Noninvasive hemoglobin monitoring during red blood cell transfusions in very preterm infants.

BACKGROUND

Noninvasive hemoglobin (NHb) correlates excellently with laboratory Hb in both preterm and term infants. Our aim was to monitor NHb during red blood cell (RBC) transfusions to assess its changes and evaluate the correlation between NHb and capillary hemoglobin (cHb).

METHODS

We studied 30 transfused preterm infants with a gestational age of 25.8 ± 2.7 weeks. NHb and cHb values were measured before and at the end of transfusions. Regression analysis and The Bland-Altman plot were performed to determine the relationship, the bias, and level of agreement (LOA) between NHb and cHb.

RESULTS

There was a significant increase of both NHb (9.1 ± 1.2 vs. 12.4 ± 1.5 g/dL; p < .001) and cHb (8.2 ± 0.7 vs. 12.2 ± 0.8 g/dL; p < .001) from before to after the transfusions. Linear regression analysis demonstrated a moderately positive relationship between NHb and cHb (r = 0.753; p < .001). The bias and LOA for the NHb and cHb values were -0.256 ± 1.412 g/dL with a LOA of -3.024 to 2.513 g/dL. However, the bias and LOA for NHb and cHb values were significantly higher in anemic patients with low cHb (bias = -0.858 ± 1.290, LOA -3.386 to 1.670 g/dL).

DISCUSSION: NHb significantly increased during RBC transfusions in anemic preterm infants. We found a moderately positive correlation between NHb and cHb values, but bias and agreement were significantly worse when Hb levels were lower before transfusion. Our results suggest that NHb may be useful to assess its relative changes and not to assess the absolute value of Hb.