These guidelines are intended to be a resource for health care professionals to identify and treat infants and children in the prearrest, intra-arrest, and post–cardiac arrest states. These apply to infants and children in multiple settings: the community, prehospital, and hospital environments. Prearrest, intra-arrest, and post–cardiac arrest topics are reviewed, including cardiac arrest in special circumstances, such as infants and children with congenital heart disease.

For the purposes of the PALS guidelines, pediatric patients are infants and children up to 18 years of age. In contrast, pediatric basic life support guidelines apply to infants and children without signs of puberty. Neither pediatric advanced nor basic life support guidelines address the resuscitation of newborn infants, who are transitioning from a fluid-filled to an air-filled environment. Resuscitation of the newborn infant is addressed in “Part 5: Neonatal Resuscitation”¹⁶. Although pediatric basic and advanced life support guidelines may be applied to newborn infants younger than 28 days of age based on pathophysiology and institutional practice, neonatal guidelines should be followed at birth to address unique aspects of transitional physiology.¹⁷

The PALS Writing Group consisted of pediatric clinicians from the American Heart Association (AHA) and the American Academy of Pediatrics (AAP) including intensivists, cardiac intensivists, cardiologists, emergency medicine physicians, and nurses. A call for candidates was distributed to the AHA Emergency Cardiovascular Care (ECC) Committee and AAP subject matter experts, and volunteers with recognized expertise in pediatric resuscitation were nominated by the writing group co-chairs. Writing group members were selected by the AHA ECC Science Subcommittee and AAP Executive Committee and then approved by the AHA Manuscript Oversight Committee. The AHA and AAP have rigorous conflict of interest policies and procedures to minimize the risk of bias or improper influence during the development of the guidelines. Before their appointment, writing group members and peer reviewers disclosed all commercial relationships and other potential (including intellectual) conflicts. Writing group members whose research informed guideline recommendations were required to declare those conflicts during discussions and abstain from voting on those specific recommendations. This process is described more fully in “Part 2: Evidence Evaluation and Guidelines Development.”¹⁸ Comprehensive disclosure information for writing group members and peer reviewers is listed in Appendixes 1 and 2.

These pediatric guidelines are based on the extensive evidence evaluation performed in conjunction with the International Liaison Committee on Resuscitation (ILCOR) and affiliated ILCOR member councils. Three different types of evidence reviews (systematic reviews, scoping reviews, and evidence updates) were used in the 2025 process. This process is described more fully in “Part 2: Evidence Evaluation and Guidelines Development.”¹⁸

The writing group reviewed all relevant and current AHA guidelines for CPR and ECC and all relevant ILCOR consensus on CPR and ECC science with treatment recommendations from 2020, 2022, 2023, and 2024.^19-22 Evidence and recommendations were reviewed to determine if current guidelines should be reaffirmed, revised, or retired or if new recommendations were needed. The writing group then drafted, reviewed, and approved recommendations, assigning a class of recommendation (COR; ie, strength) and level of evidence (LOE; ie, quality, certainty) to each. Criteria for each COR and LOE are described in Table 1.

The 2025 Guidelines are organized in discrete modules of information on specific topics or management issues.²³ Each modular knowledge chunk includes a table of recommendations using standard AHA nomenclature of COR and LOE. Recommendations are presented in order of COR: most potential benefit (Class 1), followed by lesser certainty of benefit (Class 2), and finally no benefit or potential for harm (Class 3). Following the COR, recommendations are ordered by the certainty of supporting LOE: Level A (high-quality randomized controlled trials) to Level C-EO (expert opinion). This order does not reflect the order in which care should be provided.

A brief synopsis is provided to put the recommendations into context with important background information and overarching management or treatment concepts. Recommendation-specific supportive text clarifies the rationale and key study data supporting the recommendations. When appropriate, illustrations are included. Hyperlinked references are provided to facilitate quick access and review.

The writing group consists of AHA and AAP representatives who voted on and approved all guideline recommendations. The guideline was submitted for blinded peer review to 10 subject matter experts nominated by the AHA and AAP. Before their appointment, all peer reviewers were required to disclose relationships with industry and any other conflicts of interest, and all disclosures were reviewed by AHA staff. The guideline was also reviewed and approved for publication by the AHA Science Advisory and Coordinating Committee, the AHA Executive Committee, and the AAP Board of Directors. Comprehensive disclosure information for peer reviewers is listed in Appendixes 1 and 2.

These recommendations supersede the last full set of AHA recommendations for PALS²⁴ made in 2020 unless otherwise specified. For topics that did not undergo full evidence review by the 2025 writing group, recommendations, recommendation supportive text and references from the 2020 pediatric basic and advanced life support guidelines²⁴ were not updated and were carried over and remain as the current guidelines for 2025. These topics are noted within the synopsis of their respective sections.

Historically, cardiac arrest care has largely focused on the management of the cardiac arrest itself, highlighting high-quality CPR, early defibrillation, and effective teamwork. However, there are aspects of prearrest and post–cardiac arrest care that are critical to improve outcomes. As pediatric IHCA survival rates have plateaued², the prevention of cardiac arrest becomes even more important. IHCA prevention includes early recognition and treatment of cardiac arrest such as children undergoing high-risk procedures (eg, infants undergoing cardiac surgery or catheterization procedures), children with high-risk diagnoses (eg, shock, pulmonary hypertension, or acute respiratory distress syndrome), and children with severely abnormal vital signs or other signs of deterioration. In the out-of-hospital environment, lay responder CPR training, sudden unexpected infant death prevention, safety initiatives (eg, bike helmet laws), and early access to emergency care are imperative. When OHCA occurs, early lay responder CPR is critical in improving outcomes.

Following resuscitation from cardiac arrest, management of the post–cardiac arrest syndrome (which may include brain dysfunction, myocardial dysfunction with low cardiac output, and ischemia/reperfusion injury) is important to avoid known contributors to secondary injury, such as hypotension.^8,9 There is recognition of growing differential use of the terms return of spontaneous circulation (ROSC) versus return of circulation (ROC) in the literature. For the purposes of these guidelines, after CPR, when ROC is achieved by native cardiac function, we use ROSC; when it has been achieved by either mechanical support (eg, extracorporeal membrane oxygenation, ECMO) or native cardiac function, we use ROC. Accurate neuroprognostication is important to guide caregiver discussions and decision-making.¹⁰ Finally, given the high risk of neurodevelopmental impairment in cardiac arrest survivors, early referral for rehabilitation assessment and intervention is key.

A single Cardiac Arrest Chain of Survival (Figure 1) that supports the paradigm of prevention and early recognition through recovery after cardiac arrest has now been standardized across infants, children, and adults (outside of neonatal care).

Administration of vasoactive and antiarrhythmic medications is a key component of PALS during CPR. Medications, such as epinephrine, have historically been administered via multiple routes, including endotracheal, intravenous (IV) and intraosseous (IO). Due to limited transalveolar drug absorption and limited pulmonary blood flow during CPR, IV and IO administration of medications are preferred over the endotracheal route.^1,2 Selecting the timeliest route to deliver medication based on clinical needs and available resources is appropriate.

Recommendation-Specific Supportive Text

1. Timely provision of vasoactive and antiarrhythmic medications during cardiac arrest is dependent on rapid securement of vascular access across all pediatric age ranges. IV or IO administration of epinephrine is preferred over endotracheal administration when possible.^1,2
2. IO access is a rapid, safe, and an effective initial vascular access route for pediatric cardiac arrest.^3-5 A recent systematic review did not identify any studies comparing IV versus IO for vascular access in pediatric cardiac arrest.⁶ Two retrospective nontraumatic OHCA registry studies showed that younger, more severely ill patients with lower rates of survival were more likely to have received IO access while older patients with higher rates of survival were more likely to have received IV access.^7,8 Health care professionals should consider resource availability, expertise, and timeliness (time to confirmed vascular access) when choosing vascular access routes for drug administration.

Administration of medications is a key component of PALS. Vasoactive agents, such as epinephrine, are used to increase coronary perfusion, primarily through alpha-1 receptor agonism. The optimal timing of initial dosing may differ in shockable cardiac arrest given the need to prioritize defibrillation. Antiarrhythmic agents are used to treat cardiac arrest with ventricular fibrillation (VF) and pulseless ventricular tachycardia (pVT). The existing literature does not support routine use of other adjunctive therapies such as sodium bicarbonate and calcium except in special circumstances.

