Article Text
PDF
Abstract
Objective Provision of chest compressions (CCs) and/or medications in the delivery room is associated with poor outcomes. Based on the physiology of perinatal asphyxia, we aimed to provide an overview of current recommendations and explore potential determinants of effective neonatal cardiopulmonary resuscitation (CPR): balancing ventilations and CC, CC rate, depth, full chest recoil, CC technique and adrenaline.
Design A search in the databases MEDLINE (Ovid) and EMBASE until 10 April 2015.
Setting Delivery room.
Patients Asphyxiated newborn infants.
Interventions CCs.
Main outcome measures Haemodynamics, recovery and survival.
Results Current evidence is derived from mathematical models, manikin and animal studies, and small case series. No randomised clinical trials examining neonatal CC have been performed. There is no evidence to refute a CC to ventilation (C:V) ratio of 3:1. Raising the intrathoracic pressure, for example, by superimposing a sustained inflation on uninterrupted CC, and a CC rate >120/min may be beneficial. The optimal neonatal CC depth is unknown, but factors influencing depth and consistency include the C:V ratio. Incomplete chest wall recoil can cause less negative intrathoracic pressure between CC and reduced CPR effectiveness. CC should be performed with the two-thumb method over the lower third of the sternum. The optimal dose, route and timing of adrenaline administration remain to be determined.
Conclusions Successful CPR requires the delivery of high-quality CC, encompassing optimal (A) C:V ratio (B) rate, (C) depth, (D) chest recoil between CC, (E) technique and (F) adrenaline dosage. More animal studies with high translational value and randomised clinical trials are needed.
- Neonatology
- Resuscitation
Statistics from Altmetric.com
What is already known on this topic
The cause of cardiovascular collapse in most newborns is asphyxia.
Treatment recommendations derived from adult ventricular fibrillation cardiac arrest cannot be readily extrapolated to neonatal cardiopulmonary resuscitation.
However, no randomised trials of neonatal cardiopulmonary resuscitation have been performed.
What this study adds
Despite having been challenged in animal and manikin studies, there is no evidence to refute the currently recommended synchronised chest compression (CC) to ventilation ratio of 3:1.
Raising the intrathoracic pressure, for example, by superimposing a sustained inflation on uninterrupted CCs, may be beneficial.
A CC rate >120/min may be beneficial.
Background
The majority of newborn infants successfully make the fetal to neonatal transition without help.1 However, an estimated 10% of newborns need help to establish lung aeration, which remains the most critical step of neonatal resuscitation. Fortunately, only about 0.1% of term infants receive chest compression (CC) or medications in the delivery room (DR). However, a recent review of newborns that received prolonged CC and adrenaline but had no signs of life at 10 min following birth noted 83% mortality, with 93% of survivors suffering moderate-to-severe disability.2 The poor prognosis associated with CC with or without medications in the DR raises questions as to whether improved cardiopulmonary resuscitation (CPR) specifically tailored to the newborn could improve outcomes.
The inability to predict which newborns need CPR, and the infrequent occurrence of DR CPR limit the opportunities to perform rigorous clinical studies on neonatal CC. The American Academy of Pediatrics/American Heart Association Neonatal Resuscitation Program guidelines3 recognise that cardiovascular collapse in most newborns is caused by asphyxia. However, the guidelines often rely on data from adult or animal studies. Such data may not be wholly applicable to newborn infants because the most common cause of cardiovascular collapse in the adult is ventricular fibrillation, not asphyxia. Thus, further studies are needed to determine the optimal method for improving haemodynamics during neonatal resuscitation. The objective of this review was to provide an overview of the physiology of birth asphyxia and the current recommendations for neonatal CPR with an emphasis on CC, as well as a physiological and theoretical rationale for supporting or refuting current practice. We also present the results of studies of neonatal CPR and suggestions for future directions.
Methods
We searched the databases MEDLINE (Ovid) and EMBASE until 10 April 2015.
Eligibility criteria
The search was designed to identify studies investigating neonatal CPR including CC. Search concepts included ‘perinatal asphyxia’, ‘newborn’, ‘chest compression’ and ‘cardiopulmonary resuscitation’.
