Abstract
SIGNIFICANCE: Anemia is a common problem in preterm neonates, and red blood cell transfusion (RBCT) is used to improve oxygen delivery. However, RBCT is associated with complications, although an increase in cerebral oxygenation has been documented, and no universally accepted biomarker for the need for transfusion (i.e., the concentration of hemoglobin in the blood) has been defined. AIM: We used a hybrid optical device (BabyLux device) that merges time-domain near-infrared spectroscopy (TD-NIRS) and diffuse correlation spectroscopy (DCS) to potentially obtain a better assessment of the cerebral effects of RBCT compared with previous studies using continuous wave (CW) spatially resolved NIRS. APPROACH: Eighteen clinically stable preterm neonates were assessed before and after RBCT by the BabyLux device as five repetitions of 60 s measurement (with 1 s acquisition time), estimating the cerebral blood flow (CBF) as a blood flow index (BFI), the total hemoglobin concentration (tHb), and the cerebral tissue oxygen saturation ( StO2 ). StO2 was also continuously monitored by a commercial CW-NIRS device, as well as peripheral saturation, SpO2 . Tissue oxygen extraction (TOE) and cerebral metabolic rate of oxygen consumption ( tCMRO2 ) were computed, and the Wilcoxon signed-rank test for paired data was performed, comparing the data acquired before and after RBCT. RESULTS: The BabyLux data from four neonates did not meet quality criteria and were discarded. After the transfusion, tHb and StO2 (measured both with TD-NIRS and CW-NIRS devices) significantly increased, causing a significant decrease in TOE. CW-NIRS showed a wider dispersion of StO2 data compared with TD-NIRS. However, CBF did not decrease proportionally but the variation was high, as well as for tCMRO2 . CONCLUSIONS: The results confirm previous CW-NIRS studies, but the wide variability of BFI makes the effects of RBCT on cerebral metabolism uncertain.