Hemodynamics and gas exchange during chest compressions in neonatal resuscitation

Abstract

Purpose: Current knowledge about pulmonary/systemic hemodynamics and gas exchange during neonatal resuscitation in a model of transitioning fetal circulation with fetal shunts and fluid-filled alveoli is limited. Using a fetal lamb asphyxia model, we sought to determine whether hemodynamic or gas-exchange parameters predicted successful return of spontaneous circulation (ROSC).

Methods: The umbilical cord was occluded in 22 lambs to induce asphyxial cardiac arrest. Following five minutes of asystole, resuscitation as per AHA-Neonatal Resuscitation Program guidelines was initiated. Hemodynamic parameters and serial arterial blood gases were assessed during resuscitation.

Results: ROSC occurred in 18 lambs (82%) at a median (IQR) time of 120 (105-180) seconds. There were no differences in hemodynamic parameters at baseline and at any given time point during resuscitation between the lambs that achieved ROSC and those that did not. Blood gases at arrest prior to resuscitation were comparable between groups. However, lambs that achieved ROSC had lower PaO2, higher PaCO2, and lower lactate during resuscitation. Increase in diastolic blood pressures induced by epinephrine in lambs that achieved ROSC (11 ±4 mmHg) did not differ from those that were not resuscitated (10 ±6 mmHg). Low diastolic blood pressures were adequate to achieve ROSC.

Conclusions: Hemodynamic parameters in a neonatal lamb asphyxia model with transitioning circulation did not predict success of ROSC. Lactic acidosis, higher PaO2 and lower PaCO2 observed in the lambs that did not achieve ROSC may represent a state of inadequate tissue perfusion and/or mitochondrial dysfunction.

Fig 1. Illustration of methodology depicting invasive and non-invasive monitoring.

Fig 2. Arterial blood pressures during study period.

There is no difference in mean systolic blood pressures (A) and mean diastolic blood pressures (B) between lambs that achieve return of spontaneous circulation (ROSC) and those with no ROSC. Epinephrine did not statistically increase blood pressures in either group. Data are mean ± SEM bars.

Fig 3. Mean left pulmonary artery flow.

There is no appreciable increase in flow following initiation of chest compressions or after epinephrine administration. Data are mean ± SEM bars.

Fig 4. Mean left carotid artery flow.

There is only a modest increase in flow following initiation of chest compressions with no further increase after epinephrine use. Data are mean ± SEM bars.

Fig 5. Left pulmonary artery flow.

Maximum flow increases with initiation of chest compressions in lambs with ROSC and no ROSC. Reversal of flow occurs in diastole (minimum flow) at baseline (in utero), and remains retrograde during the relaxation phase of chest compressions in lambs of both groups. Flow becomes exclusively anterograde upon ROSC and cessation of chest compressions. Data are mean ± SEM bars.

Fig 6. Left carotid artery flow.

Maximum left carotid flow increases with initiation of chest compressions in lambs with ROSC and no ROSC. During the relaxation phase of chest compressions (minimum flow), retrograde flow is observed in lambs of both groups. Flow becomes exclusively anterograde upon ROSC and cessation of chest compressions. Data are mean ± SEM bars.

Fig 7. Comparison of arterial blood gas analysis between groups during resuscitation.

(A) PaO₂, (B) PaCO₂, (C) pH, (D) Lactate. Gas sample “0” depicts “arrest gas.” Data are mean ± SEM bars. The number of lambs in the ROSC group drops with each successful ROSC (at “0” n = 18, at “1” n = 17, at “2” n = 8, at “3” n = 4, at “4” n = 2, at “6” n = 1).

Fig 8. Comparison of arterial blood gas analysis following ROSC.

(A) PaO₂; the secondary y-axis represents FIO₂, (B) PaCO₂, (C) pH, (D) Lactate. Data are mean ± SEM bars.

Fig 9. Cerebral rSO₂, preductal SpO₂, SaO₂ and SvO₂ during study period in lambs with ROSC.

Pulse oximetry and NIRS are not reliable during resuscitation as shown by a discrepancy between SaO₂ and SpO₂, and SvO₂ and cerebral rSO₂ during chest compressions as well as in the first few minutes into ROSC. Data are mean ± SEM bars. * P <0.05 SvO₂ vs. SpO₂; † P <0.05 SvO2 vs. cerebral rSO₂.