Rudolph's Pediatrics, 22nd Ed.

CHAPTER 108. Cardiopulmonary Resuscitation

Mary Fran Hazinski

Although resuscitation is a widely taught routine, a formalized approach to resuscitating an infant or child is a relatively recent development. The first published report of successful closed-chest resuscitation in 1960 contained details about the resuscitation of a 9-year-old with respiratory arrest and a 12-year-old with anesthesia-induced cardiac arrest. The first guidelines for pediatric basic life support (PBLS) and neonatal resuscitation were in 1980.1 Initial guidelines for pediatric advanced life support (PALS) were published in 1986.2 Recent research, prehospital series, and inhospital registry data have provided additional information about the epidemiology, presentation, and outcome of pediatric cardiopulmonary arrest (CPA) at different ages. As a result, the 2005 AHA guidelines recommended applying pediatric basic life support guidelines to children up to approximately age 12 (or whenever the physical signs of puberty begin) and emphasized differing priorities based on type and circumstances of arrest. For all victims of CPA, the guidelines now emphasize the importance of providing effective chest compressions with minimal interruptions.

It is now clear that resuscitation must be tailored to the type of cardiopulmonary arrest. Evidence in adults suggests that effective protocol-driven postresuscitation care can improve intact neurological survival from cardiopulmonary arrest. Every step in resuscitation is important, including recognizing prearrest conditions, identifying the arrest itself, supporting appropriate oxygenation and ventilation, delivering effective chest compressions with minimal interruptions, and, if defibrillation is needed, seamlessly integrating shock delivery with CPR. Thus, pediatric health care providers must prevent cardiac arrest when possible and must be prepared to perform skilled resuscitation and postresuscitation care when needed.


Estimating the incidence of cardiopulmonary arrest in children has been complicated by inconsistent definitions of arrest and limited data. International resuscitation experts have developed two templates for reporting outcomes of pediatric resuscitation. An international consensus template for reporting inhospital arrest was reported in 19974 and was updated in 2002.5 This in-hospital template is now used in data collection for the National Registry of Cardiopulmonary Resuscitation (NRCPR), a prospective multicentered observational registry of in-hospital cardiac arrests and resuscitation, established by the American Heart Association.6 The National Registry of Cardiopulmonary Resuscitation (NRCPR) is a prospective multicentered observational registry of in-hospital cardiac arrests and resuscitation, established by the American Heart Association. This registry (available at now includes information from more than 100,000 resuscitations (including thousands of pediatric resuscitations) from more than 30 hospitals in North America.7

Respiratory arrest is defined as the cessation of spontaneous respiratory activity, or apnea, in the presence of spontaneous circulation (palpable central pulses). Detection and immediate treatment of respiratory arrest often prevents progression to cardiopulmonary arrest. Cardiopulmonary arrest is defined as the cessation of effective cardiac mechanical function and systemic perfusion. The victim is unconscious/unresponsive, with apnea or agonal gasps and with no palpable central pulses. The patient with cardiopulmonary arrest may demonstrate agonal gasps that may be mistaken for effective spontaneous breathing.8,9

There are two major types of cardiopulmonary arrest (CPA): those that occur secondary to progression of respiratory failure or shock (“asphyxial” arrest) and those that occur as a sudden primary cardiac event, called sudden cardiac arrest (SCA). These types of CPA can be identified by terminal cardiac rhythm. Bradycardia and pulseless electrical activity are the most common terminal rhythms associated with asphyxial arrest, while ventricular tachycardia (VT) and ventricular fibrillation (VF) are the most common terminal rhythms associated with SCA. The incidence of each type of CPA varies with the child’s age and condition and with the location and circumstances of the arrest. It is important to identify the likely type of arrest, because effective resuscitation will require different priorities based on arrest type, the number of rescuers, and the location of the arrest.

Asphyxial arrest is the most common type of arrest observed in infants and children. However, this type of arrest can occur in both adult and pediatric patients with conditions such as respiratory failure, trauma, drowning, poisonings, metabolic disorders, and shock. The most common causes of asphyxial arrest are consistent with the most common causes of childhood death including birth-associated asphyxia and sudden infant death syndrome (SIDS).