Recommendation-Specific Supportive Text

1. A recent meta-analysis of retrospective observational studies showed that shorter time to administration of epinephrine during pediatric cardiac arrest with initial nonshockable rhythm was associated with favorable outcomes.^9-14
2. The optimal timing of administration of epinephrine in relation to defibrillation in pediatric cardiac arrest with shockable rhythm is unknown⁹ and may differ from nonshockable arrest given the need to prioritize defibrillation. Observational studies of pediatric OHCA have included patients with shockable rhythm, but none reported the relationship between initial epinephrine and timing of defibrillation.^11,13,15,16 Two observational studies in adults with shockable IHCA showed that early epinephrine, defined as before first defibrillation or within 2 minutes after first defibrillation, respectively, was associated with worse outcomes.^17,18 No studies have examined the effect of epinephrine for pediatric shockable arrest when early defibrillation is not possible.
3. The optimal interval between doses of epinephrine for pediatric cardiac arrest is unknown.⁹ Observational studies for conventional CPR have shown mixed results. One study demonstrated that average intervals greater than 5 minutes between doses were associated with increased survival compared to 1 to 5 minute intervals.¹⁹ A second study showed that average epinephrine intervals of 3 to less than 5 minutes were associated with increased survival.²⁰ Two additional studies found that intervals of less than 3 minutes between epinephrine doses were associated with better outcomes.^21,22 For pediatric patients undergoing extracorporeal cardiopulmonary resuscitation (ECPR), one observational study showed that after the first 10 minutes of CPR, survivors received fewer epinephrine doses during each subsequent 10-minute interval compared to nonsurvivors. There was no difference in survival between epinephrine dosing intervals of ≤5 minutes and greater than 5 minutes during minutes 10 to 30 of CPR on adjusted analysis.²³
4. In an observational study of pediatric IHCA with VF/pVT, administration of lidocaine was associated with higher rates of ROSC and 24-hour survival although neither lidocaine nor amiodarone significantly affected the odds of survival to hospital discharge.²⁴ A subsequent propensity-matched study of pediatric IHCA with initial VF/pVT demonstrated no difference in outcomes between patients who received lidocaine compared to amiodarone.²⁵
5. A meta-analysis of sodium bicarbonate administration during pediatric IHCA demonstrated lower rates of survival to hospital discharge in patients who received bicarbonate.^26-33 A subsequent propensity score-weighted cohort study found that sodium bicarbonate was associated with a lower rate of hospital survival, but no difference in rates of ROSC were observed.³⁴ However, these studies do not account for the fact that patients with longer durations of resuscitation, which are associated with lower survival rates, have greater exposure to medications like sodium bicarbonate. This phenomenon, known as resuscitation time bias, may falsely implicate the medication with poor survival if timing of drug administration is not accounted for in the analysis.³⁵ There are special circumstances in which bicarbonate is used, such as sodium channel blocker toxicity (eg, tricyclic antidepressants and cocaine)³⁶ and hyperkalemia. While a comprehensive list of special circumstances in which sodium bicarbonate may be beneficial has not been determined, it is important to note that sodium bicarbonate is routinely used in some of these circumstances (eg, hyperkalemia) which may significantly limit the emergence of new evidence due to lack of equipoise.³⁷
6. A recent meta-analysis examining the administration of calcium during pediatric IHCA demonstrated decreased survival to hospital discharge with calcium administration.^32,38-40 Two subsequent studies in pediatric IHCA and arrests managed in the emergency department showed similar associations.^41,42 However, as noted for sodium bicarbonate, none of the studies accounted for resuscitation time bias and calcium may be falsely implicated in the decreased survival rates in patients who undergo prolonged resuscitation.³⁵ There are special circumstances in which calcium administration is used, such as hypocalcemia, calcium channel blocker overdose, and hyperkalemia. It is important to note that calcium is routinely used in some of these circumstances (eg, hyperkalemia) which may significantly limit the emergence of new evidence due to lack of equipoise.³⁷

Medication dosing for children is based on weight, which is often difficult to obtain in an emergency setting. There are numerous approaches to estimating weight when an actual weight cannot be obtained.⁴³

This topic was last reviewed in the 2020 AHA Guidelines for CPR and ECC. These recommendations have not been re-reviewed and updated for this edition of the Guidelines.⁴⁴

Recommendation-Specific Supportive Text

1. There are many theoretical concerns about the use of actual body weight (especially in overweight or obese patients).^45-47 However, there are no data about the safety and efficacy of adjusting medication dosing in obese patients. Such adjustments could result in inaccurate dosing of medications.^48,49
2. Several studies suggest that inclusion of body habitus or anthropometric measurements further refines and improves weight estimations using length-based measures.⁴³ However, there are considerable variations in these methods, and the training required to employ these measures may not be practical in every context.
3. Cognitive aids such as the Broselow, PAWPER XL, and Mercy tapes can assist in the accurate approximation of body weight (described as being within 10% to 20% of measured total body weight). Several studies demonstrated high variability of weight estimates, with a tendency toward underestimation of total body weight yet closely approximating ideal body weight.^47,50,51

Hypoxic ischemic brain injury is a leading cause of morbidity and mortality following pediatric cardiac arrest. Hypoxic ischemic brain injury leads to free radical production, cellular apoptosis and necrosis.¹ Primary and secondary brain injury can result in cerebral ischemia or hyperemia, encephalopathy, seizures, and cerebral edema. Fever is common post–cardiac arrest and is associated with worse neurologic outcomes. Fever is also associated with increased central nervous system metabolic demands which may exacerbate primary brain injury after arrest.^2,3 Targeted temperature management (TTM) refers to actively maintaining a patient’s temperature within a closely prescribed temperature range while continuously monitoring central temperature. All forms of TTM actively prevent fever. Maintaining TTM between 32 °C and 34 °C attempts to treat systemic ischemic reperfusion injury.^4,5 A systematic review of pooled animal studies comparing TTM between 32 °C and 36 °C to control groups showed a strong effect of TTM on favorable neurologic outcome and reduced mortality.⁶

Recommendation-Specific Supportive Text

1. Accurate measurement of temperature is most reliably obtained at a core site (eg, rectal, esophageal, or bladder). Continuous temperature monitoring assesses temperature swings in patients at high risk for temperature instability. Continuous core temperature monitoring was used for the 5 days of TTM in Therapeutic Hypothermia After Pediatric Cardiac Arrest (THAPCA) trials.^7,8
2. Hyperthermia in the post–cardiac arrest period is common³ and is associated with decreased survival from both IHCA and OHCA. Avoiding hyperthermia offers a potential means to improve neurologic outcome.^9-11 Effective prevention of fever has been shown with active cooling with servo-regulated devices compared to air-cooled devices or passive cooling techniques¹² and with adoption of lower target temperature.¹³
3. The THAPCA randomized clinical trials of TTM (32 °C–34 °C for 48 hours followed by 3 days of TTM 36 °C–37.5 °C versus TTM 36 °C–37.5 °C for a total of 5 days) after IHCA or OHCA in children with coma following ROSC found no difference in 1-year survival with a favorable neurologic outcome.^7,8 Secondary analyses of the THAPCA trials showed no difference between temperature groups in any of the following subgroups: ECMO or ECPR, hypotension post-ROSC, open chest resuscitation, combined cohort of IHCA and OHCA, and acute kidney injury.^14-19 A Bayesian reanalysis of the THAPCA-OH trial found a high probability that hypothermia provides a modest benefit in neurobehavioral outcome and survival at 1 year.²⁰ A recent retrospective observational study found that children who received TTM 33 °C per clinician decision had higher health-related quality of life scores versus those who received TTM 36 °C evaluated at 3 years after the cardiac arrest.²¹

Hypotension (less than fifth percentile for age and sex) is common following ROC from cardiac arrest, occurring in 25% to 50% of infants and children.^15,22 In addition to the primary causes of cardiac arrest, hypotension can be related to myocardial dysfunction and systemic reperfusion, which is associated with inflammation and vasoplegia. Hypotension can exacerbate brain ischemia and myocardial ischemic injury, systemic hypoperfusion and resultant tissue hypoxia. Myocardial dysfunction, which is often present regardless of arrest etiology, typically occurs within hours of arrest and resolves in 48 to 72 hours.²³ Both the presence and severity of post arrest hypotension are associated with lower rates of survival to discharge.^15,22,24 whereas normal or high blood pressures are associated with survival to discharge.^14,25 When post–cardiac arrest hypotension is present, there is paucity of data regarding the association between interventions to treat hypotension and outcomes.

