Optimal oxygen use in neonatal advanced cardiopulmonary resuscitation-a literature review

Abstract

Background and objectives: Oxygen (O₂) use during neonatal cardiopulmonary resuscitation (CPR) remains a subject of controversy. The inspired O₂ concentration during neonatal CPR, that hastens return of spontaneous circulation (ROSC), allows adequate cerebral and myocardial O₂ delivery, and enhances survival to discharge, is not known. The optimal FiO₂ during CPR should decrease incidence of hypoxia but also avoid hyperoxia, and ultimately lead to improved neurodevelopmental outcomes. Due to infrequent need for extensive resuscitation, and emergent circumstances surrounding neonatal CPR, conducting randomized clinical trials continues to be a challenge. The goal of this study was to review the evolution of oxygen use during neonatal CPR, the evidence from animal and clinical studies on oxygen use during neonatal CPR and after ROSC, the pertinent physiology including myocardial oxygen consumption and cerebral oxygen delivery during CPR, and outcomes following CPR in the DR and in the neonatal intensive care unit.

Methods: This narrative review is based on recent and historic English literature in PubMed and Google scholar over the past 35 years (January 1, 1985 - May 1, 2021).

Key content and findings: Several studies in animal models have compared ventilation with different inspired O₂ concentrations (mostly 21% and 100%) during chest compressions and after ROSC. These studies reported no difference in short-term outcomes, even with as low as 18% O₂. However, in lamb models of cardiac arrest and CPR, 100% O₂ during chest compressions is associated with better oxygen delivery to the brain compared to 21% O₂. Abrupt weaning to 21% O₂ following ROSC followed by titration to achieve preductal SpO₂ of 85-95% minimizes systemic hyperoxia and oxidative stress compared to slow weaning from 100% O₂ following ROSC.

Conclusions: Clinical research is needed to arrive at the best strategy for assessment of oxygenation and choice of FiO₂ during neonatal CPR that lead to improved survival and outcomes. In this article, we have reviewed the literature on evidence behind O₂ use during neonatal advanced CPR and after ROSC.

Keywords: Oxygen (O2); chest compressions (CC); hyperoxia; neonatal cardiopulmonary resuscitation; neonatal resuscitation.

Figure 1

Change in carotid arterial and jugular venous oxygen content after recovery from cardiac arrest and bradycardia. Change in carotid arterial and jugular venous O₂ content following ROSC in a lamb model of cardiac arrest are shown in (A). The AVDO₂ is low in the first 10 min after ROSC indicating reduced cerebral O₂ consumption after recovery from asphyxial arrest in term newborn lambs. The typical AVDO₂ in fetal and neonatal lambs is shown in the yellow box in (A). Carotid arterial and jugular venous oxygen content, and arteriovenous differences (AVDO₂) after ROSC from lamb models of cardiac arrest induced by umbilical cord compression [B, (i)] and bradycardia [with meconium aspiration, (B), (ii)] are presented as bar graphs. These suggest that oxygen consumption by the brain is low in neonatal lambs following recovery from cardiac arrest and bradycardia. Copyright Satyan Lakshminrusimha. O₂, oxygen; ROSC, return of spontaneous circulation; AVDO₂, arteriovenous difference in oxygen content.

