Paediatric cardiac arrest is the utmost medical emergency in a child, and the survival chances and the subsequent quality of life depend on timely quality resuscitation.
In an ideal world, at least in hospitals, identifying the early signs of a serious illness should prevent the occurrence of cardiac arrest (CA), in children as in adults. Unfortunately, in many patients, cardiac arrest occurs outside of hospitals and little can be done to prevent it. Even in cases of in-hospital CA (IHCA), with significantly more resources available, the prevention is difficult.
Unsurprisingly, if the patient has preexistent comorbidities (chronic illnesses, malnutrition or metabolic abnormalities), or if hypoxia and shock are present prior to the CA, the likelihood of return of spontaneous circulation (ROSC) will be much lower than if CA occurs in a previously normal patient(1).
Professional adult and paediatric societies have responded over the years to the need of better outcomes of CA, with guidelines for its management that have been updated regularly (every five years) in the past two decades. These documents refer to basic life support and advanced life-support in adults and children, with specific recommendations in certain special circumstances of resuscitation, such as in adult, CA in pregnancy, opioid overdose, acute coronary syndromes, as well as neonatal resuscitation, paediatric basic and advanced life support. Changes have been made over the years, as new evidence has become available from experimental and human observational, retrospective or prospective studies and meta-analyses. The paediatric guidelines have been developed based on extrapolations from adult data, since there currently are only simulation studies and extremely few prospective human trials or small case-series of paediatric CPR, resuscitation techniques, or post-resuscitation care and, therefore, the quality of evidence is relatively poor.
Out-of-hospital cardiac arrest (OHCA)
When it comes to paediatric OHCA, the survival is generally lower than in IHCA, with a significant proportion managing to achieve the return of spontaneous circulation (25-50%), but few surviving to discharge (8-20%) and even less (under 16%) with a good neurological outcome(2,3). The factors that are associated with survival are witnessed arrest, first rhythm VF of VT and bystander CPR(4).
The overall survival is higher in cases that benefit from by-stander cardio-pulmonary resuscitation (CPR) – 13.2% versus 9.5%, and the same advantage was noted with respect to a good neurological outcome (10.3% versus 7.6%) compared with those who did not. The neurological outcomes are even better if a bystander witnesses the OHCA and promptly initiates CPR – 20.3% versus 6.5%. Furthermore, only CPR that combines ventilation with chest compressions appears to result in better neurological outcome than no CPR (12.9% versus 9.6%), with an odds ratio (OR) of 2.0 (compressions-only CPR has similar results to no CPR)(5).
If non-traumatic cases are initially cared for in a paediatric emergency department (PED) versus general ED (GED), the survival is better (33.8% versus 18.9%), but with similar outcomes in traumatic cases(6).
In paediatric OHCA, the survival to hospital discharge is twice higher in children and adolescents than in infants (infants: 3.3%; children: 9.1%; adolescents: 8.9%; adults: 4.5%) and higher in children who experience VT/VF than in those with asystole/PEA(7). These outcomes are expected to improve as more automated external defibrillators will be more widely available.
Three factors seem to be important for survival in infants after OHCA – immediate and short-term witnessing of the arrest, bystander rescue breathing, and an interval time of less than 18 minutes from arrest to hospital arrival. The infants who received rescue breaths were twice more likely to survive, which seems to indicate that in this age group the majority of CAs are of respiratory origin(8).
In-hospital cardiac arrest (IHCA)
Survival after in-hospital cardiac arrest has improved over the years (from 25.9% to 41% in a Spanish population(9), and to 43-45% in the USA(10,11)). Although the time to the initiation of CPR is usually considerably shorter than in OHCA, this does not seem to influence outcomes.
Preexistent haemato-oncological diseases, the duration of CPR, the need for vasoactive drugs, bicarbonate and volume are associated with in-hospital mortality(9).
There are few studies on the subject of the optimal duration of CPR in children – a model proposed by Nehme et al. suggested no chance of survival at or after a CPR duration of 42 minutes(12).
Causes of cardiac arrest
In adults, the most frequent cause of CA is cardiac (acute coronary syndromes or arrhythmias) or, less common, non-cardiac – respiratory (including drowning, carbon monoxide poisoning), trauma, drug overdoses, oncologic or neurologic (stroke).
In children, the respiratory causes are at the origin of most cases of CA, followed by trauma, infections, oncological causes, endocrinological and neurological causes and, less common, but not less dangerous, cardiac causes such as congenital heart disease, arrhythmia or coronary artery disease.