Sudden cardiac arrest (SCA) typically results from an arrhythmia. This type of cardiac arrest is the most common arrest studied in adults but is less common in children. SCA has been reported in young athletes following a blow to the chest (commotio cordis) and in children with genetic impulse formation and conduction anomalies (eg, prolonged QT syndrome), congenital heart disease, coronary artery abnormalities, cardiac inflammation, cardiomyopathy, and drug toxicity. Ventricular arrhythmia with the development of ventricular tachycardia and ventricular fibrillation typically precipitates SCA.


Sudden cardiac arrest (SCA) is not a reportable cause of death in adults or children in the United States, so estimates of the incidence of asphyxial arrest versus SCA are based on extrapolation from published case series that vary widely in inclusion and exclusion criteria.

A 2006 US prehospital case series of 272 children, ages 1 to 18 years with cardiopulmonary arrest documented important age-related differences in cardiac arrest rhythm and differences among victims with witnessed versus unwitnessed arrest.12 Important qualifiers of this report are that it excluded victims of trauma (the leading cause of death), and approximately two thirds of the victims received bystander CPR (about double the national average). Overall incidence of ventricular tachycardia (VT) or ventricular fibrillation (VF) was 18.6%. VT or VF was documented in only 7.5% of children ages 1 to 7 years but was documented in 27% of children and adolescents ages 8 to 18 years. The incidence of VT/VF was higher in witnessed (31%) than in unwitnessed (10%) arrests and was highest (nearly 50%) in victims who were ages 13 to 18 years with witnessed arrest.12

Cardiac arrest in schools, particularly in athletes, has received a great deal of attention in the media and has stimulated grassroots efforts regarding screening of athletes and developing CPR and AED programs in the schools. Extrapolation from a Minnesota statewide database suggests that approximately 1 per 100,000 athletes develops SCA annually.13 In the United States, more than half of the sudden cardiac arrest deaths in athletes are attributed to hypertrophic cardiomyopathy, commotio cordis (a blow to the chest that triggers ventricular tachycardia or fibrillation), or coronary artery anomalies.13 Many episodes of sudden cardiac arrest in athletes are witnessed.

In a 15-year retrospective review of EMS responses to cardiac arrests in Seattle and King County, Washington, a cardiac arrest occurred, on average, in 1 per 111 elementary schools, high schools, or colleges per year.14 Most cardiac arrests occurred among faculty, staff, or visiting adults. Of the 97 documented arrests that occurred in schools over a 15-year period, 12 arrests were among students; half of the student victims were ages 3 to 18 years and had a prior history of cardiopulmonary disease or severe cardiopulmonary disability. The AHA recommends that schools develop a planned and practiced response for dealing with all medical emergencies, including cardiac arrest, with many teachers and high school students trained in CPR.15


A series of publications from the NRCPR enable characterization of in-hospital arrest in children ages 18 or younger. In 880 pediatric cardiopulmonary arrests,7 most (58%) children demonstrated prearrest respiratory insufficiency, and many demonstrated shock (36%) or congestive heart failure (31%), which are consistent with other reports.16,17 The most common intermediate causes of arrest were hypotension (61%), acute respiratory insufficiency (57%), and arrhythmias (49%). Asystole was the presenting rhythm in 40% of cases, with an initial rhythm of pulseless electrical activity (PEA) in 24% of cases and VF/VT in 14% of cases. In 27% of the hospitalized victims, VF/VT was present at some time during the arrest.18


For many years, most information about the pathophysiology of pediatric cardiopulmonary arrest was extrapolated from adult clinical or laboratory (animal) data. Recent published pediatric case series in the prehospital and inhospital settings have provided more specific information about pediatric arrest. Both asphyxial and arrhythmic arrest can be divided into four major phases: the prearrest phase, the arrest phase, the CPR phase, and the postresus-citation phase.


The child with asphyxial arrest typically develops respiratory failure or shock and often develops bradycardia prior to the arrest. Bradycardia in children may also result from vagal stimulation (eg, suctioning, gagging) or from hypoxia. In the minutes before an asphyxial arrest, oxygen delivery to the tissues is compromised both by low arterial oxygen content and by inadequate blood flow, and major organ ischemia is likely to be present even before the arrest occurs.