Results
Asphyxia at birth
Asphyxia, a condition of impaired gas exchange with simultaneous hypoxia and hypercapnia leading to a mixed metabolic and respiratory acidosis, is the most common reason that newborns fail to make successful transition.4 The focus of resuscitation after birth asphyxia is to restore oxygenation and perfusion of vital organs including the myocardium. CC should increase coronary perfusion pressure (CPP), the main determinant of coronary blood flow,5 and mechanically pump the blood through the body until the myocardium becomes sufficiently oxygenated to regain function.4
CC mechanics
The ‘cardiac pump theory’ postulates that by compressing the heart between the sternum and vertebral column CC directly eject blood from the heart into the circulation with each compression.6 In comparison, the ‘thoracic pump theory’ states that a phasic increase in intrathoracic pressure creates a pressure gradient between the arterial and venous compartments, that is, a driving force for antegrade blood flow.7 ,8 During CC, coronary blood flow occurs during diastole with CPP determined by the difference between the aortic and right atrial pressure.9 In adult animals and humans, uninterrupted CC and systemic vasoconstrictors such as adrenaline enhance diastolic blood flow during CPR. Studies in 2–3 month-old asphyxiated pigs10 as well as randomised trials of adult cardiac arrest11 showed that continuous CC without rescue breaths significantly improved return of spontaneous circulation (ROSC) and increased survival due to improved haemodynamics. Thus, for adults, bystander CPR should be performed without rescue breaths and in lone rescuer advanced CPR, CC should be uninterrupted for the first few minutes.12 In adult and paediatric advanced life support, continuous CC with asynchronous ventilations (CCaV) is recommended after a secure airway has been established.12 ,13 In contrast, neonatal resuscitation guidelines recommend a 3:1 compression:ventilation (C:V) ratio with a pause after every third CC to deliver one effective ventilation.14
Factors influencing the overall efficacy of CC during CPR
High-quality CPR encompasses several factors including (A) balancing CC and ventilation, (B) adequate CC rate, (C) CC depth, (D) full chest recoil between CC and (E) correct technique (eg, the two-thumb (TT) technique and optimal positioning of the thumbs on the chest) (figure 1). In addition, to optimise the CPP generated with CC, (F) adrenaline may be needed. However, neither of these factors has been extensively studied to optimise coronary and cerebral perfusion while providing adequate ventilation of an asphyxiated newborn.
Factors influencing successful cardiopulmonary resuscitation with restoration of myocardial function.
Factor A: balancing CCs and ventilation to achieve ROSC
A1: CC to ventilation ratio
The recommendation for a 3:1 C:V ratio is based on expert opinion and consensus rather than strong scientific evidence.15 Rationales for this ratio include the higher physiological heart rate of 120–160/min and breathing rates of 40–60/min in newborns compared with adults. Furthermore, as bradycardia or cardiac arrest in newborns is usually caused by hypoxia, providing ventilation has priority in neonatal CPR.14
In newborn piglets with asphyxia-induced cardiac arrest, combining CC with ventilations improved ROSC and 24 h neurological outcome compared with ventilations or CC alone.16 Solevåg et al17 investigated a 9:3 C:V ratio in asphyxiated piglets with cardiac arrest with the hypothesis that nine CCs would generate higher diastolic blood pressure (DBP) during CPR than only three CCs in a series. However, increasing the number of CCs in a row should not be at the expense of ventilation, hence the ratio of CC and ventilation was maintained at 3:1. The DBP during CC and time to ROSC were similar between 9:3 and 3:1 C:V CPR.17 Similarly, C:V ratios of 3:1 and 15:2 were compared using the same model.18 Although the 15:2 C:V ratio provided a higher number of CC/min, ROSC was similar between groups.18 These studies suggest that during neonatal CPR higher C:V ratios do not improve outcomes. The current recommendations are further supported by manikin studies showing higher ventilation rates19–21 and minute ventilation (MV)21 with 3:1 compared with higher C:V ratios.
The effectiveness of any given approach depends on mechanical and physiological effects, and feasibility. Pedagogical and practical aspects should be taken into account in guideline development. Rescuer preferences and compliance might be important determinants of CPR quality and clinical outcomes. Manikin studies reported that participants perceived the 3:1 C:V ratio as being more tiring than higher ratios,20 ,22 that rescuers preferred 10:2 and 15:2 C:V ratios over a 3:1 C:V ratio,20 and that participants did not deliver the intended number of ventilations during 3:1 C:V CPR.9 Foglia et al23 confirmed poor guideline compliance in simulated neonatal CPR with providers, irrespective of experience, performing CC at a significantly higher rate than the recommended 90 CC/min. This is further supported by a recent clinical observation that even experienced resuscitators do not always comply with the current algorithm of neonatal CPR.24
A2: synchronised versus asynchronous CCs with ventilation
An argument for synchronised CPR is the potential interference of non-synchronised CC with tidal volume (VT) delivery, hence impairment of oxygen delivery. A manikin study compared 3:1 C:V CPR with CCaV and reported a similar VT with significantly higher MV during CCaV.21 Schmölzer et al compared 3:1 C:V CPR with CCaV in a piglet model of neonatal asphyxia and reported similar VT and MV. However, CCaV resulted in a trend towards improved ROSC and higher survival.25 The combined results of these studies suggest that a similar oxygen delivery can be achieved with 3:1 C:V CPR or CCaV.