Recommendation-Specific Supportive Text

1 and 2. Blood pressure is often labile in the post–cardiac arrest period and recognition of hypotension is important. Continuous blood pressure monitoring allows for rapid identification of hypotension and can facilitate immediate treatment. Two observational studies associated systolic blood pressure below the fifth percentile for age in the first 12 hours following cardiac arrest with decreased rates of survival to discharge.^15,22 The severity and duration of systolic hypotension within the first 72 hours post–cardiac arrest care in the pediatric ICU (PICU) are associated with decreased survival to discharge,²⁴ whereas the combined absence of both post–cardiac arrest systolic hypotension and fever was associated with increased odds of survival to discharge.¹³ A secondary analysis of the ICU-Resuscitation trial of pediatric IHCA found higher rates of survival to hospital discharge as well as survival to hospital discharge with favorable neurologic outcome when blood pressure targets were above a threshold of systolic blood pressure greater than 10th percentile for age and diastolic blood pressure greater than 50th percentile for age during the first 6 hours post–cardiac arrest.²⁶ A multicenter study of IHCA and OHCA found that mean arterial blood pressure between the fifth and 74th percentiles for age was associated with favorable neurologic outcome.²⁷ An additional study showed that mean arterial blood pressure less than 10th percentile for age in the first 24 hours after cardiac arrest, quantified as burden of hypotension (duration and magnitude), was associated with unfavorable neurologic outcomes.²⁸

Post–cardiac arrest care is a critical component of the Chain of Survival. Monitoring of gas exchange with oxygen and ventilation titration is a key component of post–cardiac arrest care with the goal of preventing secondary end-organ injury. Although current recommendations are to administer 100% oxygen during cardiac arrest to maximize oxygenation during CPR as well as to minimize hypoxic-ischemic injury,²⁹ target ranges for post-ROC oxygenation are less certain. Animal studies^30,31 have shown associations with hyperoxia and reactive oxygen species, inflammation and brain injury, yet observational studies in infants and children post-ROC (whether spontaneous or achieved by mechanical means) have demonstrated mixed associations for hyperoxemia with survival and neurologic outcomes.^32-36 While most studies demonstrated adverse outcomes with hypoxemia, 1 recent observational study did not find an association with hypoxemia in the post–cardiac arrest period.³⁷ Associations of either duration or severity of hypoxemia or hyperoxemia with adverse outcomes in the post–cardiac arrest setting are unknown.

Extremes of arterial carbon dioxide levels lead to cerebral vasoconstriction when low (hypocapnia) and vasodilation when high (hypercapnia).³⁸ Carbon dioxide level fluctuations and their impact on cerebral blood flow in the pediatric post–cardiac arrest population remains poorly understood, although emerging evidence suggests mild hypercapnia may be associated with survival and favorable neurologic outcome.^35,37

Recommendation-Specific Supportive Text

1 and 2. Because an arterial oxyhemoglobin saturation of 100% may correspond to a Pao₂ between approximately 80 mm Hg and 500 mm Hg, it is reasonable to target an oxyhemoglobin saturation between 94% and 99%. Of note, pulse oximeters may overestimate oxygen saturation levels in patients with darker skin, which can lead to lower oxygen levels (hypoxemia) going undetected.^39,40 Six small observational studies of pediatric IHCA and OHCA did not show an association between hyperoxemia and outcome.^32,36,37,41 One larger observational study of pediatric IHCA and OHCA, as well as a secondary analysis of the ICU-Resuscitation trial, found associations between hyperoxemia after ROC and either decreased survival to PICU discharge³⁴ or decreased survival to hospital discharge with favorable neurologic outcome.³⁵

3. Four observational studies found an association between hypercapnia and increased mortality and worse neurologic outcomes.^33,35,37,41 Hypercapnia and hypocapnia impact cerebral blood flow. Targeting normocapnia (Paco₂ 35–45 mm Hg) or the patients’ baseline arterial partial pressure of carbon dioxide when chronically hypercapnic may prevent these perturbations.

Post–cardiac arrest brain injury remains a leading cause of morbidity and mortality in children because the brain has limited tolerance of ischemia, hyperemia, or edema. Post–cardiac arrest seizures occur in 5% to 30% of patients and can be nonconvulsive which can only be detected on electroencephalography.^44,45 Post–cardiac arrest status epilepticus is associated with worse outcomes including death and neurologic injury in survivors. There are no pediatric studies assessing the efficacy of antiseizure medications for either prophylaxis or treatment of seizures and their association with outcomes such as survival to hospital discharge or survival with favorable neurologic outcome.

Recommendation-Specific Supportive Text

1. Nonconvulsive seizures and nonconvulsive status epilepticus are common after pediatric cardiac arrest and are associated with worse outcomes.^44-47 The American Clinical Neurophysiology Society recommends continuous EEG monitoring for encephalopathic patients after pediatric cardiac arrest.⁴⁸ Nonconvulsive seizures and nonconvulsive status epilepticus cannot be detected without EEG monitoring.⁴⁸

2 and 3. There is insufficient evidence to determine whether treatment of convulsive or nonconvulsive seizures improves neurologic or functional outcomes after pediatric cardiac arrest. Both convulsive and nonconvulsive status epilepticus are associated with worse outcomes, but no study has evaluated treatment with antiseizure medications compared to no treatment.^44,45 A study comparing treatment to no treatment of rhythmic and periodic discharges following adult cardiac arrest found no difference in survival or neurologic outcomes.⁴⁹ The Neurocritical Care Society recommends treating status epilepticus with the goal of stopping convulsive and electrographic seizure activity.⁵⁰

Figure 3 shows the checklist for post–cardiac arrest care.

Hypoxic ischemic brain injury is the leading cause of death and disability after cardiac arrest.¹ Early and reliable neurological prognostication after resuscitation from pediatric cardiac arrest is essential to guide treatment, enable accurate counseling, and provide family support. In addition, accurate neurological prognostication is critical to avoid inappropriate withdrawal of life sustaining therapy in patients who may have a meaningful recovery while also avoiding potentially inappropriate life sustaining treatments. For further discussion of ethical considerations regarding prognostication and uncertainty refer to “Part 3: Ethics.”²

The definitions of favorable and unfavorable neurological outcome are complex and dynamic throughout the recovery period after cardiac arrest. Recovery progresses along a temporal continuum such that a child assessed as having a favorable or unfavorable outcome at hospital discharge may be assessed differently several months to years after discharge. Gross scoring systems such as the PCPC are limited in the granularity of assessment and not standardized across ages, such that the same PCPC of a 6-month-old and 10-year-old may not be comparable. Thus, when classifying patients into favorable versus unfavorable outcome categories and combining timepoints of assessment, it is important to understand that these dichotomous outcomes may not give a complete picture of the individual patient’s complex reality. However incomplete the portrait of outcomes, health care professionals must still provide guidance to families, thereby necessitating criteria to classify patient outcomes.

These 2025 recommendations are based on 2 ILCOR systematic reviews of potential prognostic modalities for neurological outcome; one for prediction of good (favorable) neurological outcome and one for prediction of poor (unfavorable) neurological outcome.^3,4 These reviews used the same data but analyzed them differently based on the outcome being assessed (good or poor). Definitions of good neurological outcome are not standardized and vary across studies, most often defined as a PCPC of 1 or 2; 1, 2, or 3; no change from baseline; or a Vineland Adaptive Behavioral Score of greater than 70.^5,6

Good neurological outcome was assessed as a threshold false positive rate (FPR) of less than 30% (ie, predicting a good neurological outcome but having a poor neurological outcome).^3,7 For poor neurological outcome, they used an FPR of less than 1% (ie, predicting a poor neurological outcome but having a good neurological outcome).⁴ Sensitivity (eg, probability of a good neurological outcome in a patient who has a positive finding) was assessed for all studies but was not the primary assessment of predictive accuracy. An FPR of less than 1% was selected for predictors of poor neurological outcome to minimize the risk of making recommendations for limitation of care in patients who would have a good neurological outcome. Adult studies have used FPR thresholds to predict poor neurological outcome between less than 0% and 5%.^8,9 Standard thresholds to predict outcome from pediatric cardiac arrest have not been established.

The systematic review for predicting good (favorable) neurological outcome-assessed associations of positive test findings (eg, bilateral reactive pupils) with good neurological outcome, whereas the systematic review for poor neurological outcome-assessed associations of negative findings (eg, bilateral unreactive pupils) with poor neurological outcome. It cannot be presumed that positive findings for favorable outcome (ie, presence of reactive pupils with good neurological outcome) will have the same predictive value as negative findings for unfavorable neurological outcome (ie, absence of reactive pupils with poor neurological outcome). For the purposes of this guideline, we have chosen the terms favorable (good) and unfavorable (poor).

Health care professionals use various assessments to guide neurological prognostication in the post–cardiac arrest period including neurological examination, biomarkers, EEG, and neurological imaging modalities (eg, brain computed tomography and magnetic resonance imaging). Studies reporting the predictive accuracy of individual assessments must be interpreted in the context of timing following ROC, outcome assessed (ie, survival versus survival with a good functional outcome), and unmeasured confounders (eg, sedation). Some modalities can predict favorable neurological outcome, unfavorable neurological outcome, or both. Each must be considered in conjunction with other modalities.

Recommendation-Specific Supportive Text

1. Numerous studies demonstrate associations between clinical examination findings, biomarkers, electrophysiology patterns, and neurological imaging findings with outcomes following pediatric cardiac arrest.^10-22 However, these studies are limited by their retrospective designs, lack of blinding, and unadjusted and unmeasured confounding. In addition, there are no established predictive accuracy thresholds (eg sensitivity, specificity, Area Under the Receiver Operating Curve) for individual or combined criteria. Thus, multiple modalities are needed for neurological prognostication.

Neurologic assessments routinely include brainstem reflexes such as pupillary response, cough and gag, and motor response to stimuli. The neurologic examination evolves in the days after cardiac arrest and a given finding (eg, pupillary reactivity) at an early timepoint may not have the same accuracy for prediction of outcome at a later timepoint. Furthermore, the predictive accuracy of an exam finding (eg, pupillary reflexes) at one timepoint for favorable neurological outcome may not have the opposite predictive accuracy when the exam finding is absent (eg, absence of pupillary reflexes for unfavorable neurological outcome). Exam findings in these studies are often obtained through retrospective chart review, thus patient condition and potential confounders may not be reported which may impact the accuracy of the assessment. Clinical examination may also be confounded by sedation administration. Therefore, caution must be used when interpreting these data and health care professionals must avoid the use of isolated exam findings to predict neurological outcome.

