Cardiopulmonary resuscitation in poorly resourced settings: better to pre-empt than to wait until it is too late

Abbreviations: APLS: advanced paediatric life support; CA: cardiac arrest; CPR: cardiopulmonary resuscitation; CPA: cardiopulmonary arrest; ETAT: emergency triage assessment and treatment; ETAT+: emergency triage assessment and treatment (includes first 48 h); ICP: intracranial pressure; LMIC: low- and middle-income countries; MET: medical emergency team; PALS: paediatric advanced life support; PEA: pulseless electrical activity; PEWS: paediatric early warning signs; PICU: paediatric intensive care unit; ROSC: return of spontaneous circulation; RRT: rapid response team; VF: ventricular fibrillation

Introduction

Cardiopulmonary resuscitation (CPR) is an emergency attempt to maintain or restore life by keeping vital organs (brain and heart) oxygenated until spontaneous respiration and circulation return.

Cardiopulmonary arrest (CPA) is often the end result of progressive respiratory or circulatory failure. The primary causes are very varied but, in children, cardiac arrest is often precipitated by respiratory failure, circulatory failure, sepsis or central nervous system infections. In adolescents, polytrauma can lead to an arrest from poor positioning when unconscious, head or chest injuries, or massive blood loss. Primary cardiac causes are less common than in adults.

CPAs are described as occurring outside or inside hospital, and inside hospital is divided into those that occur in intensive care or on general wards. These distinctions are made because access to immediate trained help differs between them and this is reflected in the outcome of cardiopulmonary resuscitation efforts.

Epidemiology

This editorial concerns children older than 1 month of age and therefore does not include neonatal resuscitation. Reports are difficult to compare because of variation in the upper age which ranges from 13 to 20 years. In high-income countries, out-of-hospital CPA occurs in 8–20 children/100,000 population/year [1,2]. In-hospital CPA occurs in approximately 0.77/1000 admissions, about two-thirds of whom are infants <1 year of age. Respiratory failure is the main reason, although cardiac disease and renal failure are not uncommon. It has been estimated that 1/1300 admissions to hospitals in the United States require CPR [3].

Cardiopulmonary arrests are more common in low- and middle-income countries (LMIC), probably because children are very ill and arrive late at hospital. Almost all CPR attempts in LMIC occur in hospital [4]. Reports from LMIC tend to come from single institutions and reflect their level of care and admission policies. In an emergency department of a Tanzanian teaching hospital, 18/165 (11%) infants <1 year of age required CPR [5]. A report from South India states that 27% (20/75) of children arriving at the hospital with agonal breathing received CPR in the emergency department [6]. In North India, 6.7% of all admissions required CPR [7]. Several studies report the prevalence and outcome from paediatric intensive care units (PICU). In a 6-month period in 18 centres in Turkey—15 PICUs and three emergency departments—24 (8%) children arrested prior to admission, 70% of CPR attempts were in PICUs and 17% in general medical wards [8].

Physiology of CPR

The aim of CPR is to maintain adequate, oxygenated, blood flow to vital organs. Blood flow is maintained in standard CPR by manual, regular compressions of the chest. During compression, intrathoracic pressure is raised and the heart is squeezed between the sternum and the spine. Pressure rises in both sides of the heart and blood is pushed forward out of the non-beating heart towards the brain, coronary arteries and other parts of the body. Heart valves prevent backflow. Coronary artery perfusion pressure is the difference between the pressure in the aorta and the right atrium but, as external cardiac compression squeezes both sides of the heart, the differential may not be large. Between compressions, the chest wall recoils and the heart fills with blood. It is important to give sufficient time for this to occur as filling is inefficient with passive chest recoil. It is important to understand that intracranial pressure (ICP) rises with increased intrathoracic pressure during chest compressions and also positive pressure ventilation reduces cerebral blood flow. Cerebral perfusion depends on gravity for venous flow back to the right side of the heart, for which the horizontal position is not ideal [9].