Arrhythmic arrests produce sudden cardiac arrest. In the adult, such ventricular arrhythmias are often associated with coronary artery disease and an acute coronary occlusion. In children, pulseless ventricular arrhythmias typically result from underlying hypertrophic cardiomyopathy, channelopathies, congenital heart disease, drug toxicities, or electrolyte imbalances. The victim of sudden cardiac arrest typically has normal arterial oxygen content and oxygen delivery and excellent organ perfusion until the moment of the arrest. Some athletes or children with arrhythmias such as those due to ion channel anomalies (eg, long QT syndrome) may have syncopal episodes prior to the cardiac arrest event. Vigorous exercise can act as a trigger for lethal ventricular arrhythmias.13

In the National Registry of Cardiopulmonary Resuscitation data of in-hospital arrest, most pediatric victims who presented with ventricular tachycardia (VT) or ventricular fibrillation (VF) had respiratory insufficiency (54%), congestive heart failure (41%), hypotension (26%), or pneumonia or sepsis (21%). Approximately half of the children who developed VT or VF arrest were classified with a medical cardiac illness or surgical cardiac condition.18 In an NRCR series of 1005 index children with in-hospital cardiac arrest, when comparing the 272 patients who developed VT or VF to those who did not, the only prearrest vasoactive infusions were identified as proarrhythmic treatments.18


Once cardiac arrest develops, oxygen delivery ceases. Unless CPR is provided or spontaneous rhythm is restored immediately, the myocardium and all tissues will become progressively ischemic, and the likelihood of successful resuscitation diminishes with every passing minute. Generation of lactic acid, increased cell membrane permeability, production of free oxygen radicals, and activation of inflammatory mediators contribute to progressive organ and tissue destruction.

PEA and Asystole

PEA is a term used to describe cardiac electrical activity that fails to produce sufficient mechanical function to generate a palpable central pulse. Although this term could encompass agonal (eg, bradyasystolic) rhythms, it is typically reserved for patients with narrow QRS complexes that fail to produce a palpable pulse. Reversible causes of PEA in children include tension pneumothorax, cardiac tamponade, hypovolemia, and, rarely, pulmonary embolus. If PEA is not promptly treated, the rhythm will ultimately deteriorate to asystole. Once the patient becomes pulseless, the myocardium will rapidly become ischemic, and survival is poor unless prompt and skilled resuscitation is provided.

Asystole is the ultimate terminal rhythm, characterized by electrical silence. This rhythm is likely to be present for any unwitnessed or prolonged arrest, and it is the most common rhythm reported in pediatric prehospital10,12 and inhospital7 arrest. Although overall survival with asystole or PEA is poor, survival is higher among children than adults (24% versus 11%) who present with these rhythms in in-hospital arrest.7 Once asystole develops, myocardial, brain, and other organ ischemia will progress rapidly.

Ventricular Tachycardia and Fibrillation

Children who present with ventricular tachycardia (VT) or ventricular fibrillation (VF) in the prehospital or in-hospital setting typically have higher survival rates than those presenting with PEA or asystole.7,12,18 Untreated pulseless VT will rapidly progress to VF, and untreated VF will deteriorate to asystole.

When VF or pulseless VT develops suddenly, the alveolar oxygen tension and arterial oxygen content should initially be normal. During the first minutes of sudden VT or VF arrest, oxygen delivery to the heart, brain, and other organs is limited more by blood flow than by oxygen content. Agonal gasps have been reported during the first minutes of arrest in patients and among laboratory animals with sudden VT or VF.

During the first approximately 4 minutes of VF, the amplitude of the VF will be high (so-called “good VF”), indicating that the myocardium initially has adequate oxygen and substrates. At this point, shock delivery is likely to result in elimination of VF, return of spontaneous cardiac rhythm, and return of spontaneous circulation.

If the VF remains untreated for approximately 4 to 10 minutes after arrest, the VF amplitude will gradually decrease (so-called “bad VF”), indicating several myocardial ischemia. At this point, shock delivery is less likely to be followed by return of spontaneous rhythm and return of spontaneous cardiac rhythm, unless some coronary perfusion and myocardial oxygen delivery is restored through CPR.

Untreated ventricular fibrillation (VF) will ultimately progress to asystole. Within approximately 10 minutes of arrest, global ischemic injury is present. Cell death occurs and formation and activation of inflammatory mediators contribute to reperfusion injury. If asystole is present, shock delivery is ineffective.20 Initiating CPR at this point is not likely to prevent progressive organ injury.

