Titration of inspired oxygen in preterm infants with hypoxemic respiratory failure using near‐infrared spectroscopy and pulse oximetry: A new approach

Abstract

Background: Titration of inspired oxygen is a challenge in preterm infants with hypoxemic respiratory failure (HRF). Monitoring of brain oxygen by near-infrared spectroscopy (NIRS) has been proven to minimize the burden of hyperoxia and hypoxemia; with a better understanding of cerebral autoregulation, integrating NIRS and pulse oximetry for titrating inspired oxygen in preterm infants is a novel approach.Methods: We studied the impact of integrated monitoring of oxygen saturation by pulse oximetry (SpO 2 ) and cerebral regional tissue oxygen (crRTO) by NIRS during a stepwise oxygen reduction test (ORT) on reducing oxygen requirement in preterm infants with HRF. The correlation between SpO 2 with crRTO, and fractional oxygen extraction (FOE) was assessed, concordance levels (r > 0.5) were determined during the assessment period and were considered as a sign of impaired autoregulation. The primary outcome was the achievement of significantly lower FiO 2 at 72 h after the start of the integrated monitoring.Results: A total of 38 preterm infants were included, 27 had normal cerebral autoregulation (CAR) (Group 1) in whom SpO 2 was poorly correlating with crRTO with (r < 0.5) and had a significantly greater percentage of reduction below baseline in FiO 2 (mean: 34%). Eleven infants had impaired CAR (Group 2) with SpO 2 significantly correlating with crRTO (r > 0.5) and had a linear trend of FOE inverse to SpO 2 and crRTO; this was considered as an arterial saturation dependent oxygen delivery (SadDO 2 ). Conclusion:Integrated monitoring of preterm infants by SpO 2 and crRTO was associated with easier weaning of oxygen with less burden of both hyperoxia and hypoxemia.

Assessment of Autoregulation of the Cerebral Circulation during Acute Lung Injury in a Neonatal Porcine Model

“…The proportion of cerebral autoregulation impairment attributable to blood pressure, and to hypoxemia, was calculated by established methods [12,13] using a computational algorithm integrating existing published nomograms for cerebral flow regulation with mBP, PaO 2 and PaCO 2 [14][15][16] to detect the degree of variation from these nomograms in experimental data.…”

“…To further analyze these data, we used an autoregulation calculation algorithm designed to best fit changes in cerebral SO 2 with contemporaneous changes in blood pressure, PaCO 2 and/or PO 2 across a range of FiO 2 values, using established nomograms from large samples of neonatal populations [14][15][16]. The output (Figure 6) identifies the proportion of dynamic autoregulation failure attributable primarily to pressure variation, O 2 tension variation or to CO 2 variation, by curve fitting to these nomograms [13]. As a computational model, the algorithm must still be externally validated, but these results track with the manually calculated comparisons made in Figures 4 and 5 and serve to integrate these findings.…”