Figure 2

Changes in hemodynamics and timing of flows during intrinsic heart beats in a BIOPAC snapshot and an illustration. Changes in aortic pressure, carotid artery blood flow, pulmonary blood flow, ductus arteriosus blood flow and coronary artery blood flow during spontaneous heartbeat are depicted in (A). Pink highlight indicates timing of systole during an intrinsic heartbeat. Except forward coronary flow that occurs during diastole, all the other blood flows (carotid, pulmonary and ductal) occur during systole (A). During chest compressions (B), there is forward flow across the aorta, carotid artery, pulmonary artery and ductus arteriosus (right-to-left) depicted by the positive deflection in the BIOPAC snapshot (pink highlight indicating time of chest compressions). However, there is no forward blood flow across the coronary artery during chest compressions, and the forward flow occurs during the decompression phase (diastole). Illustration of pressure and flow changes in the heart chambers during chest compressions for cardiac arrest are shown in (C). During chest compressions/“systole” (left hand side of the figure), the venous valves are closed (preventing back flow of blood in inferior vena cava), the right atrial pressure is increased, along with open pulmonary and aortic valves, allowing forward flow to the PA, ductus arteriosus (ductus) and ascending aorta. However, there is no forward coronary blood flow during CC as there is no gradient between aortic pressure and right atrial pressure. The pulmonary flow may be limited by compressed lung with increased pulmonary vascular resistance during CC. On the other hand, during decompression phase of CC/“diastole”, the venous valves and atrioventricular valves are open, allowing forward flow from atria to ventricles, but the pulmonary and aortic valves are closed, and forward blood flow occurs in the coronary arteries in between the CC. Furthermore, the flow from left to right across the ductus arteriosus during the decompression phase prevents the increase and build-up of diastolic pressure. Copyright Satyan Lakshminrusimha. PA, pulmonary artery; CC, chest compressions; PV, pulmonary veins; PVR, pulmonary vascular resistance.

Figure 3

Coronary perfusion pressure during spontaneous heartbeats compared to during CC. During spontaneous heartbeat, the aortic pressure is normal, with very low right atrial pressure (BIOPAC snapshot), allowing a gradient between diastolic blood pressure and right atrial pressure of ~15 mmHg which is the coronary perfusion pressure. Forward carotid blood flow, ETCO₂ and EKG are also shown. However, during CC (right hand side of the figure) for cardiac arrest, the aortic diastolic pressure is lower, and the right atrial pressure is higher approaching the diastolic pressure, thus decreasing the difference to 2–3 mmHg resulting in lower coronary perfusion pressure. Reduced carotid artery blood flow (forward flow during CC/systole) and dampened ETCO₂ (due to low pulmonary circulation) are also depicted. Copyright Satyan Lakshminrusimha. CC, chest compressions; ETCO₂, end tidal carbon dioxide; EKG, electrocardiogram.

Figure 4

Differences in carotid and pulmonary hemodynamics during CC. (A) Shows carotid flow, aortic pressure, and pulmonary flow during spontaneous heart beats immediately after birth in term lambs ventilated with 21% Oxygen after cord clamping from a BIOPAC snapshot in non-asphyxiated lambs. Immediately after birth, following clamping of the umbilical cord and ventilation of the lungs with 21% oxygen, carotid flow, aortic pressure and pulmonary flow rapidly increase. The carotid flow is always antegrade both during systole and diastole. The pulmonary flow is predominantly forward with minimal retrograde flow during diastole due to high pulmonary vascular resistance and bidirectional ductal shunting (A). Whereas, (B) is a BIOPAC snapshot depicting changes in hemodynamics during CC and at ROSC. The pink, red and blue horizontal bold lines indicate “zero” value for the carotid flow, aortic pressure and pulmonary flow respectively. During CC, pulmonary flow is bidirectional with minimal forward flow during chest compressions but significant retrograde flow during the decompression phase (B). This results in minimal effective pulmonary flow to support gas exchange during CPR for cardiac arrest. Hence, providing 100% oxygen during CC for cardiac arrest might benefit by reducing duration of CPR and increasing oxygen content in the low flow volume of pulmonary venous return (B). In contrast, carotid flow is predominantly antegrade with only minimal retrograde flow during the decompression phase. We speculate that these differences are secondary to extrathoracic location of carotid vessels—resulting in positive pressure gradient during chest compressions vs. intrathoracic location of pulmonary vessels (as shown in Figure 2C). Copyright Satyan Lakshminrusimha. CC, chest compressions; ROSC, return of spontaneous circulation; CPR, cardiopulmonary resuscitation.