Ventilation is provided by a self-inflating bag and mask. Each inflation improves oxygenation and opens up the pulmonary vasculature; it also increases intrathoracic pressure which in turn increases right ventricle afterload. Thus, positive pressure lung inflation improves the stroke volume of the heart (Starling’s law). Carbon dioxide exchange depends on adequate ventilation and its role in intracranial perfusion during CPR, when autoregulation is lost, is unclear [9].

Delivery of cardiopulmonary resuscitation

Even with excellent CPR, the brain receives <10% of the blood flow generated in spontaneous cardiopulmonary activity, which means that it is important to be as efficient as possible. If cardiac compressions are too fast or too slow, or not deep enough, the blood flow generated is poor and the outcome is jeopardised [7]. If the rate is too fast, diastolic time is short, the heart fails to refill fully with blood and there is inadequate time for coronary artery perfusion. Full chest recoil is needed during diastole to effectively decrease ICP. Breaks from compressions and ventilation should not occur. CPR should not stop while pulses are felt, rhythms reviewed and drugs requested.

Over-ventilation leads to poor myocardial and brain perfusion pressures; hypoventilation reduces cerebral oxygenation and perfusion. In paediatrics, because the cause of CPA is often respiratory, five rescue breaths are given initially, followed by 15 cardiac compressions to two positive pressure ventilations. The rate of compressions is 100–120/min and depth should be 5 cm or a third of the chest depth [10]. At least two people are required in order to undertake this.

In children, the most common abnormal cardiac rhythm is pulseless electrical activity (PEA) or asystole. Ventricular fibrillation (VF) or pulseless ventricular tachycardia occurs in less than 10% [11]. If defibrillation is required, the defibrillator should be paediatric or able to attenuate to give 50–75 Joules. Automated external defibrillators which give a large adult shock (200 Joules) have been used successfully in children but should be used only if there is no alternative [10].

If spontaneous circulation is recovered, it is not uncommon for a child to have another CPA shortly afterwards and this second episode may be owing to PEA or an abnormal cardiac rhythm requiring defibrillation [12,13]. In many low-income settings without a cardiac monitor, the rhythm may not be known.

Drugs in CPR

Drugs play a secondary role in CPR; cardiac output and ventilation must first be maintained.

Fluids are essential for severe dehydration, and blood is needed urgently for massive blood loss; other causes of circulatory failure will need cautious fluids; if inotropic support is required, it is usually undertaken in the intensive care unit, if there is one. Adrenaline every 3–5 min is included in every algorithm for CPR; it is given following every two cycles of cardiopulmonary support. Adrenaline is given because it induces vasoconstriction, increases coronary and cerebral perfusion pressure and improves myocardial contractility. It is thought to stimulate spontaneous cardiac contractions and increase the intensity of VF, making successful defibrillation more likely [14,15]. The intravenous or intra-osseous dose of adrenaline is 10 µg/kg and higher doses should be avoided as they may worsen the outcome [14,15]. The Resuscitation Council (UK) recommends other drugs in special circumstances but none is to be used routinely [14]. Outside ICUs, other drugs are unlikely to be available in most low-income settings.

Utstein-style reporting on outcome of CPR

Utstein-style definitions and reporting templates are used widely for cardiac arrest to standardise reports. They have been revised since their introduction in the 1990s. In 2002, a task force of the International Liaison Committee on Resuscitation (ILCOR) simplified the reporting template and removed differences between in-hospital and out-of-hospital reports [16]. Outcome after cardiopulmonary resuscitation depends on early intervention, effective CPR and critical life support. Important outcomes are return of spontaneous circulation (ROSC), survival to hospital discharge and short- and long-term neurological sequelae [17].