Recommendation-Specific Supportive Text

1. Three studies assessed the associations between cough or gag reflexes or evoked pain response and neurological outcome. The presence of cough and gag reflex at 24 hours predicted favorable neurological outcome with a low sensitivity of 40% for both and an FPR of 35% and 32% respectively.^22,23 The absence of cough and gag predicted an unfavorable neurological outcome with an FPR of 60% and a sensitivity of 65% to 69%.²³ The sensitivity for evoked pain response for favorable neurological outcome was 100%, but the FPR was as high as 67%.^22,24 The absence of an evoked pain response at 6 and 12 hours to predict unfavorable neurological outcome in 1 study had an FPR of 0% (0%–15%) with a sensitivity of 33%. ²² The FPR for these assessments did not meet predefined thresholds to predict favorable or unfavorable neurological outcome.
2. In a small study of 29 patients, any motor response at 48 and 72 hours after ROC had a sensitivity 80% to 100% and low FPR (23%–27%) for favorable neurological outcome.²⁵ In that same study, the absence of any motor response to predict unfavorable neurological outcome at less than 1 hour, 48 hours, and 72 hours after ROC had a sensitivity 61% to 73%. The FPR for unfavorable neurological outcome was 62% at less than 1 hour, 20% at 48 hours and 0% at 72 hours. These data are limited to 1 small single-center study and require further evaluation.²⁵
3. Three studies including 296 patients assessed the association of mental status measured by the total GCS score greater than 4, 7 or 8, or GCS motor score greater than 4 within 24 hours of ROC with neurological outcome at ICU or hospital discharge or 6-month follow-up. They had low sensitivity (less than 50%), and low FPR (less than 14%) for favorable neurological outcome,^26-28 but only 1 study assessed each timepoint and threshold and thus data were limited.
4. Three studies evaluating the presence of bilateral pupillary reflexes within 12 hours of ROC demonstrated a sensitivity (ie, probability of patients with a favorable neurological outcome having bilateral pupillary response) of at least 82% for favorable neurological outcome (PCPC 1–2 or 1–3) and FPRs (ie, bilaterally reactive pupils with an unfavorable neurological outcome) of 16% to 31%.^22,25,27 At timepoints greater than 24 hours from ROC, while sensitivity of bilateral pupillary reactivity was 75% to 100%, a high FPR of 68% was observed thus demonstrating its lack of reliability for prediction.^23-25,28-30 Therefore, findings of bilateral pupillary reactivity greater than 24 hours post–cardiac arrest are less reliable and pupillary response alone is inadequate for neurological prognostication at any timepoint.
5. Three studies evaluated the absence of pupillary reflexes to predict unfavorable neurological outcome at 48 and 72 hours after ROC with an FPR less than 1% but with wide confidence intervals (95% CI, 0%–40%) and low sensitivity of 12% to 46%.^24,25,31 These FPRs met the prespecified FPR threshold and are moderately reliable, but should only be used in conjunction with other predictors.
6. Six of 7 studies that assessed the absence of bilateral pupillary reflexes to predict unfavorable neurological outcome between less than 1 hour and 24 hours after ROC had an FPR greater than 10% up to 60% with sensitivities of 33% to 93%.^23,25,27-30 These studies show that the absence of pupillary reflex within 24 hours after ROC are not accurate to predict outcome.
7. Three studies including 296 patients assessed the association of mental status measured by the total GCS or GCS motor score within 24 hours of ROC with neurological outcome at ICU or hospital discharge or 6 month follow-up.^26-28 A GCS motor score of less than 4 at 1 hour after ROC to predict unfavorable neurological outcome had high FPRs of 83% and 50%, respectively, with high sensitivities of 93% to 94%.²⁷ A total GCS score of less than 4 at resuscitation or within 1 hour of ROC predicted unfavorable neurological outcome with a high FPR of 70% and a high sensitivity of 86%.²⁶ A total GCS score less than 7 to predict unfavorable neurological outcome had a high FPR of 69% and high sensitivity of 92%.²⁸ These data demonstrate that these GCS data are not accurate to predict outcome.

Serum biomarkers are blood-based tests that measure levels of proteins that are found in the central nervous system or measure inflammation or systemic ischemic reperfusion. Central nervous system proteins are released across the blood brain barrier when the brain is injured: Neuron-specific enolase is released from injured neurons, glial fibrillary acidic protein from injured glia, neurofilament light from injured axons, and S100B from injured astrocytes. Hypoxic ischemic brain injury and reperfusion injury may cause injury to these cells differently and release of these proteins may occur and peak at different times after injury. Studies of these neuronal biomarkers are limited by different assessment platforms, different baseline values across platforms and extracerebral sources for some proteins.

Commonly measured blood-based markers of inflammation or systemic ischemic reperfusion are pH and lactate levels. Lactate levels may reflect the severity of systemic hypoxia and ischemia before, during, or after cardiac arrest. pH is impacted by both systemic acidosis, of which lactate acidosis may be a major contributor, as well as respiratory acidosis, which may be due to inadequate ventilation. These systemic markers change over time after arrest. pH can be modified by titration of the ventilator to correct respiratory acidosis as well as administration of sodium bicarbonate to treat metabolic acidosis. Lactate clearance refers to the decrease in lactate over time after cardiac arrest; earlier clearance is associated with decreased mortality in adults.³²

Recommendation-Specific Supportive Text

1. Serum lactate less than 2 mmol/L within 12 hours of ROC had low sensitivity of 16% to 28% for favorable neurological outcome, but a very low FPR ranging from 4% to 7%.^33-35 Lactate thresholds of less than 2 mmol/L at 24 and 48 hours and less than 5 mmol/L at 1 and 24 hours had high FPRs of 17% to 68% and were moderately sensitive (61%–89%).^33,36 Thus only timepoints before 12 hours should be considered to predict outcomes in conjunction with other findings.
2. Neuronal biomarkers S100B, neuron-specific enolase, myelin basic protein , neurofilament light, ubiquitin C-terminal hydrolase-L1, glial fibrillary acidic protein and tau have been assessed at timepoints ranging from less than 1 hour through 96 hours after ROC.^20,29,37,38 Various thresholds were assessed for biomarkers to predict favorable neurological outcome at different timepoints resulting in a very wide ranges of FPR (0% to 96%) and sensitivities (5% to 100%). Three studies of various thresholds for S100B had an FPR of 0% with sensitivities of 29% to 38% for unfavorable neurological outcome and for neuron-specific enolase had FPR of 0% with sensitivities of 19% to 26%.^20,34,45 While these FPRs for favorable outcome were low, there were no consistent thresholds evaluated for these markers. Further study is needed to validate specific biomarker thresholds and discrete timepoints.
3. Lactate was assessed in 6 studies.^{20,34,35,39-41} Two studies had an FPR of less than 1% for unfavorable neurological outcome, 1 with a sensitivity of 11% with a threshold greater than 28.8 mmol/L 1 hour after ROC and 1 with a lactate clearance less than 2 mmol/L by 48 hours post-ROC.^20,39 All other studies had FPR ranging from 11% to 83% for unfavorable neurological outcome.^33-35,41
4. pH level was assessed in 4 studies.^20,34,35,39 A pH threshold of greater than 7.0 at 1 hour and 6 to 12 hours post–cardiac arrest had high sensitivities of 71% to 96%, but high FPRs ranging from 45% to 97% for favorable neurological outcome. A pH threshold of greater than 7.3 at 1 hour and 24 hours post–cardiac arrest had sensitivities of 49% and 89% respectively but FPRs of 38% and 81% respectively for favorable neurological outcome.⁴⁰ For unfavorable neurological outcome, 3 studies had FPRs greater than 5%, but sensitivities of 3% to 14%.^34,35 pH is not accurate enough to predict favorable or unfavorable neurological outcome.