V. Nadkarni et al. conducted a large study on pulseless cardiac arrest in adults and children: the first documented rhythms were asystole (40% in children and 35% in adults), and pulseless electrical activity (PEA – 24% in children, 32% in adults). Kitamura et al. reported in children 76.8% with asystole and 18.4% with PEA(13). Pulseless ventricular tachycardia (VT) or VF were present in 14% in children and 23% in adults. Of the patients with asystole and PEA, more children survived to discharge than adults – 24% versus 11%(14). In a large prospective IHCA resuscitation registry, VF/VT were present in 27% of children at one time during resuscitation, with the incidence increasing with age – 21% of neonates and infants, 25% of 1-7-year-old children, and 37% of 8-17-year-old children. The survival to hospital discharge was much better when VF/VT was the initial rhythm than when VF/VT appeared later during resuscitation – 35% versus 11%, with similar results with respect to good neurological outcomes – 33% versus 8%(15). These outcomes may have been overestimated because the authors included paediatric cerebral performance (PCP) category 3 as a good outcome (in most studies a good outcome is considered for PCP 1 and 2 only).
Resuscitation sequences – ABC and CAB
The prompt initiation of CPR, in both OHCA and IHCA, contributes to better outcomes.
The ABC of cardio-pulmonary resuscitation refers to three steps: Airway – ensuring the patency of upper respiratory airway (includes positioning the head in order to facilitate oxygen intake and carbon dioxide elimination); Breathing – providing initial rescue breaths and later basic or advanced ventilatory support, and Circulation – providing circulatory support (chest compressions). Conventional CPR includes rescue breaths (RB) with chest compressions (CCs).
Over the years, given the fact that many CAs in children occur as a result of respiratory disease, the traditional resuscitation sequence has been ABC: Airway – Breathing – Circulation.
In neonatal resuscitation, conventional CPR (ventilations plus chest compressions) provides the necessary lung expansion and circulatory support in the ABC sequence. All current CPR guidelines (Neonatal Resuscitation Program – NRP; American Heart Association – AHA; European Resuscitation Council – ERC) agree that this is the optimal approach. At birth, because lungs are filled with foetal pulmonary fluid that needs to be cleared and replaced by air, ventilation is essential to ensure optimal oxygen delivery to tissues, especially to the heart and brain.
Similar agreement exists with respect to adult CPR – given the fact that most of CAs are generated by a cardiac cause, the circulatory support by CC is essential to ensure an adequate coronary and cerebral perfusion.
In experimental studies on adult animals with ventricular fibrillation, CCs with RBs were associated with lower myocardial blood flow, aortic diastolic pressure, coronary and cerebral perfusion pressures during the seconds spent on performing the RB, when no compressions were provided(16). Although coronary artery oxygen content was better when chest compressions were associated with rescue breaths, left ventricular blood flow was lower, in spite of similar myocardial oxygen delivery, and neurological outcomes were worse in this group(17). Ideally, if the two RBs recommended prior to the 15 or 30 CCs would be provided within the space of only four seconds, coronary and cerebral perfusion pressures would drop less, but such a short time is not feasible in real life – it generally takes longer, about 14-16 seconds, which is longer than the time spent on 15 CCs(16).
In such animal models, PaO2 and PaCO2 are maintained for 4-8 minutes of continuous CCs, without RBs(18). In CAs of respiratory origin, from the outset arterial hypoxemia and acidemia are present and RBs are essential to correct these abnormalities and ultimately to ensure survival, in combination with CCs.
In a study based on a simulation of adult CA model, the teams of two physicians required a significantly shorter time to check the pulse and initiate CCs if they used CAB, rather than ABC algorithm(19).
In a similar design, of adult CA with CPR simulation conducted by nurses and physical therapists, Sekiguchi et al. have demonstrated a shorter time to the initiation of CCs in CAB than in ABC sequence (15.4±3 seconds versus 36±4.1 seconds; p<0.001). Considerable time was spent on preparing and performing rescue breathing (15.9±3.8 seconds)(20). CAB sequence is easier to teach and is retained longer(21).
The overventilation during CPR has detrimental haemodynamic effects – increased intrathoracic pressure decreases venous return and subsequently the cardiac output(22,23).
CCs are associated with increases in arterial and pulmonary blood pressure, but also of intracranial pressure(24), which may be of concern in patients with traumatic brain injury, both adults and children. In an experimental study on pigs, this effect was limited when the patient was positioned with the head up at 30°(25).
In human adults, in OHCA, a single rescuer is less likely to provide optimal quality conventional CPR, with seconds being lost between transition from CC to RB and back, with lower rates and quality of CCs. Even when CPR is performed by two rescuers, the time spent on RBs is associated with a significant decrease of coronary and cerebral perfusion pressure(16,26). As a consequence, the concept of “cardio-cerebral resuscitation” was proposed by G.A. Ewy(16,27) and consists of providing cycles of 200 CCs (CC-only) for the first few minutes of CPR, in order to minimize interruptions in OHCA of cardiac origin in adults. Its application has significantly improved survival (from 20% to 37%) and the rates of neurologically intact survival (from 15% to 39%)(28,29).
