place. Other drugs used during cardiac arrest include the antiarrhythmics lidocaine and amiodarone. There have been some studies in the recent years that have demonstrated improved rates of ROSC with lidocaine compared with either amiodarone or no antiarrhythmic. But since amiodarone was in the guidelines first, it got to stick around. So the guideline calls for either amiodarone or lidocaine to be used for shock, refractory, VF, or pulseless. And that's really it for the core pharmacotherapy that's a component of our primary cardiac arrest algorithms. Two others to touch on are calcium and sodium bicarbonate. Really, we always have to talk about these because they're both commonly used in observational studies, though these same studies show that both are associated with worse survival outcomes. So they're really only recommended when there is a specific indication, such as calcium channel blocker overdoses, sodium channel blocker overdoses, or hyperkalemia. Extracorporeal CPR has been increasingly used over the last many years for refractory cardiac arrest. In this study highlighted here by Havi Lassa using the Get With The Guidelines resuscitation database, they demonstrated that in patients who required at least 10 minutes of CPR, those who received eCPR fared better than those who had conventional CPR. This was one of several observational studies that informed the AHA in their most recent guidelines in which they stated that eCPR may be considered for pediatric patients with cardiac diagnoses who have in-hospital cardiac arrest in settings with existing ECMO protocols, expertise, and equipment. And that's not to say that pediatric patients without cardiac diagnoses should not receive eCPR, just that there isn't quite enough evidence to routinely support its use. Whenever we discuss extracorporeal CPR, we have to talk about CPR duration. Here we see data from Matos and colleagues from 2013, representing data from the first decade of the century, actually. And what I want to point out here is the black hashed line, which is all patients. This shows that with increasing duration of CPR, as you would expect, the survival likelihood goes down. However, this really demonstrates that prolonged CPR is not necessarily futile. So this is not ROSC, this is survival to hospital discharge, and it includes both patients with and without extracorporeal support as a part of their resuscitation. But look, for example, at the 60-minute mark, more than 10% of patients receiving 60 minutes of CPR actually can be predicted to survive the hospital discharge, which is really striking. And these data are all more than 10 years old. We'll now briefly discuss principles of resuscitation physiology and how we can use this to better care for our patients. So we've learned our guidelines, we're going to follow them, we're going to push hard, we're going to push fast, but not too fast, we're going to minimize interruptions, give up an effort every three to five minutes. Isn't that enough? Well, I'm going to give you a few reasons why it may not be. Doing those things, focusing on the guidelines, focusing on our compliance, and really a view of pediatric and hospital cardiac arrest as an intervenable problem that we can do something about, was in large part responsible for the improvements that we saw in outcomes over the course of 2000 to 2010. But as I mentioned earlier, our outcomes have plateaued, and if we want to continue to improve our outcomes from this disease process, we may have to think about something different. Second, our algorithmic approaches to CPR are particularly useful when we have a single rescuer, when we're training lay people, when you're hurtling down a highway in the back of an ambulance, or when you have an institution that's not quite used to caring for these patients. But the vast majority of children who have cardiac arrests are cared for by pediatric intensivists at the time that they have an arrest. We have well-trained, well-equipped teams that perhaps can have a more sophisticated approach to certain aspects of resuscitation care. And lastly, our patients are not all the same prior to their cardiac arrest, and they're not the same during their cardiac arrest either. In-hospital cardiac arrest patients have cancer, maybe they've received CART therapy, they have coronary artery disease, or in our case, more likely congenital heart disease, they have ARDS, and they have refractory septic shock. And yet at the moment that they're at their sickest, we try to boil everything down to these two cards. And that's not to say that we shouldn't use those CPR quality metrics, but that we should actually be measuring in real time whether or not we're meeting them, and whether or not our patient is benefiting from them. This isn't a new concept, it's been endorsed by the AHA for several years, and there's a couple of different overarching themes for how to do this. The first, most of us do, and that's measuring how we're actually doing in terms of meeting those metrics. A lot of our defibrillators now have CPR mechanics monitoring that tells you whether or not you're going too fast, too slow, not deep enough. There's also apps that you can use and put your cell phone right on the chest. We have dedicated CPR coaches who can ensure that compression by compression, we're doing what we need to do. What we're really going to focus on here is the physiologic response, or how the patient is responding to our resuscitation therapies. The three primary ways that we have good evidence for are