EEG is broadly used to monitor for subclinical seizures and to assess background states after brain injury. Patients who are encephalopathic after cardiac arrest may have subclinical seizures due to cortical injury which cannot be detected without EEG monitoring. Electroencephalography for neurological prognostication has promise as it directly assesses neurologic activity. However, studies are limited because most are single center, utilize nonstandardized terminology, lack clinician blinding to EEG data when caring for patients, and do not describe the impact of medications on EEG background. These systematic reviews assessed multiple EEG features including the presence and absence of normal and abnormal EEG features with favorable and unfavorable neurological outcomes.^5,7

Recommendation-Specific Supportive Text

1. EEG monitoring up to 72 hours after ROC identifies EEG background features and diagnoses electrographic seizures. In patients who are encephalopathic, clinical exams can be unreliable and seizures can be subclinical. Electroencephalography is the only way to identify subclinical seizures.^{21,22,30,42-46}
2. Ten studies with more than 560 patients evaluated the association of either continuous or routine (brief) EEG at either less than 1 hour; 6 to 12 hours; 24, 48,72 hours; or 4 to 6 days with favorable neurological outcome at PICU or hospital discharge or 6 months following cardiac arrest. Among the studies evaluating EEG performed at 6 to 12 and 24 hours post-ROC, 6 of 7 studies had FPRs less than 29% for favorable neurological outcome with sensitivities that ranged from 7% to 100%.^{23,30,42,44,45,47,48} Studies at 48 and 72 hours reported FPRs ranging from 0% to 50% for favorable neurological outcome, although 50% of the studied timepoints had FPRs less than 30%.^21,24,29,47
3. The presence of sleep architecture at 6 to 12 hours or sleep spindles at 24 hours had an FPR of 16% and 8% for favorable neurological outcome, with sensitivities of 57% and 80%, respectively. Sleep spindles or sleep II architecture should not be used as the sole assessment to prognosticate at any timepoint.^23,30
4. At 6 to 12 and 24 hours post-ROC, the presence of reactivity had an FPR of less than 27% with a sensitivity of 53% to 63% for favorable neurological outcome.^42,44 At 48 hours, 1 study had an FPR of 50%.⁴⁴ Thus EEG reactivity, based on these data, is accurate only at 6 to 24 hours, not 48 hours for favorable neurological outcome.
5. The presence of status epilepticus predicted unfavorable neurological outcomes with a low FPR of 0% at 24 to 72 hours after ROC in 3 studies,^21,24,48 but an FPR of 4% to 5% at 4 to 6 hours and 6 to 12 hours, all with low sensitivity.^42,45 Thus, the presence of status epilepticus at 24 to 72 hours is moderately predictive for unfavorable neurological outcome. The presence of burst suppression, burst attenuation or generalized periodic epileptiform discharges within 24 hours of ROC had an FPR that ranged from 0% to 19% and sensitivity of 9% to 30% for unfavorable neurological outcome.^23,44,45,49 From 48 to 72 hours after ROC, 3 studies had FPRs less than 1% (95% CI, upper limit range 16%–54%) with sensitivities of 0% to 67% for unfavorable neurological outcome.^21,22,24 The prediction of unfavorable neurological outcome was moderately reliable from 24 to 72 hours.
6. Two studies of 61 patients, of which 8 had myoclonic status epilepticus, had an FPR of 0% (95% CI, 0%–34%) and sensitivity of 17% to 21% at PICU/hospital discharge.^22,44 The presence of attenuated, isoelectric, or flat EEG before 24 hours in 5 studies had FPRs ranging from 6% to 95%,^{23,42,45,47,49} except for 2 studies that had and FPR of 0% with upper limit confidence intervals of 4% to 31%.^30,50 In 6 studies from 48 hours to 7 days, 2 had FPRs less than 1% with 95% CI up to 34% to 52%^22,44; while other studies had FPRs that ranged from 3% to 71%.^21,29,47,51 There are conflicted data and further study is needed.
7. Of 10 studies, the presence of seizures between 4 to 6 hours and 24 hours post-ROC had an FPR of 0% to 20%, of which only 1 had an FPR of less than 1% with wide confidence intervals^44,45,50 and a sensitivity of 2% to 38% for predicting unfavorable neurological outcome.^{27,30,34,42,44,45,48-50} At 48 hours and onward only 2 studies reported an FPR for predicting unfavorable outcome of less than 1%^41,44; others had FPRs up to 58%. The presence of seizures is accurate to predict unfavorable neurologic outcome at any of these timepoints. The absence of a normal/continuous EEG background patterns (defined as normal, continuous, and reactive; continuous and unreactive; and nearly continuous) by ACNS definitions⁵² had variability and high FPRs ranging from 0% to 90%^{21-24,29,30,42,44,45,47-50}; only 2 studies had an FPR less than 1%.^48,51 The absence of a normal/continuous EEG background pattern is not accurate to predict unfavorable neurological outcome. The absence of EEG reactivity had an FPR of 0% to 93% and sensitivity of 36% to 100%^21,42,44; absence of sleep II architecture had an FPR 2 of 0% to 43% and sensitivity of 84% to 92%^23,30,42;and absence of variability on EEG^42,44 had an FPR of 0% to 80% and sensitivity of 21% to 82% for unfavorable neurological outcome prediction. These were unreliable tests for unfavorable outcome prediction.
8. Absence of attenuated, isoelectric or flat EEG to predict favorable neurological outcome was assessed in 10 studies.^{21-23,29,30,42,44,45,47,51} The FPR was greater than 40% in most studies and the sensitivity was high in most studies at all timepoints (71%–100%).^{23,30,42,44,45,47,51} While the absence of burst suppression, burst attenuation, or generalized periodic epileptiform discharges had a sensitivity for favorable neurological outcome of greater than 81% at all timepoints, the FPR was greater than 67% at all timepoints.^{21-23,42,44,45} These EEG features should not be used to predict favorable neurological outcome.
9. The absence of seizures had FPRs of 58% to 100% and sensitivities that ranged from 50% to 100% for favorable neurological outcome at PICU or hospital discharge, or 6 or 12 months post–cardiac arrest.^{21,22,27,30,34,35,42,44,45,51} Absence of status epilepticus at 6 to 12 or 72 hours had a high sensitivity (greater than 96%) for favorable neurological outcome, but very high FPRs (75% to 91%).^21,24,42,45 The absence of myoclonus in the 48 hours post–cardiac arrest was 100% sensitive for favorable neurological outcome but had high FPRs (79% to 83%).^22,44 Thus these EEG features should not be used to predict favorable neurological outcome at any timepoint.

Early neuroimaging after cardiac arrest is useful to detect structural brain injury and potential etiologies of cardiac arrest that may be amenable to intervention such as evacuation of intracranial hemorrhage. MRI can assess the severity of cytotoxic injury through diffusion weighted imaging. Ischemic injury on diffusion weighted imaging peaks at days 3 to 7 after cardiac arrest and may have the appearance of normalizing, or “pseudonormalization,” 2 weeks after injury. Both computed tomography (CT) and MRI findings evolve over the days after cardiac arrest and must be interpreted in the context of other findings and based on time from injury. Studies assessing neuroimaging for the prediction of neurological outcome require careful interpretation as they are limited by retrospective designs where health care professionals have likely used imaging results to guide treatment decisions.

Recommendation-Specific Supportive Text

1. In 3 studies, absence of diffusion restriction or absence of any abnormality on MRI had a sensitivity of 42% to 88% with FPRs of 0% to 2% for favorable neurological outcome.^24,51,53 Apparent diffusion coefficient thresholds of greater than 600×10^–6 mm²/s in greater than 93% and greater than 650×10^–6mm²/s in greater than 89% of brain volume, at a median of 4 days after ROC, predicted favorable neurological outcome with a sensitivity of 100% and low FPR of 20%.⁵⁴ While the absence of specific regional abnormalities on MRIs were highly sensitive for favorable neurological outcome, they had very high FPRs.^31,55
2. In 3 studies, the presence of high burden of ischemia defined as apparent diffusion coefficient threshold less than 650×10^–6 mm²/s in ≥10% of brain volume at a median of 4 days after ROC, predicted unfavorable neurological outcome with a sensitivity of 49% to 52% and an FPR of 0% to 6% (95% CI, 1%–21%)^49,54,56 but only 1 had an FPR less than 1% with a sensitivity of 80%.⁵⁴ The presence of diffusion restriction based in any brain region based on nonstandardized definitions at a median of 4 days and up to 14 days had high FPR (12%–58%) for unfavorable neurological outcome.^24,56 The studies found that the abnormalities in diffusion weighted imaging, T1-, and T2- weighted imaging in individual regions of the brain, at 4 to 6 days post-ROC, predicted unfavorable outcome with FPR of 0% to 10% but wide confidence intervals up to 50%.^24,31,55
3. One study of 78 children had an FPR 0% (CI 0%–12%) and sensitivity of 65% for the loss of gray-white matter differentiation on CT within 24 hours of arrest to predict unfavorable neurological outcome (PCPC greater than 3) at hospital discharge.¹⁶ Clinicians were not blinded to the CT results in any study.
4. In 2 studies, the absence of grey-white matter differentiation on CT within 24 hours from ROC had a sensitivity of 64% to 100% with an FPR between 35% and 75% for favorable neurological outcomes at hospital discharge.^16,57 Absence of a reversal sign and cistern or sulcal effacement had high FPRs that ranged from 14% to 80% sensitivities despite sensitivities that ranged from 93% to 100%.^16,57 Normal CT early after cardiac arrest is not accurate to predict favorable neurologic outcome.

Shock is the failure of oxygen delivery to meet tissue metabolic demands and can be life threatening. The most common type of pediatric shock is hypovolemic, including hemorrhagic shock, with distributive, cardiogenic, and obstructive shock occurring less frequently. Multiple types of shock can occur simultaneously, and early presentations can be subtle; health care professionals must be vigilant.