Outcome of CPR

Globally, the outcome of CPR is poor. In well-resourced settings, survival from out-of-hospital arrests is, on average, 7% and for in-hospital is <30%, but up to 50% of survivors will have long-term neurological disability [18]. In a single PICU in India, only 14.5% (n = 25/170) survived to discharge. Survival was greater if CPR was for <10 min (32% discharged) vs >10 min (10% discharged). Only one of 10 cases requiring defibrillation survived [19]. Predictors of outcome are age (younger children do better), pre-arrest morbidity, where the arrest occurred, cardiac rhythm and duration of CPR [20]. Reported predictors of poor neurological outcome include duration of CPR >10 min (p = 0.013); out-of-hospital CA (p = 0.005); arterial pH <6.8 (p = 0.014); arterial lactate >2 mmol/L (p = 0.004); lack of pupil activity, absent brain stem reflexes and, on EEG, background suppression (all p < 0.001) [21,22]. In ICUs, if there has been a CA pre-admission, a re-arrest is six times more likely than an in-ICU arrest, and, if the arrest occurred within 4 h of admission, there is a 50% chance of re-arrest, both circumstances leading to a poor long-term outcome [23]. In a Malawian tertiary-care teaching hospital over a 6-month period in 2017, there were 135 attempts at CPR in children >1 month and <13 years of age. Almost all CPR efforts (90%) were on the general wards and 58% of children had an underlying diagnosis of malaria: only 6% had ROSC and none survived to discharge. This is a salutary reminder that seriously ill children arriving late at hospital in LMIC will not recover, especially if resuscitation attempts are not ideal [24].

When to start and when to stop

CPR should commence as soon as respiration or circulation is inadequate. It is better to initiate resuscitation before breathing and the heart actually stop. Gasping respirations are more effective at increasing intrathoracic negative pressure than positive pressure ventilation [9] and bag and mask ventilation can be synchronised with gasping attempts [25]. It was reported that 40.7% of 1853 children in whom CPR was commenced for bradycardia survived compared with 24.5% of 1489 in whom asystole/PEA had already occurred [26]. The sooner CPR commences the better.

Compressions should be as continuous as possible. Any pause to check for pulse or rhythm should be very brief. CPR must be effective and chest compressions should be to the depth of one-third of the chest diameter and at the rate of 100–120 beats/min; chest recoil must be full and free. Ventilation should allow for full lung inflation and then sufficient time given for expiration.

High-flow 100% oxygen is attached to the self-inflating bag which, for young infants, should have a pressure release valve so that the lungs are not over-inflated.

In a successful resuscitation, CPR can be stopped when the heart rate and volume support circulation and breathing is spontaneous and regular; in other words, ROSC has occurred. Careful vigilance must be maintained because of the possibility of a further arrest.

The decision as to when to stop if CPR is unsuccessful can be more difficult. Sepsis-related CA is known to have a poor outcome [27]. Sepsis leads to hypoxia and metabolic acidosis compounded by poor vascular tone and myocardial dysfunction which makes successful resuscitation difficult [28]. The required duration of CPR is strongly correlated with outcome; old and more recent reports suggest that attempts fail after 15–30 min of CPR [7,29–31]. Success is also less likely when more than one dose of adrenaline is required, or if other vaso-active drugs are required [7,30]. When more than one dose of adrenaline was required, it was found that survival was reduced from 77.8% to 20.7% [27]. Before withdrawing CPR, the leader of a team should spell out clearly to all involved what has been done and ask if anything has been missed; not all the team may have been privy to all that has been going on. Together the team needs to agree when CPR should be withdrawn. It is also important to consider what happens after resuscitation. If endotracheal intubation is undertaken and an ICU with mechanical ventilation is available, further difficult decisions may follow as to when to withdraw active treatment.

There are always some exceptions to rules and likelihoods and one is drug overdose. If a child has taken a toxic dose of, for example, an opiate, anxiolytic or anti-epileptic, every attempt should be made to continue resuscitation until the drug effect has passed. Some drugs have an antidote which will hasten recovery and reverse some of the toxic effects of drugs. Another exception is drowning in very cold water. This is unlikely in hot climates but, should such an accident occur, slow warming and continued resuscitation over time are appropriate. However, in a study of 160 Dutch children <16 years old with cardiac arrest and hypothermia after drowning during 1993–2012, 98 (61%) required >30 min of CPR and 87 (89%) died. The 11 survivors were all neurologically damaged whereas, of the 62 (39%) who required CPR for <30 min, 17 survived, 10 were neurologically normal, 5 had mild disabilities and 2 had moderate disabilities [32].