A3: alternative ways of combining CCs with ventilation
Raising the intrathoracic pressure can significantly improve carotid blood flow during CPR.7 ,11 Chandra et al provided ventilation at high airway pressure (80–135 cm H2O) while simultaneously performing CC in an animal model and demonstrated increased carotid flow without compromising oxygenation.7 Studies in preterm lambs demonstrated that a sustained inflation (SI) increased intrathoracic pressure without impeding blood flow.26 In the resuscitation of asphyxiated newborn piglets, Schmölzer et al27 achieved passive ventilation with CC, by superimposing uninterrupted CC with SI, which significantly improved haemodynamics, MV, ROSC and survival compared with a 3:1 C:V ratio.
Factor B: adequate CC rate
Current guidelines recommend a 3:1 C:V ratio with 90 CC and 30 breaths to achieve approximately 120 events/min.14 However, the optimal CC rate during neonatal CPR remains unclear. Since natural heart rates in neonates are 120–160 bpm, an increase in CC frequency in neonatal patients might have the potential to boost artificial cardiac output compared with the recommended CC frequencies, which are based largely on experimental work in larger animal. A recent mathematical study suggested that the most effective CC frequency depends upon body size and weight.28 Although this study suggested that CC rates for term infants should be around 180 CC/min and even higher for preterm infants to optimise systemic perfusion and survival, these CC rates are impossible to achieve with any technique.28
In the above-mentioned piglet study by Schmölzer et al,27 using a CC frequency of 120/min superimposed by SI significantly improved outcomes compared with 3:1 C:V CPR, and we speculate that uninterrupted CC at a higher rate achieved a higher CPP and a faster recovery of systemic and regional blood flow,27 in agreement with the mathematical model.
Factor C: optimal CC depth
In infants and neonates, CC depth should be approximately 33% of the anterior-posterior (AP) chest diameter, which is relatively greater than that recommended for adults (20% of AP diameter).14 Studies in adult animals and humans show a positive correlation between adequate CC depth and improved outcomes.8 Unfortunately, our own observations suggest that even experienced resuscitators do not achieve the required CC depth. Adequate CC depth is important to optimise cardiac output. However, overcompressing the chest might cause rib fractures, cardiac contusion and other thoracic injuries.29
The optimal CC depth has not been rigorously evaluated in neonates. A recent analysis of CT images of neonates predicted better left ventricular ejection fraction using a third AP CC depth compared with a fourth AP CC depth.30 Maher et al31 performed a retrospective review of six infants (2 weeks to 7.3 months old) with indwelling arterial blood pressure monitoring after cardiac surgery. CC was performed with the TT technique, and a half AP CC depth produced significantly higher systolic blood pressure but similar DBP compared with a third AP CC depth. Manikin studies compared the CC depth using C:V ratios of 3:1, 5:1 and 15:2 during 2-min simulated CPR.19 Participants had higher and more consistent CC depth during 3:1 C:V CPR, however the CC rate was lower compared with a 15:2 C:V ratio. In addition, the 2-min depth decay during CC was significantly higher during 5:1 and 15:2 compared with a 3:1 C:V ratio.19
Factor D: effect of leaning
CC circulates blood partly by creating negative intrathoracic pressure during chest re-expansion/decompression, thereby facilitating venous return through a ‘bellows’ action.32 CC leaning or incomplete recoil is the incomplete release of the downward force on the chest wall after a CC. CC leaning may prevent or reduce the negative intrathoracic pressure during decompression, thereby impeding blood flow and reducing CC effectiveness.33 Laboratory investigations have suggested that CC leaning decreases coronary and cerebral perfusion pressures lasting for a significant period of time after the cessation of leaning,33 and current guidelines recommend complete chest wall decompression during CPR.14 A small number of investigations have documented leaning during adult CPR.22 ,32 CC leaning was very common and exhibited a wide distribution, with most leaning within a subset of resuscitations. Leaning decreased with time during continuous CC, suggesting that either leaning is not caused by rescuer fatigue, or that it may be mitigated by automated feedback during CPR. Our own observations from manikin studies also suggest leaning during neonatal CPR,34 however this has not been systematically analysed.