Shock progresses over a continuum of severity, from a compensated to a decompensated (hypotensive) state. Compensatory mechanisms include tachycardia and increased systemic vascular resistance to maintain cardiac output and end-organ perfusion. As compensatory mechanisms fail, hypotension and signs of inadequate end-organ perfusion develop, such as depressed mental status, decreased urine output, acidosis, and weak central pulses.

Sepsis is a life-threatening response to infection, where the body’s immune system overreacts, leading to widespread inflammation and organ dysfunction. Septic shock is characterized as sepsis with cardiovascular dysfunction including hypotension. Early recognition and prompt treatment (eg, antibiotics, fluids) are critical for improving outcomes in infants and children. Mortality from pediatric sepsis has declined, concurrent with implementation of guidelines emphasizing early antibiotic and fluid administration.¹ Controversies in septic shock management include volume and type of fluid administration, vasopressor timing and choice, and use of corticosteroids. Previous AHA guidelines² had considered large studies of patients with malaria, sickle cell anemia, and dengue shock syndrome; however, generalization of results from these studies is problematic.

This topic was last reviewed in the 2020 AHA Guidelines for CPR and ECC. These recommendations have not been updated for this edition of the Guidelines, with the exception of No. 6, which was reviewed and updated.³

Recommendation-Specific Supportive Text

1. Although fluids remain the mainstay initial therapy for infants and children in shock, especially in hypovolemic and septic shock, fluid overload can lead to increased morbidity.⁴ In 2 randomized trials of patients with septic shock, those who received higher fluid volumes⁵ or faster fluid resuscitation⁶ were more likely to develop clinically significant fluid overload characterized by increased rates of mechanical ventilation and worsening oxygenation.
2. In a systematic review, 12 relevant studies were identified, though 11 assessed colloid or crystalloid fluid resuscitation in patients with malaria, dengue shock syndrome, or “febrile illness” in sub-Saharan Africa.⁷ There was no clear benefit to crystalloid or colloid solutions as first-line fluid therapy in any of the identified studies.
3. One randomized controlled trial compared the use of balanced (lactated Ringer’s solution) to unbalanced (0.9% saline) crystalloid solutions as the initial resuscitation fluid and showed no difference in relevant clinical outcomes.⁸ A matched retrospective cohort study of pediatric patients with septic shock showed no difference in outcomes,⁹ though a propensity-matched database study showed an association with increased 72-hour mortality and vasoactive infusion days with unbalanced crystalloid fluid resuscitation.¹⁰
4. In a small, randomized controlled study, there were no significant differences in outcomes with the use of 20 mL/kg as the initial fluid bolus volume in septic shock, compared to 10 mL/kg; however, the study was limited by a small sample size.⁵
5. Two pediatric randomized controlled trials comparing escalating doses of dopamine or epinephrine demonstrated improvement in timing of resolution of septic shock¹¹ and 28-day mortality¹² with the use of epinephrine over dopamine. Both studies were conducted in resource-limited settings, and the doses of inotropes used may not have been directly comparable, limiting conclusions from the studies. Medications that increase systemic vascular resistance, such as norepinephrine, may also be a reasonable initial vasopressor therapy in septic shock patients.^1,13-15 International sepsis guidelines recommend the choice of medications to be guided by patient physiology and clinician preferences.¹
6. No studies support deviations from standard life-support algorithms to improve outcomes in patients with sepsis-associated cardiac arrest. Sepsis-associated cardiac arrest is associated with worse outcomes than other causes of cardiac arrest.¹⁶
7. A meta-analysis¹⁷ showed no change in survival with corticosteroid use in pediatric septic shock, though a subsequent randomized controlled trial suggested a shorter time to reversal of shock with steroid use.¹⁸ Two observational studies^19,20 suggested there may be specific subpopulations, based on genomics, that would either benefit or experience harm from steroid administration, though these subpopulations are difficult to identify clinically. Patients at risk for adrenal insufficiency (eg, those on chronic steroids, patients with purpura fulminans) are more likely to benefit from steroid therapy.¹⁴
8. In situations when epinephrine or norepinephrine are not available, dopamine is a reasonable alternative initial vasoactive infusion in patients with fluid-refractory septic shock.^11,12 Patients with vasodilatory shock may require a higher dose of dopamine.¹³

Cardiogenic shock occurs when the heart is unable to pump blood efficiently, resulting in decreased cardiac output and inadequate oxygen delivery to tissues. While it is relatively rare in children compared to other forms of shock, it is associated with high rates of mortality.²¹ Causes of cardiogenic shock include congenital heart disease, myocarditis, cardiomyopathies, and arrythmias. Clinical manifestations vary by severity and underlying cause. Initial compensatory mechanisms include tachycardia, tachypnea, and increased systemic vascular resistance. As compensatory mechanisms fail, hypotension and signs of inadequate end-organ perfusion develop, such as depressed mental status, decreased urine output, lactic acidosis, and weak central pulses. Initial treatment is focused on improving oxygen delivery, minimizing oxygen consumption, and restoring cardiac function through administration of vasoactive medications. It is important to note that cardiogenic shock can occur with other forms of shock simultaneously. In its early stages, it can be difficult to diagnose, so a high index of suspicion is warranted. This topic was last reviewed in the 2020 AHA Guidelines for CPR and ECC. These recommendations have not been updated for this edition of the Guidelines.³

Recommendation-Specific Supportive Text

1 and 2. Cardiogenic shock in infants and children is uncommon and associated with high mortality rates. No studies were identified comparing outcomes between vasoactive medications. For patients with hypotension, vasoactive medications such as epinephrine may be more appropriate as an initial inotropic therapy. Because of the rarity and complexity of these presentations, expert consultation is recommended when managing infants and children in cardiogenic shock.

Resuscitation guidance for children with hemorrhagic shock is evolving, as crystalloid-then-blood paradigms are being challenged by resuscitation protocols using blood products early in resuscitation. However, the ideal resuscitation strategy for a given type of injury is often unknown.

This topic was last reviewed in the 2020 AHA Guidelines for CPR and ECC. These recommendations have not been updated for this edition of the Guidelines.³

Recommendation-Specific Supportive Text

1. There are no prospective pediatric data comparing the administration of early blood products versus early crystalloid for traumatic hemorrhagic shock. A scoping review identified 6 retrospective studies that compared patient outcomes with the total volume of crystalloid resuscitation received in the first 24 to 48 hours among children with hemorrhagic shock.^22-27 Four studies reported no differences in survival to 24 hours, survival at 30 days with good neurological outcome, or survival to discharge.^22,24-26,28 Large-volume resuscitation was associated with increased hospital/ICU length of stay in 5 of the 6 studies.^23-28 One study reported lower survival to hospital discharge among children who received more than 60 mL/kg crystalloid compared to lower volume groups.²⁷ Despite limited pediatric data, guidelines for adults from the Eastern Association for the Surgery of Trauma,²⁹ the American College of Surgeons, and the National Institute for Health and Care Excellence³⁰ suggest the early use of balanced ratios of packed red blood cells, fresh frozen plasma, and platelets for trauma-related hemorrhagic shock.²⁹ The American College of Surgeons, and the National Institute for Health and Care Excellence³⁰ suggest the early use of balanced ratios of packed red blood cells, fresh frozen plasma, and platelets for trauma-related hemorrhagic shock.

Cuffed and uncuffed ETTs have been studied in the perioperative period and up to 48 hours postintubation. These studies have shown that cuffed ETTs are safe in infants and children and are not associated with significant differences in postextubation airway complications.^1-6 The primary advantage of cuffed ETTs over uncuffed tubes is a reduction in the need for ETT changes, which reduces high-risk reintubations and interruptions in chest compressions during CPR that may occur during reintubation. Additional potential benefits include improved capnography accuracy and a lower risk of atelectasis.^2,5-13 Although several studies have identified that cuffed tubes may decrease airway trauma by decreasing tube exchanges, attention must be directed to selecting the correct tube size³ and cuff inflation pressure.¹⁴ ETT cuff pressures are dynamic during transport at altitude,¹⁵ and with increasing airway edema or pressure, make it necessary to monitor the cuff pressure to avoid damage to the airway mucosa.

