[Measurement of brain regional oxygen saturation in neonates in China: a multicenter randomized clinical trial]

To understand the value of measuring neonatal cerebral regional oxygen saturation (rSO2) using near infrared spectroscopy (NIRS) in assessing cerebral oxygenation, to establish the normal range of neonatal cerebral rSO2 and to collect data of the changes of cerebral rSO2 under certain disease status.

Nine large hospitals participated in the multicenter randomized clinical trial from Jan 2007 to Apr 2008. Using the NIRS human tissue oximeter (TSAH-100) independently developed in China, the cerebral rSO2 of 223 normal full-term and 95 otherwise healthy preterm neonates without any special disease, was detected at 1, 2 and 3 days after birth, respectively. The cerebral rSO2 of 102 neonates with diseases which may affect the cerebral oxygenation, was also detected during the severe phases. The pulse oxygen saturation (SpO2) measured at the finger tip, and also the arterial oxygen saturation (SaO2) measured by blood gas analysis, which could indicate the oxygen supply of the whole body, were obtained simultaneously. The correlations among cerebral rSO2, pulse SpO2 and arterial SaO2 were analyzed.

(1) The cerebral rSO2 of the normal full-term neonates was (62+/-2)%. Cerebral hypoxia was defined as rSO2 lower than 58%. The cerebral rSO2 of the normal full-terms was steady at 1, 2 and 3 days after birth respectively, without any significant differences among them (F=0.610, P>0.05). The cerebral rSO2 of the neonates with diseases was (55+/-7)%, which was significantly lower than that of the normal full-term neonates (t=15.492, P<0.05). (2) The cerebral rSO2 was positively correlated with the SpO2 (r=0.74, P<0.01) and the SaO2 (r=0.71, P<0.01). (3) Under some special diseases, the changes of cerebral rSO2 was asynchronous with those of the SpO2: (1) For 18 cases under severe cerebral damages or under relatively low hemoglobin concentration, the cerebral rSO2 was significantly low (50%-58%), but the SpO2 was still normal (above 90%). (2) During the recovery of some critically ill neonates, the increase of cerebral rSO2 was lagged as compared with that of pulse SpO2. Especially, during the severe phases of 6 cases with multi-organ failure, the SpO2 and the cerebral rSO2 were both significantly low (55%-80% for SpO2, and 44%-50% for cerebral rSO2); when the diseases were alleviated, although the SpO2 recovered to above 85%, the cerebral rSO2 was still significantly low (around 50%). (3) In 3 cases, during the severe phases of serious hypoxic-ischemic encephalopathy (HIE), the cerebral rSO2 significantly increased to 70%-72%, which was significantly higher than the normal value (62%).

The range of cerebral rSO2 of the normal full-term neonates was (62+/-2)%. Cerebral oxygenation can be externally indicated by the rSO2 noninvasively and continuously measured by NIRS, which was positively correlated with traditional pulse SpO2 and arterial SaO2. In some special diseases, the rSO2 measured by NIRS can be helpful for clinical diagnoses and treatments.