Factor E: optimal CC technique
Based on three radiographic studies in 1986 suggesting that the infant heart was located beneath the lower third of the sternum,35–37 CC should be over the lower third of the sternum rather than the mid sternum.14
Neonatal CCs have traditionally been performed with the TT or the two-finger (TF) method.5 With the TT method the hands encircle the chest while the thumbs depress the sternum. The two thumbs can either be side-by-side or superimposed on each other (figure 2). With the TF method the index and middle fingers depress the sternum while the other hand provides support to the infant's back. The TT method has been shown to generate higher systolic38 and mean39 blood pressures compared with the TF method in a small number of newborn infants with indwelling catheters. In infant pigs, the TT method resulted in higher CPP40 and applied CC pressure41 compared with TF CC. In manikins and human infants, the TT method resulted in a more appropriate CC depth with less variance, improper finger placement and fatigue than the TF method.42–44 Thus, guidelines recommend the TT over the TF method.14 Two newer CC methods have been described: The thumb and index finger (TIF) method45 and the use of an adhesive glove device for active decompression during CC.46–48 In an infant manikin, the TIF method was superior to the TF method and equivalent to the TT method with respect to CC quality.45 More effective decompression,46 vital organ perfusion and haemodynamics47 ,48 might be achieved during paediatric CPR using an adhesive glove device. However, the evidence about the effectiveness of these methods in newborn infants is currently limited.
Different techniques for neonatal chest compressions. Showing from left to right: Two-thumb technique with thumbs superimposed on each other, two-thumb technique with thumbs side by side, two-finger technique, and thumb and index finger technique.
Factor F: adrenaline
The optimal dose49 and route50 of adrenaline in newborn resuscitation have been much debated. Also, whether adrenaline has a place in newborn CPR at all remains controversial49–51 as the scarce amount of evidence is conflicting. In post-transitional asphyxiated pigs, 14% of animals achieved ROSC without vasopressors,49 and administration of 10 µg/kg adrenaline prior to initiation of CC did not improve ROSC or cerebral circulation and oxygenation compared with placebo.51 In contrast, 10–15 µg/kg adrenaline was needed to achieve ROSC in a transitional lamb asphyxia-model.52 In agreement with this Solevåg et al17 ,18 showed that CC alone did not generate a sufficient DBP, a proxy for CPP, even when CC and ventilations were performed at a ratio of 9:317 or15:2.18 In a ventricular fibrillation pig model, Kleinman et al50 found that endotracheal adrenaline resulted in lower plasma adrenaline concentrations compared with administration intravenously or in the right atrium. However, ROSC rates were similar between groups. McNamara et al49 compared 10 µg/kg and 30 µg/kg of adrenaline in asphyxiated pigs and reported that animals receiving 10 µg/kg adrenaline were less likely to survive and more likely to need defibrillation. Interestingly, 0.4 U/kg vasopressin resulted in improved survival, lower postresuscitation troponin I, and less haemodynamic compromise after cardiac arrest compared with adrenaline.49
Future directions
Much of our knowledge about the physiology and mechanics of newborn CC is derived from adult, manikin and/or more or less appropriate animal models that often bear little resemblance to an asphyxic neonate with poor circulation and open fetal shunts. Good transitional and survival models to examine neurodevelopmental outcomes are missing. There is also an urgent need for randomised clinical trials (RCTs). To date no RCT examining any aspect of neonatal CC has been performed. CPR effectiveness depends on interplay between physiological, mechanical and pedagogical aspects. Real-time feedback systems for CC rate and depth in neonatal CPR are evolving53 ,54 and may potentially be useful in teaching and training, as well as clinical CPR. Such systems also allow for force-targeted and pressure-targeted rather than depth-targeted CC. These are interesting aspects to keep in mind when designing neonatal CC studies.
Conclusion
Successful CPR requires the delivery of high-quality CC, encompassing several factors including (A) optimal C:V ratio (B) adequate CC rate, (C) depth, (D) full recoil between CC, (E) correct technique and (F) adrenaline. Animal studies with high translational value, as well as RCTs are urgently required to understand and improve neonatal CC.
References
Footnotes
Contributors ALS, P-YC, MOR and GMS: Conception and design; collection and assembly of data; drafting of the article; critical revision of the article for important intellectual content; and final approval of the article.
Funding ALS is supported by the Canadian Institutes of Health Research (operating grant MOP-CIA-299111 to P-YC and travel grant to ALS) and the South-Eastern Norway Regional Health Authority. MOR is supported by a Molly Towell Perinatal Research Foundation Fellowship. GMS is a recipient of the Heart and Stroke Foundation/University of Alberta Professorship of Neonatal Resuscitation and a Heart and Stroke Foundation Canada Research Scholarship. The authors have no financial relationships relevant to this article to disclose, real or perceived. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.