Recommendation-Specific Supportive Text

1. A retrospective study of 155 neonates and children under 12 years of age undergoing cardiac surgery with cardiopulmonary bypass and intubated for at least 4 hours postoperatively found larger ETT size as the only significant factor associated with airway complications.³ A retrospective study of 2,953 children found that applying 25 cm H₂O of cuff pressure with a slight leak around the ETT resulted in no cases of clinically significant subglottic stenosis and the incidence of reintubation for stridor was less than 1%.¹
2. Three systematic reviews, 5 randomized controlled trials, and 3 retrospective reviews support the safety of cuffed ETTs in infants and children.^2-5,7-13 While these studies were almost entirely performed in the perioperative patient population, 2 studies included patients who remained intubated for 4 hours³ or over 12 hours postoperatively.⁵ An additional retrospective review evaluating the use of perioperative cuffed ETTs in 1162 infants weighing 2 to 5 kg demonstrated no increase in postoperative airway complications.⁶ The use of cuffed ETTs is associated with lower reintubation rates, more successful ventilation, and improved accuracy of capnography without increasing the risk of complications.^{2,5-9,11-13,16-18} Cuffed ETTs may also decrease the risk of aspiration and atelectasis.^5,19,20

Cricoid pressure during bag mask ventilation and during intubation has historically been utilized to minimize the risk of regurgitation of gastric contents into the airway while potentially aiding in visualization of laryngeal structures.^21,22 However, uniform application of cricoid pressure may impede bag mask ventilation or obscure visualization of the airway during laryngoscopy.^23-25 Cricoid pressure should be distinguished from external laryngeal manipulation where pressure is applied to the larger thyroid cartilage to facilitate airway visualization.²⁶

Recommendation-Specific Supportive Text

1. A retrospective study from a large international pediatric ICU intubation registry showed that cricoid pressure during bag-mask ventilation before tracheal intubation was not associated with lower rates of regurgitation.²⁶

2 and 3. A bronchoscopic study of 30 children revealed distortion of the airway with the application of cricoid pressure at forces below those recommended by usual intubation guidelines.²⁷ Acknowledging anatomic differences in the airway from infancy through adulthood, relevant adult data was considered, given limited pediatric data for the use of cricoid pressure in pediatric patients. A systematic review of cricoid pressure during rapid sequence intubation in adult emergency department patients found insufficient evidence to conclude if cricoid pressure affected the rate of first pass intubation success or the incidence of complications such as aspiration.²⁵ In a study of 60 adult patients undergoing elective anesthesia, cricoid pressure interfered with bag-mask ventilation in over 50% of patients.²⁴

Pediatric emergency intubation is a lifesaving, high-risk procedure that can be complicated by hypoxemia, bradycardia, hypotension, and cardiac arrest. Peri-intubation bradycardia may occur in response to (1) the underlying disease process that led to respiratory failure, (2) hypoxemia, (3) medications given for rapid sequence intubation, (4) the transition from negative to positive pressure ventilation, and (5) vagus nerve stimulation during laryngoscopy and endotracheal intubation, particularly in infants.^28,29 Previous guidelines have discouraged the routine administration of atropine before intubation, but state that atropine pretreatment may be reasonable to prevent vagally mediated bradycardia in select populations.³⁰ Observational studies demonstrate an association between atropine pretreatment and decreased bradycardia, higher heart rates without ventricular arrhythmias, and increased ICU survival.^31,32 However, there is little evidence to support the use of atropine as a pretreatment for the prevention of peri-intubation cardiopulmonary compromise or cardiac arrest.

Recommendation-Specific Supportive Text

1. Two prospective, observational studies conducted between 2007 and 2009 in patients less than 8 years of age intubated in the PICU or by critical care transport teams showed associations between atropine pretreatment and (1) decreased ICU mortality after propensity score adjustment and (2) significant acceleration in heart rate without provoking arrhythmias and a decrease in bradycardia.^31,32 Conversely, 4 retrospective, observational studies did not demonstrate an association between atropine use and occurrence of bradycardia during the peri-intubation period.^33-36
2. Historically, a minimum dose of atropine (0.1 mg) was recommended to prevent paradoxical bradycardia based on a 1971 study of 79 patients undergoing elective surgery, of which 5 participants were between 6 weeks and 3 years of age. These infants and children had a small and statistically insignificant decrease in heart rate after receiving doses of atropine ranging from 0.0018 to 0.0036 mg/kg, approximately 10% to 20% of the recommended dose for atropine pretreatment for intubation.³⁷ In 2015, in a prospective, observational study of 60 infants less than 15 kg undergoing elective surgery all of whom received less than 0.1 mg of atropine before intubation, no patient experienced paradoxical bradycardia or arrhythmias.³⁸

Confirmation of ETT placement in patients with a perfusing rhythm is not reliably achieved by auscultation of breath sounds, mist in the tube, or chest rise. Either colorimetric exhaled CO₂ detection or capnography is a more reliable method to assess initial ETT placement. In patients with decreased pulmonary blood flow from low cardiac output or cardiac arrest, exhaled CO₂ may not be reliable.

Recommendation-Specific Supportive Text

1. Although there are no randomized controlled trials linking use of exhaled CO₂ detection with clinical outcomes, the Fourth National Audit Project of the Royal College of Anesthetists and the Difficult Airway Society concluded that the failure to use capnography contributed to adverse events, including death and persistent neurological injury, in a study of adults and children in the ICU and emergency department settings.³⁹ One small, randomized study showed that capnography was faster than clinical assessment in premature newborns intubated in the delivery room.⁴⁰ There has been no difference in patient outcomes in pediatric studies comparing qualitative (colorimetric) and quantitative (capnography or numeric display) exhaled CO₂ detectors in the delivery room, ICU, or emergency department.^41-43
2. Adult literature suggests monitoring and correctly interpreting capnography in intubated patients may prevent adverse events.^39,44,45 This has been demonstrated in pediatric simulation studies, in which capnography improved health care professional recognition and response to ETT dislodgement.^46,47

Regular, narrow-complex tachyarrhythmias (QRS duration 0.09 second or less) are most commonly caused by re-entrant circuits, although other mechanisms such as ectopic atrial tachycardia or junctional tachycardia may occur. Regular, wide-complex tachyarrhythmias (WCT; QRS greater than 0.09 second) can have multiple mechanisms, including SVT with aberrant conduction, antidromic SVT or VT. SVT with aberrant conduction is the most common cause of WCT in infants and children.¹

The hemodynamic impact of SVT in the pediatric patient can be variable, with signs of cardiopulmonary compromise such as altered mental status, signs of shock or hypotension occurring in a minority of patients. In infants and children without cardiopulmonary compromise, re-entrant SVT can often be terminated with vagal maneuvers.^2-5 Adenosine remains the preferred medication initially unresponsive to vagal maneuvers.^3,6-14 For patients with hemodynamically stable SVT that recurs after initial successful treatment, expert consultation is important to diagnose etiology and customize treatment. The use of other IV antiarrhythmics including procainamide, amiodarone and sotalol have been studied with varying levels of success.^15-20 When cardioversion is provided, the administration of sedation before synchronized cardioversion can minimize the pain associated with the shock. However, extreme caution is warranted in the selection and dosage of sedatives in a patient with cardiopulmonary compromise.

Recommendation-Specific Supportive Text

1. Vagal maneuvers are noninvasive, have few adverse effects, and effectively terminate SVT in many cases; exact success rates for each type of maneuver (ie, ice water to face, postural modification) are unknown.³ Although improved success rates have been reported with a postural modification to the standard Valsalva maneuver in adults,² published pediatric experience with this technique is very limited. Upside-down positioning may be an additional form of a vagal maneuver that is effective in children.⁴
2. IV adenosine remains generally effective for terminating re-entrant SVT within the first 2 doses. In retrospective observational studies on the management of tachyarrhythmias, none directly compared adenosine to other drugs.^6,7,9-14 A Cochrane review that included mostly adult studies and 1 pediatric study showed that either adenosine or calcium channel blockers can be used for quick termination of SVT.⁸ Given the fast onset of action, short half-life and favorable side effect profile, adenosine is preferred for infants and children.²¹
3. For infants and children with SVT without cardiopulmonary compromise that is refractory to vagal maneuvers or adenosine, expert consultation can guide the choice of alternative second-line agents. The risk for proarrhythmic and life-threatening hemodynamic collapse increases with the administration of multiple antiarrhythmic agents. Multiple medications have been used as second-line agents for the management of adenosine-refractory SVT, including IV verapamil, β-blockers, amiodarone, procainamide, and sotalol.^{3,6,7,10,11,15,16,18-20,22,23} Few comparative studies exist.
4. Direct current transcutaneous synchronized cardioversion remains the treatment of choice for infants and children with SVT and cardiopulmonary compromise (characterized by altered mental status, signs of shock, or hypotension). However, these cases are uncommon, and there are few studies that report outcomes from cardioversion of SVT.^7,13,24 The administration of sedation before synchronized cardioversion can minimize the pain associated with the shock. However, extreme caution is warranted in the selection and dosage of sedatives in a patient with cardiopulmonary compromise.
5. Procainamide and amiodarone are moderately effective treatments for adenosine-resistant SVT.¹⁶ Procainamide may be slightly more efficacious; adverse effects are frequent with both therapies. IV sotalol was approved for the treatment of SVT in 2009. Since its approval, several single-center and 1 multicenter registry have shown that IV sotalol is effective in acute conversion of SVT, with a 60% to 100% termination rate of SVT and atrial tachyarrhythmias.^15,17,18-20 In these studies, IV sotalol was administered under the guidance of pediatric electrophysiologists in acute care settings and no significant adverse events were reported.