In the NRCR data, some children with initial asystole or PEA developed a secondary VF during the course of attempted resuscitation. The survival among children with such secondary VF (ie, VF that appears during resuscitation and is not the primary event in the arrest) is poor (11%), even lower than 27% survival reported among children with presenting rhythms of asystole or PEA.18


Once cardiopulmonary arrest is present, immediate bystander CPR is needed. CPR maintains a small (estimated range of 10% to 33% of normal) but critical amount of blood flow and oxygen and substrate delivery to the heart (ie, coronary perfusion) and brain.21 CPR can prolong VF and can maintain or increase VF amplitude for several minutes. This increases the window of opportunity for shock delivery and increases the likelihood that the shock delivery will be followed by return of spontaneous cardiac rhythm.

Chest compressions create blood flow by increasing intrathoracic pressure and by directly compressing the heart.

Excessive ventilation and incomplete chest wall recoil after each compression will reduce blood flow generated by chest compressions. Excessive ventilation during CPR creates positive intrathoracic pressure, reduces venous return to the heart, and reduces coronary perfusion and systemic and cerebral blood flow. Hyperventilation of intubated adults during resuscitation has been linked with poor survival.24 If rescuers fail to allow the chest wall to recoil completely after each compression, venous return to the heart is decreased, and blood flow and coronary perfusion pressure fall.25,26 Pulmonary blood flow, carbon dioxide delivery to the lungs, and oxygen uptake from the lungs are approximately 10% to 33% lower than normal during cardiac arrest, and victims need only a fraction of normal minute ventilation during CPR, particularly following insertion of an advanced airway.


Following restoration of circulation, ischemic and reperfusion injuries can cause cardiorespiratory instability, perfusion abnormalities, and organ dysfunction. Reperfusion injury is characterized by cell death, calcium entry into cells, and activation of inflammatory mediators. This inflammatory response includes the development of endothelial injury, capillary leak, neutrophial activation, platelet aggregation, and increases in mediators such as tumor necrosis factor and interleukins. Increased production of free oxygen radicals and decreased production of nitric oxide can cause vasoconstriction and further ischemic injury. Hyperglycemia is common in children after an arrest, and both hyperglycemia27 and hypoglycemia28 have been linked with increased mortality in studies of critically ill children.

Children often demonstrate temperature instability during the first 24 hours after resuscitation from cardiopulmonary arrest. A brief period of spontaneous hypothermia is typically followed by the development of hyperthermia.29This hyperthermia is associated with worse neurological outcome.


Treatment priorities differ for the prearrest, arrest, CPR, and postarrest intervals. In recent years, there have been significant advances in treatment approaches during the prearrest, CPR, and postarrest intervals. Ideally, successful treatment of prearrest conditions will prevent the arrest or at least minimize potential organ ischemia before the arrest. Rapid response or medical emergency teams may reduce the incidence of non-ICU pediatric arrests,30 especially respiratory arrests.31 However, the published success rates of these teams have varied widely based on format, activation criteria, and team members.32

Prompt recognition of arrest and initiation of CPR by a bystander should minimize the arrest (no flow) interval. During the CPR interval, all rescuers must provide high-quality CPR to optimize blood flow and oxygen delivery. The AHA 2005 CPR guidelines emphasized the importance of delivering effective chest compressions for all victims.33 The AHA PALS guidelines recommend organizing advanced life support interventions (eg, drug delivery, intubation) around 2-minute periods of uninterrupted CPR. Extracorporeal cardiopulmonary resuscitation (ECPR) is extracorporeal membrane oxygenation instituted during attempted resuscitation, and it has been successful for pediatric in-hospital resuscitation, particularly among children with heart disease (see Chapter 109). In recent series reporting ECPR, intact neurological survival has averaged approximately 30% among children with heart disease34,35 and 55% among children with septic shock who required chest compressions.36

In adults, therapeutic hypothermia and protocols for therapy such as hemodynamic support and respiratory care improved outcomes of patients admitted following cardiopulmonary arrest (CPA) and ROSC.37Although similar studies have not been reported in children, it is likely that improved postresuscitation care can increase survival following CPA.