Post resuscitation

The problems do not end with successful resuscitation. Re-perfusion injury occurs when circulation is restored to tissues after a time of profound lack of oxygen and nutrients. Reoxygenation causes oxidative stress leading to inflammation which can induce mitochondrial and endothelial dysfunction and eventual cell death. Controlled hypothermia appears to have benefit but is rarely feasible; however, raised temperature must be avoided or treated aggressively [9,33].

Training for resuscitation

The response to an emergency requires to be pre-planned and organised. A team needs to practice together, to know and understand their roles so that there is very little confusion and tasks are undertaken quickly. This requires that there is an adequate number of trained staff, the correct equipment, drugs in the ‘right’ place and a method of communication to alert staff when they are needed. In most high-income settings, it is mandatory that all clinical staff have attended and passed a life support course such as the American Heart Association’s PALS or the UK’s APLS [34,35]. ETAT and ETAT+, courses which target care at district-level hospitals in low-income settings, do not include cardiac compressions beyond the neonatal period [36,37]. This is because the outcome of CPR is poor in settings where very sick children present late and there is an insufficient number of staff. Instead, emphasis is placed on identifying emergency and priority signs and trying to forestall a cardiac arrest.

Early interventions to prevent CPA

Paediatric early-warning scores (PEWS) are designed to alert health workers to possible clinical deterioration in the hope of averting it [38]. Signs and symptoms are given a numerical score according to their seriousness. An abnormal score should elicit an action such as increasing the frequency of monitoring or calling for assistance. This assumes that there is a rapid response to the request for help and appropriate action can be taken. Many different scores are in use; some are quite complicated with many indicators, but others are simple, based on vital signs. Comparison of the scores is difficult because they differ but when the sensitivity and specificity of the scores have been reported they range, respectively, from 64% to 71% and from 82% to 95% [39–42].

Though early-warning scores are used widely in well-resourced settings [43], in low-resourced settings it is assumed that scoring is too time-consuming and complicated for routine use [44]. However, a simple early-warning score was validated successfully in the paediatric department of the University Hospital of Rwanda in Kigali [45]. Sick children can be identified and care escalated before cardiac arrest occurs. A word of caution is that a multi-centre study of EWS in 2018 found no reduction in all-cause mortality compared with usual care [46]. This makes it all the more important that signs of impending deterioration are identified early [47].

Response teams are called rapid response teams (RRT) or medical emergency teams (MET) [48,49]. They have the advantage of being trained with a history of working together. Few hospitals in low-resource settings have sufficient trained staff to maintain an emergency team.

Monitoring is not undertaken well in most poorly resourced settings, yet it is the foundation on which EWS and subsequent clinical interventions depend. Emergency care would improve greatly if monitoring was done well, nurses and doctors were trained in emergency care and necessary equipment was at hand. This needs leadership and local champions [50]. Staff must feel empowered and listened to; it must be seen as adding value to their work [51].

What can be done to improve the outcome of CPR?

Here is a situation where prevention is much, much better than cure. A child with an invasive bacterial infection will become hypoxic, hypotensive and acidotic before cardiac arrest occurs. It is important to anticipate these problems, look for danger signs and prevent a worsening situation. Respiratory failure has a much better outcome than cardiopulmonary arrest. It is better to support respiration early with non-invasive ventilation than to await an arrest. The most sick children should be nursed in beds near the nursing station and they must be regularly monitored for vital signs; if a PEWS can be used, so much the better. A concerned health worker should have easy access to the opinion of a more senior member of staff who can assess the situation and make a rapid, informed decision. Emergency equipment and drugs should be readily available on a ward or in the outpatient department.

Mothers and guardians need to know the danger signs to look for when a child is ill and to bring them to hospital quickly. Children should be triaged on arrival and receive appropriate and timely treatment.

The basics must be got right: careful history-taking and examination, ongoing monitoring and early interventions. It is better to pre-empt trouble than to try to manage it.