Vaillancourt et al. used the notion of chest compression fractions (the proportion of CPR time spent providing CCs) to illustrate this concept in a study on OHCA in adults without ventricular fibrillation (VF): the shorter the interruptions in CC (CC fraction 71%), the more likely was the patient to achieve ROSC (24.2%), although survival to discharge remained disappointingly low (only 2%)(30). Survival after by-stander initiated CPR can be similar with both CC only and CC + RB sequences(31).
Therefore, current guidelines recommend the CAB sequence in adults: ERC recommends, since 2005, initial 30 CCs, followed by a 30:2 ratio between CCs and RBs in single trained rescuer CPR(32); AHA recommends continuous CCs – up to three cycles of 200 CCs, with delayed ventilation, after 2-3 minutes(33).
Lubrano et al. have shown that, in a simulation model of CA in children, the diagnosis of CA is achieved in a shorter time in the CAB sequence than the ABC one (17.48±2.19 seconds versus 41.67±4.95 seconds; p<0.05) and CCs are initiated faster (19.27±2.64 seconds versus 43.4±5.03 seconds; p<0.05); the ventilation was initiated faster in both cardiac and respiratory scenarios of CA in the CAB sequence (19.13±1.47 versus 22.66±3.07; p<0.05)(34).
In children (including newborns after birth), the optimal CPR sequence is controversial – in the last two decades, studies have raised the issue of a potentially detrimental delay in initiating chest compressions (CC) and its impact on immediate (ROSC) and long-term (neurological) outcomes.
In a large prospective study, CA of non-cardiac causes predominated in children (71%), whilst the cardiac ones were less common, but still considerably more than in previous studies (29%). Only 1.7% of infants (aged under 1 year) survived with good neurological outcome. In children, the survival with favourable neurological outcome was significantly higher when by-stander CPR was provided – 4.5% versus 1.9% in the cases in which CPR was not given. In non-cardiac CA, the neurological outcomes were better when conventional CPR (rescue breaths + CC) was performed by bystanders than with CC-only CPR (7.2% vesus 1.6%); in CA of cardiac origin, the results were similar with both techniques(13).
In another large paediatric sample Japanese study, Goto et al. stated that conventional bystander CPR was associated with greater likelihood of neurologically intact survival, compared to chest-compression-only CPR, irrespective of cardiac arrest etiology, but the overall one-month survival rate and the one-month favourable neurological outcome (CPC 1 and 2) were disappointingly low – only 10.5% and 3.5% respectively(35).
In a population of children with OHCA overwhelmingly of respiratory origin, ROSC was obtained in 29% and survival to hospital discharge in 20%. Early by-stander CPR and shorter times to emergency medical services intervention were associated to higher ROSC and survival to hospital discharge(36).
In cardiac causes of CA in children, the survival is similar with conventional CPR and chest compressions-only CPR(13). As a consequence, since 2010, AHA has been recommending the CAB sequence(37,38), whilst ERC continues to support the traditional ABC one, arguing that it is easier to continue a well-practiced technique(39).
For CA of cardiac origin in children, important factors for good neurological outcomes (20%) are: public AED use, initial ECG, rhythm, defibrillation, age and witnessed cardiac arrest, whilst for non-cardiac OHCA (6.4%) these factors are: bystander rescue breathing, witnessed cardiac arrest, and age(40).
A logical question that arises when it comes to paediatric cardiac arrest is: “What is the right sequence – ABC or CAB?” A subsequent one would be “Is there a single right approach to paediatric cardiac arrest?” Given the variety of situations in which cardiac arrest can occur in infants and children, it would be logical to have more than one approach. And, as it is unanimously accepted, respiratory causes are more often at the origin of cardiac arrest in this age group (in some studies >80%(41)), the ABC sequence would be preferable, as adequate ventilation and oxygenation are priorities. Nevertheless, cardiac causes do exist in a significant number of infants and more in children – up to 30% (CA has been often reported in adolescent or young apparently healthy athletes during exercise)(15,40). In such cases, reestablishing circulation is of paramount importance and thus a CAB sequence would be preferable, CC mobilizing the O2 contained in the lungs (functional residual capacity) and delivering it to the tissues, primarily to the brain and the heart.
In spite of experimental and simulation data favouring the CAB sequence, there are no human or animal data which demonstrate that either ABC or CAB are superior when it comes to outcomes.
For the time being, in the absence of compelling evidence, a practical and logical conclusion would be the use of the ABC sequence in cardiac arrest of respiratory origin and CAB in those of cardiac origin – as long as the cause can be rapidly identified, based on the patient’s status and the context in which cardiac arrest has occurred.
Conflict of interests: The author declares no conflict of interests.