Baseline haemoglobin variability by measurement technique in pregnant people on the day of childbirth

Point-of-care haemoglobin measurement devices may play an important role in the antenatal detection of anaemia in pregnant people and may be useful in guiding blood transfusion during resuscitation in obstetric haemorrhage. We compared baseline haemoglobin variability of venous and capillary HemoCue^® haemoglobin, and Masimo^® Rad-67 Pulse CO-Oximeter haemoglobin with laboratory haemoglobin in people on the day of their planned vaginal birth. A total of 180 people undergoing planned vaginal birth were enrolled in this prospective observational study. Laboratory haemoglobin was compared with HemoCue and Masimo Rad-67 Pulse CO-Oximeter measurements using Bland-Altman analysis, calculating mean difference (bias) and limits of agreement. Five (2.8%) people had anaemia (haemoglobin < 110 g.l^-1 ). Laboratory haemoglobin and HemoCue venous haemoglobin comparison showed an acceptable bias (SD) 0.7 (7.54) g.l^-1 (95%CI -0.43-1.79), with limits of agreement -14.10-15.46 g.l^-1 and acceptable agreement range of 29.6 g.l^-1 . Laboratory and HemoCue capillary haemoglobin comparison showed an unacceptable bias (SD) 13.3 (14.12) g.l^-1 (95%CI 11.17-15.34), with limits of agreement - 14.42-40.93 g.l^-1 and unacceptable agreement range of 55.3 g.l^-1 . Laboratory and Masimo haemoglobin comparison showed an unacceptable bias (SD) -14.0 (11.15) g.l^-1 (95%CI -15.63 to -12.34), with limits of agreement to -35.85 to 7.87 g.l^-1 and acceptable agreement range of 43.7 g.l^-1 . Venous HemoCue, with its acceptable bias and limits of agreement, should be applied more widely in the antenatal setting to detect, manage and risk stratify pregnant people with anaemia. HemoCue capillary measurement under-estimated haemoglobin and Masimo haemoglobin measurement over-estimated, limiting their clinical use. Serial studies are needed to determine if the accuracy of venous HemoCue haemoglobin measurement is sustained in other obstetric settings.