measuring the coronary perfusion pressure, the diastolic blood pressure, and end tidal CO2. To understand why we would care about blood pressure during cardiac arrest, we have to discuss CPR physiology briefly. So like any organ, the heart has a perfusion pressure determined by the difference between the upstream and downstream pressures, which in this case are the aorta and the right atrium during diastole. The difference between these is the determinant of myocardial blood flow or coronary perfusion pressure and is directly correlated with the likelihood of achieving return of spontaneous circulation in patients with cardiac arrest. This is long-standing data that's more than 30 years old now showing the correlation between coronary perfusion pressure during cardiac arrest and survival outcomes. Since most pediatric and hospital cardiac arrests happen in ICUs, and about half of those patients have an arterial catheter in place at the time of arrest, measuring arterial blood pressure during cardiac arrest is actually feasible in many of our patients. Because simultaneous measurement of right atrial or central venous pressure and arterial blood pressure and calculation of coronary perfusion pressure isn't necessarily feasible, diastolic blood pressure is recommended as a surrogate for coronary perfusion pressure. And in a relatively recent multi-center observational study of PICU patients with an arterial catheter in place at the time of in-hospital cardiac arrest, these diastolic blood pressure measurements were associated with outcomes with thresholds of greater than 25 in infants and greater than 30 in children over the course of a cardiac arrest event being associated with improved rates of survival. So let's now look at the physiologic rationale for monitoring end-tidal carbon dioxide during CPR. So here you have a patient who's receiving chest compressions, which during cardiac arrest is the source of cardiac output. This allows for gas exchange in the tissues where CO2 is picked up as a result of cellular metabolism. We then have pulmonary blood flow, which is also generated by those chest compressions. We exhale CO2 such that better chest compressions lead to more blood flow, both cardiac output and blood flow to the lungs, thereby generating higher values of end-tidal CO2. Not surprisingly then, end-tidal CO2 has been demonstrated to be correlated with chest compression depth during CPR. The relationship between end-tidal CO2 and survival is a little bit more complicated. So we do know that very low values such as those less than 10 millimeters of mercury in both classic and modern literature have been nearly universally incompatible with survival. However, when we include values that are higher than 10 millimeters of mercury and try to correlate those with the likelihood of return of spontaneous circulation, it gets a bit more murky. So two recent pediatric and hospital cardiac arrest studies have actually had conflicting results regarding whether or not there is an association between end-tidal CO2 values and survival. Regardless, it is important to note that a rapid increase in end-tidal CO2 during CPR could signify that underlying ROSC is present. As I've alluded to, both of these methods of measuring CPR physiology do have their limitations. First, when thinking about measuring blood pressures, we're really trying to get at how much blood flow is going out to the body or to a specific organ, and pressure is not necessarily correlated with flow as resistance also factors in. And you do need an invasive arterial catheter in order to measure diastolic blood pressure. For end-tidal CO2, vasopressors have an effect by increasing VQ mismatch. Concomitant lung disease, which most of our patients have at the time of cardiac arrest, can play a role in the end-tidal CO2 values, and you do need to have a patent airway. Other modalities that have been increasingly used include point-of-care cardiac ultrasound, as well as NEARs to measure cerebral physiology. There's a lot of really interesting observational data on both of these, but we're not going to get into that in detail. So how can understanding and monitoring this physiology actually impact what we do during cardiac arrest? And I'll share a couple different perspectives on translational animal work that's been done in this area. Our group at CHOP has explored a hemodynamic-directed model of cardiopulmonary resuscitation, whereby we measure blood pressure during CPR, titrate chest compression force and depth to the systolic blood pressure goals that we set, and then titrate vasopressor administration to maintain a coronary perfusion pressure or diastolic blood pressure above our goals. This has resulted in superior rates of ROSC and 24-hour survival, as well as markers of brain perfusion and injury, relative to guidelines for cardiac arrest. Similarly, our colleagues down at Johns Hopkins have looked at end-tidal CO2-guided CPR, in which chest compression depth and rate are changed according to the end-tidal value, and they've also demonstrated superior physiologic parameters and superior rates of survival with this model compared to standard CPR. Physiologic monitoring is one way to personalize CPR care, and we've done this across several models. Physiologic monitoring is one way to personalize CPR care for an individual patient, but we also must explore on a patient-by-patient basis reversible causes of cardiac arrest and the underlying physiology that can potentially be targeted. The classic H's and T's provide a teachable