The appropriate sequence for pediatric resuscitation is determined by the type and location of the arrest and the type of rescuers and equipment available. Although the single rescuer must choose a sequence of actions, multiple rescuers can perform several actions simultaneously. The lone health care provider (HCP) should tailor the resuscitation sequence to the likely cause of the arrest.

For the prehospital arrest of a child or likely victim of asphyxial arrest, CPR focuses on establishing both ventilation and oxygenation. The lone HCP should deliver about 20 cycles or about 2 minutes of CPR before leaving the victim to activate the emergency response system and retrieve an AED. The rescuer should then return to the victim to resume CPR and apply and use the AED. If multiple bystanders are present during prehospital arrest, one rescuer should begin the steps of CPR while others activate the emergency response system and retrieve an AED. Following the sudden collapse of a victim of any age, focus is on immediate defibrillation and support of blood flow.


The child with respiratory arrest (apnea) typically develops a bradycardia with poor perfusion before developing pulseless arrest. If rescuers detect and treat respiratory arrest before the patient develops pulseless arrest, survival is typically 75% or higher, whether the arrest occurs in the prehospital11 or the inhospital7 setting.


High survival has been reported if in-hospital providers detect and treat bradycardia before it progresses to arrest and if they provide chest compressions for symptomatic bradycardia (HR ≤ 60/minute with poor perfusion despite adequate oxygenation and ventilation). In the National Registry of Cardiopulmonary Resuscitation in-hospital series, survival from bradycardia with pulses is higher (60%) than survival from cardiopulmonary arrest (27%).7

Epinephrine is the drug of choice for symptomatic bradycardia unresponsive to support of airway, oxygenation, and ventilation. Rescuers should consider using atropine for symptomatic bradycardia with increased vagal tone or primary heart block. Cardiac pacing may be needed.33


The optimal compression-to-ventilation ratio for pediatric resuscitation is unknown. The AHA recommends a 30:2 compression-to-ventilation ratio for single rescuers and a 15:2 ratio for two or more rescuers, based on expert consensus and prevalence of asphyxial (rather than arrhythmic) arrest in the pediatric population. The ratios are designed to optimize myocardial and systemic blood flow by minimizing interruptions in chest compressions while maintaining adequate arterial oxygen content. Once an advanced airway is placed, rescuers should no longer deliver cycles of compressions and ventilations. Instead, the rescuer performing compressions should deliver continuous chest compressions, and the rescuer providing ventilations should deliver approximately 8 to 10 breaths/minute.

Rescuers must give compressions of adequate rate (approximately 100/minute) and depth (one third to one half the depth of the chest), allowing full chest recoil after each compression. Rescuers should minimize the number and duration of any interruptions in chest compressions (eg, those required for delivery of ventilations, attempted defibrillation, or insertion of an advanced airway). If shock delivery is indicated, rescuers should deliver a shock within 10 seconds or less of the last chest compression, then should resume chest compressions immediately after shock delivery. Shock effectiveness (ie, likelihood of termination of ventricular fibrillation (VF) and ultimate ROSC) decreases with every 10 seconds that elapse between the last compression and the shock delivery.39 When a shock eliminates VF, the most common rhythm for 30 to 60 seconds after shock delivery is asystole or PEA, so compressions are needed for the first minutes after defibrillation attempts.

Open-chest CPR can produce nearly normal blood flow but is impractical for most resuscitation situations. This technique of CPR should be considered in the immediate postoperative period for pediatric cardiovascular surgical patients, when the chest is left open or the sternotomy can be opened quickly.40


Because AED algorithms have been shown to be accurate in interpreting pediatric rhythms,41,42 the AHA recommends AED use for children ages 1 to 8 years in cardiac arrest.43 Ideally, rescuers should use pediatric AED pads that contain attenuators to reduce the shock dose. However, if no pediatric pads are available, rescuers should use the adult AED pads. There is insufficient data regarding AED use in infants to recommend or not recommend using AEDs in infants up to 1 year of age.43

There are very limited data available regarding optimal manual biphasic waveform shock dose for defibrillation of children. Current recommendations for an initial 2 J/kg shock dose for pediatric defibrillation are based on a series of 71 monophasic shocks delivered to 27 children ages 3 days to 15 years.44 In general, weight-based defibrillation dosing may not be optimal, because dose delivered is influenced not only by paddle or pad size and energy dose but also by transthoracic impedance. The impedance is not linearly related to body weight. For in-hospital manual defibrillation, rescuers should use infant pads or paddles for defibrillating infants up to 1 year of age and should use larger pads or paddles for patients 1 year of age and older. If the 2 J/kg dose is ineffective, rescuers should use a 4 J/kg dose for subsequent shocks. More research regarding defibrillation dose is needed.