The occurrence of WCT (QRS duration greater than 0.9 second) with a pulse is rare in children and may originate from either the ventricles or the atria.²⁵ SVT with aberrant conduction is the most common cause of WCT in infants and children; however, other etiologies can include antidromic SVT or VT.¹ For patients with WCT and no cardiopulmonary compromise, expert consultation is important to diagnose etiology and customize treatment. The administration of sedation before synchronized cardioversion can minimize the pain associated with the shock. However, extreme caution is warranted in the selection and dosage of sedatives in a patient with cardiopulmonary compromise.

Recommendation-Specific Supportive Text

1. Both pediatric and adult studies have identified potential populations at risk of proarrhythmic complications from antiarrhythmic therapies, including patients with underlying cardiomyopathies, long-QT syndrome, Brugada syndrome, and Wolff-Parkinson-White syndrome.^26-30 Given this potential risk, expert consultation can help provide a tailored antiarrhythmic choice to minimize the chances of arrhythmia.
2. Given that the most common cause of WCT in infants and children is SVT with aberrancy,¹ the use of adenosine can be either therapeutic or diagnostic. If the cause of the WCT is SVT with aberrancy, administration of adenosine is likely to terminate the arrhythmia. If the cause of the WCT is VT, the administration of adenosine can be diagnostic if it demonstrates ventriculo-atrial dissociation. Adenosine should only be administered in WCT if the tachycardia is monomorphic and regular. If WCT is irregular, it can represent pre-excited atrial fibrillation, and the administration of adenosine can lead to VF. When possible, the administration of adenosine should be done with a running rhythm strip or multilead electrocardiogram to help with interpretation of the adenosine response.^1,31
3. Cardiopulmonary compromise is a key factor in determining the use of electrical therapy over primary pharmacologic management in children with WCT and a pulse. In patients with cardiopulmonary compromise, electrical direct current synchronized cardioversion should be provided urgently, regardless of suspected atrial or ventricular origin. There is insufficient evidence describing the incidence of WCT without cardiopulmonary compromise, and there is no support for or against the use of specific antiarrhythmic drugs in the management of children with wide-complex tachycardia with a pulse.
   Figure 7 shows the algorithm for pediatric tachyarrhythmia with a pulse.

Pulmonary hypertensive crises are acute rapid increases in pulmonary artery pressure that can lead to systemic hypotension, myocardial ischemia, cardiac arrest, and death. Optimization of physiologic conditions and avoidance of potential triggers of pulmonary hypertensive crises can potentially avoid these events. Acidosis and hypoxemia are both potent pulmonary vasoconstrictors.^1-4 Inadequate sedation or uncontrolled pain can lead to increased sympathetic activity, which raises pulmonary vascular resistance and can exacerbate pulmonary hypertension.^1,5-7 Hypovolemia in children with pulmonary hypertension decreases preload and potentially compromises cardiac output, worsening the condition and increasing hypoxia risk. Conversely, fluid overload causes venous congestion and impairs heart function in the setting of right ventricular dysfunction.⁸ Anemia can be problematic in patients with pulmonary hypertension due to reduced oxygen-carrying capacity.⁸

Recommendation-Specific Supportive Text

1. One review, 1 consensus statement, and 1 scientific statement highlight the importance of identifying and treating environmental factors that contribute to pediatric pulmonary hypertensive crises, such as hypoxia and acidosis, especially in high-risk populations with preexisting idiopathic pulmonary arterial hypertension or with congenital heart disease and pulmonary arterial hypertension.^7,9,10 Two physiologic reviews, 1 randomized clinical trial, and 2 retrospective observational studies demonstrated that hypercarbia, hypoxemia, acidosis, atelectasis, and ventilation-perfusion mismatch can lead to increases in pulmonary vascular resistance and, hence, elevation of pulmonary artery pressures in the immediate postoperative period.^1-4,11
2. Two observational studies in select high-risk postoperative cardiac patients found an attenuation in the stress response in those patients receiving fentanyl in the postoperative period.^5,6 Three studies indicate that triggering factors, such as pain and anxiety, can precipitate pulmonary hypertension crises both during the perioperative period and outside of it, particularly in the presence of acute lung injury, infection, or noncardiac interventions. Managing and preventing these aggravating factors with sedatives and analgesics can help prevent recurrent, more severe, and prolonged pulmonary hypertension crises.^7,10,12 Neuromuscular blockade can be used in addition to sedation and analgesia in critically ill, mechanically ventilated children with pulmonary hypertension.^7,12
3. Hypovolemia and hypervolemia can be deleterious to the cardiac function. Right ventricular decompensation may result from conditions leading to increases in cardiac demand or increases in ventricular afterload. Two consensus statements on the treatment of pediatric pulmonary hypertension emphasize the importance of managing triggering factors such as volume overload, dehydration, and anemia in critically ill children with acute pulmonary hypertension.^8,10

During a pulmonary hypertensive crisis, elevated pulmonary vascular resistance reduces pulmonary blood flow, impairing left-heart (or single ventricle) preload and ultimately leading to a fall in cardiac output. Hypoxia and hypercarbia can significantly increase pulmonary vascular resistance and trigger pulmonary hypertension crises.^1,8,10 Supplemental oxygen helps maintain oxygen levels and lowers pulmonary artery pressures, as does inducing alkalosis. Pulmonary vasodilators, including inhaled nitric oxide, inhaled prostacyclin and IV prostacyclin analogs have been used to treat acute pulmonary hypertensive crises while IV and oral phosphodiesterase type-5 inhibitors, oral endothelin receptor antagonists and oral soluble guanylate cyclase stimulators are used as long-term therapies to lower pulmonary vascular resistance and prevent pulmonary hypertensive crises.^13-16 ECMO may stabilize infants with severe pulmonary hypertension, low cardiac output, or severe respiratory failure when other treatments fail,¹² or during high-risk cardiac interventions.¹⁷ Children with pulmonary hypertension who require ECMO have higher mortality rates compared to those without pulmonary hypertension.^18,19

Recommendation-Specific Supportive Text

1 and 2. Treatment with inhaled nitric oxide reduces the frequency of pulmonary hypertensive crises and shortens time to extubation.²⁰ In patients with atrioventricular septal defect repair and severe postoperative pulmonary hypertension, inhaled nitric oxide administration is associated with reduced mortality.^15,21 Inhaled prostacyclin transiently produces pulmonary vasodilation and improves oxygenation, but the alkalinity of the drug can irritate airways, and precise dosing can be complicated by drug loss in the nebulization circuit.^12,22 Two physiologic reviews and 1 randomized clinical trial demonstrated that hypercarbia, hypoxemia, acidosis, atelectasis, and ventilation-perfusion mismatch can lead to increases in pulmonary vascular resistance and elevation of pulmonary artery pressures in the immediate postoperative period after cardiac surgery.^2-4 One prospective study on the right ventricular response to hypercarbia following cardiac surgery showed that hypercarbia significantly increased pulmonary vascular resistance by 54% and mean pulmonary arterial pressure by 34%.²³ In addition, a prospective study in infants after cardiopulmonary bypass demonstrated that increasing the arterial pH by the administration of sodium bicarbonate resulted in significant reduction of pulmonary vascular resistance with both a decrease in mean pulmonary arterial pressure and a concomitant increase in cardiac index.²⁴

3. Extracorporeal membrane oxygenation has been used in children with pulmonary vascular disease in refractory low cardiac output states and cardiac arrest.^25,26 Although outcomes remain poor in certain populations,²⁷ advances in technology of extracorporeal devices may allow for bridging to durable MCS or to transplantation.^17,28 Although patients with pulmonary hypertension who require ECMO have a high mortality rate, provision of ECMO can be lifesaving.^12,29,30

Among children with in-hospital cardiac arrest, pulmonary hypertension is a common preexisting condition, and several observational studies have found an association between pulmonary hypertension and lower cardiac arrest survival rates.^19,31-34 Despite inhaled nitric oxide showing efficacy in preclinical large animal models of cardiac arrest associated with pulmonary hypertension,^35,36 studies on optimal management for children with pulmonary hypertension-associated cardiac arrest are scarce, and the efficacy of specific intra-arrest therapies is unknown.

Recommendation-Specific Supportive Text

1. No intra-arrest therapies have been prospectively compared to standard resuscitation therapies for pulmonary hypertension-associated cardiac arrest. One observational study reported that 55% of children with pulmonary hypertension and in-hospital cardiac arrest were receiving inhaled nitric oxide at the time of arrest, but did not identify associations with survival.³² No studies have prospectively investigated pulmonary vasodilator therapy during pediatric cardiac arrest.

• Javier J. Lasa, MD, Co-Chair
• Gurpreet S. Dhillon, MD
• Jonathan P. Duff, MD, Med
• Jennifer Hayes, RN, MSN
• Beena D. Kamath-Rayne, MD, MPH
• Arielle Levy, MD, Med
• Melissa Mahgoub, PhD
• Ryan W. Morgan, MD, MTR
• Taylor McCormick, MD
• Joan S. Roberts, MD
• Catherine E. Ross, MD
• Stephen M. Schexnayder, MD
• Todd Sweberg, MD, MBA
• Santiago O. Valdés, MD
• Alexis A. Topjian, MD, MSCE, Co-Chair

Contributor: Rebecca N. Ichord, MD