framework by which we can remember many of these, but it's important to have a more focused approach when able, based both on the likely etiology of the arrest and the patient's underlying conditions, as well as the characteristics of the arrest. For example, when you have a patient with a known history of venous thromboembolism, and then they have very low end-tidal CO2 despite what appears to be high-quality CPR, think PE. Let's go through a couple of relevant examples. So here we have a five-year-old girl who's admitted to the PICU with pneumococcal pneumonia and hemolytic uremic syndrome. She's intubated and mechanically ventilated for acute respiratory failure, has anuria and a serum potassium level of 6.4 on this morning's labs. You're planning to insert a percutaneous dialysis catheter in her right IJ. While prepping the site, you note a widened QRS on the bedside monitor. Minutes later, the nurse points out a declining end-tidal CO2 value in the monitor alarms for asystole. In addition to high-quality CPR, what additional therapies should be considered? So the widened QRS here alludes to the fact that the patient is having a hyperkalemic cardiac arrest. So calcium gluconate, sodium bicarbonate, and repeat serum electrolyte assessment would be the best next step. Same patient, same problems. Now your fellow is inserting a percutaneous dialysis catheter in the right IJ. While dilating the vessel, you note tachycardia, hypotension, and increasing peak inspiratory pressures. Moments later, the nurse points out a declining end-tidal value and you no longer feel ephemeral pulse. In addition to high-quality CPR again, what additional therapies should be considered? Here we're steering you towards the patient having an iatrogenic chemothorax. Rapid bedside evaluation of this, immediate thoracostomy, and crystalloid resuscitation while awaiting blood products would be the correct answer here. We'll now discuss elements of post-cardiac arrest care in children. We've discussed that a remarkable 80% of our patients who require CPR in the hospital will have return of spontaneous circulation, but this isn't a time to pat ourselves on the back as only 40 to 45% actually survive to discharge. So our patients are twice as likely to die post-arrest as they are during arrest itself. The post-cardiac arrest syndrome formally described in 2008 is a SERS-like inflammatory response that's similar to sepsis, or the post-pump slump that we see after cardiopulmonary bypass. It has four main parts, which are brain injury, myocardial injury, the systemic ischemia reperfusion response, and persistence of the precipitating pathophysiology of the arrest. The principles of what we do for our post-arrest patients are pretty simple. First, keep everything normal. Normal blood pressures, normal or low temperatures, normal oxygen levels, normal CO2 tensions, and treat and prevent seizures. In order to successfully do these things in real time, we have to monitor our patients. Comatose survivors of cardiac arrest should, if able, have an arterial catheter in place, have continuous temperature monitoring, have serial blood gases in addition to continuous monitoring with pulse ox and end-tidal CO2, and should be placed in continuous EEG if available. I'll go into a bit of detail on specific aspects of post-cardiac arrest physiology and management. We'll start with the heart. So, post-arrest myocardial dysfunction is common, and in and of itself is associated with an increased odds of death. Combined with the systemic inflammatory response, it frequently causes hypotension, and hypotension, as determined by the systolic blood pressure being less than the fifth percentile for myocardial dysfunction, as determined by the systolic blood pressure being less than the fifth percentile for age at any point after arrest, is also associated with increased odds of pre-discharge death. So blood pressure monitoring, preferably with an arterial catheter, is important after cardiac arrest, as is echocardiography, and use of both fluids and inotropes or vasoactive drugs, depending on the physiology that's being observed at the bedside, is important to maintain a systolic blood pressure greater than the fifth percentile for age. Therapeutic hypothermia and targeted temperature management are always hot topics in cardiac arrest care. We know that fever is common after cardiac arrest, and that occurrence of fever is associated with unfavorable outcomes. Earlier cardiac arrest studies looking at therapeutic hypothermia were unable to discern whether or not the potential beneficial effects of hypothermia were due to hypothermia itself, or the fact that the patients in those arms of studies avoided fever. So the THAPCA trials compared targeted temperature management at normothermia versus targeted temperature management with hypothermia in both out-of-hospital and in-hospital cardiac arrests. In the out-of-hospital arm, there was no statistically significant difference in the primary outcome of favorable neurologic outcome at one year. Since this is the rare randomized control trial in our field, I thought I would include the actual numbers here. It's important to note that the absolute survival was 20% versus 12%, with a p-value of 0.14. So there is some speculation among experts in the field that statistical power or patient selection may have affected the primary outcome. In terms of adverse effects, hypokalemia and thrombocytopenia were more common in the low temperature group. Renal replacement therapy was more commonly used