During resuscitation, the potential benefits of inserting an advanced airway should be weighed against the detrimental effects of the intubation attempt itself. Bag-mask ventilation for short periods may be as effective as ventilation through an advanced airway, but this form of ventilation requires training and periodic retraining.47 Inserting an advanced airway will enable delivery of uninterrupted chest compressions and may reduce gastric insufflation (and its attendant risks of regurgitation and aspiration), but it does require training and experience and will require interruption of chest compressions and oxygen delivery.

Experienced providers should perform intubation, with careful preparation and coordination of rescuer activities to minimize interruptions in chest compressions. Once the advanced airway is inserted, rescuers should confirm placement using clinical examination plus a device such as a qualitative or quantitative exhaled carbon dioxide detector.


Although no drug has been shown to increase survival from pediatric cardiac arrest, the use of vasoconstrictors does increase blood pressure and coronary and cerebral blood flow and return of spontaneous cardiac rhythm in animals. Drugs with beta-adrenergic effects also increase spontaneous myocardial depolarization and contractility. The AHA recommends administering a standard dose of intravenous (IV) epinephrine every 3 to 5 minutes for cardiac arrest.33

During resuscitation, intravenous or intraosseous drug administration is preferable to endotracheal administration (see Chapter 107). Although lipid-soluble drugs can be administered by endotracheal route, drug absorption is poor and unpredictable, and optimal drug doses are unknown.

Intravenous high-dose epinephrine (HDE) is no longer recommended for routine use in pediatric resuscitation, because it was associated with decreased survival and neurological outcomes in a randomized, controlled study.49Although high-dose epinephrine can increase return of spontaneous cardiac rhythm, it can also increase postresuscitation myocardial oxygen consumption, myocardial dysfunction, and hemodynamic instability. High-dose epinephrine may still be considered for some resuscitations, such as beta-blocker overdose.33

Amiodarone is the recommended antiar-rhythmic for use during resuscitation from shock-refractory ventricular tachycardia or ventricular fibrillation, based on extrapolation from reports of its effectiveness in treating life-threatening pediatric arrhythmias.50 The drug can produce hypotension and arrhythmias, so expert consultation is advised when the drug is used.

Magnesium administration is indicated for treating torsades de pointes and when hypomagnesemia is suspected. There is no evidence to support use of lidocaine for treating shock-refractory ventricular tachycardia or ventricular fibrillation in children, but the drug may be considered as an alternative antiarrhythmic if amiodarone is not available.


Once effective spontaneous circulation is established, rescuers should optimize hemodynamic support, provide oxygenation and ventilation, support end-organ function, and control temperature. Therapeutic hypothermia in adults following cardiac arrest51,52 and in neonates with hypoxic encephalopathy53 have improved survival and neurological outcome, but data regarding hypothermia are not available. Rescuers should prevent or promptly treat postresuscitation hyperthermia.29


Unfortunately, there are no reliable predictors that can be used to make the decision to terminate resuscitation efforts. Higher survival rates are associated with witnessed, ventricular tachycardia (VT), or ventricular fibrillation (VF) arrest and with prompt bystander CPR. However, prolonged efforts have been successful, particularly in the treatment of VT or VF sudden cardiac arrest, drug toxicity, in-hospital arrest, and in small children with primary hypothermia.16 Recent emphases on CPR quality, postresuscitation care, and peri-arrest use of life-support devices could all have a significant positive impact on CPR outcomes.


Programs providing resuscitation should develop a process of continuous quality improvement. The process should monitor resuscitation outcomes and techniques and should provide opportunities for retraining on a regular basis. Training with the use of realistic simulators holds promise to improve health care provider skills in and outcomes of pediatric resuscitation. Such processes increase resuscitation quality and outcomes.