in the higher temperature group. The in-hospital arm of THAPCA, published two years later in the New England Journal of Medicine, was stopped early due to futility, with those two arms being much closer to each other in terms of the favorable neurologic outcome at 36% versus 39%. So really no signal of a difference there. So what do we do with that? First, aggressively treat and prevent fever. In order to do that, as alluded to earlier, we have to measure and monitor core temperature. Recommended to either be in the esophagus, the rectum, or the bladder. Apply targeted temperature management, either through hypothermia for two days, followed by normothermia for three days, or a full five days of normothermia. Prevent shivering through the use of sedation and neuromuscular blockade as needed. Monitor blood pressure as hypotension can happen, especially during re-warming, which can of course be exacerbated by the myocardial dysfunction that we see, and continue to prevent fever after re-warming if the patient is still comatose. Abnormalities of oxygenation and ventilation are common following cardiac arrest due both to the arrest itself as well as concomitant respiratory failure in many of our patients. There's inconsistent but relatively limited observational data from the bedside regarding the importance of these values in terms of their impact on cardiac arrest outcomes. There is of course physiologic rationale for why these things would matter. Hypoxemia of course can exacerbate ischemia reperfusion injury. Hyperoxia can drive the promotion of reactive oxygen species as was seen in this animal study on the right. Hypercarbia may drive cerebral hyperemia and can also cause acidosis that might exacerbate myocardial dysfunction, and hypocarbia can harmfully decrease cerebral blood flow. So in the absence of convincing data otherwise, the answer is to keep everything as normal as possible, to titrate oxygen immediately after arrest to avoid both hypoxemia or significant hyperoxia, targeting normoxemia with SpO2 of 94 to 99 percent, and to target normocapnia with arterial CO2s in the 35 to 45 range or values that are specific to the patient's condition, so higher values are okay in your chronic CO2 retainers. And we should both monitor end-tidal CO2 and follow serial blood gases as many of these patients have VQ mismatch or dead space ventilation. Finally, seizures are common following cardiac arrest occurring in between 10 and 50 percent of comatose cardiac arrest survivors. These are associated with worse outcomes, though whether they are a driver of those outcomes or just a symptom of bad brain injury is unclear. Continuous EEG monitoring is recommended, and it's recommended that we treat both clinical and electrographic seizures, though prospective studies are needed to determine whether or not this actually improves neurologic outcomes after arrest. Other common elements of post-cardiac arrest care include glucose control, hypoglycemia is harmful and should be avoided, and severe hyperglycemia should be avoided as well. Routine corticosteroids are not recommended, nor is the use of other immunomodulators unless there's a specific indication. Sedation and neuromuscular blockade are often necessary, even in the brain injured patient, to facilitate other therapies and to avoid shivering that often comes with targeted temperature management. Serious infections do occur frequently after cardiac arrest, most notably pneumonia and sepsis, so we should be monitoring for signs of infection, but there's no clear role for either routine cultures or routine antimicrobial agents. And acute kidney injury is a common organ dysfunction following cardiac arrest, so we should always be monitoring urine output and electrolytes cautiously using nephrotoxic or renally cleared medications. There are multiple patient characteristics from prior to cardiac arrest, intra-arrest characteristics and features, and post-arrest signs, symptoms, and findings that are associated with outcomes from cardiac arrest. It's important to note that no single variable can adequately or accurately predict outcome in and of itself, so we really must take into account the overall picture of the patient at all stages of arrest and post-arrest. It's also important to know that neurologic recovery may take time. There are many adult studies now that show that recovery can happen after being comatose for several days, and the guidelines are to wait at least 72 hours after return to normal thermia for patients being treated with targeted temperature management or 72 hours after cardiac arrest for patients not being treated with TTM. In the absence of specific pediatric evidence or guidelines regarding this, we usually abide by the same timeline. In summary, pediatric cardiac arrest has diverse etiologies and characteristics. We have pretty reasonable rates of survival, but progress has stalled in recent years. Pediatric CPR targets are based on limited data, but they do give us something to go by, and achieving those targets has been associated with improved outcomes. Measuring and targeting physiology during CPR can also be a way to improve resuscitation care, and the essentials of post-arrest care are to maintain normalcy and to measure what we do. Thank you for your time and attention, and have a great day.

pediatric cardiac arrest

cardiopulmonary resuscitation

high-quality CPR

ventilation rates

coronary perfusion pressure

etiologies of cardiac arrest

post-cardiac arrest care

seizure prevention

resuscitation outcomes