The Masimo Radical 7 (Masimo Corp., Irvine, CA, USA) pulse co-oximeter^® calculates haemoglobin concentration (SpHb) non-invasively using transcutaneous spectrophotometry. We compared SpHb with invasive satellite-lab haemoglobin monitoring (Hb_satlab) during major hepatic resections both under steady-state conditions and in a dynamic phase with fluid administration of crystalloid and colloid solutions.

Thirty patients undergoing major hepatic resection were included and randomized to receive a fluid bolus of 15 ml kg^–1 colloid (n=15) or crystalloid (n=15) solution over 30 min. SpHb was continuously measured on the index finger, and venous blood samples were analysed in both the steady-state phase (from induction until completion of parenchymal transection) and the dynamic phase (during fluid bolus).

Correlation was significant between SpHb and Hb_satlab (R²=0.50, n=543). The modified Bland–Altman analysis for repeated measurements showed a bias (precision) of –0.27 (1.06) and –0.02 (1.07) g dl^–1 for the steady-state and dynamic phases, respectively. SpHb accuracy increased when Hb_satlab was –1, with a bias (precision) of 0.41 (0.47) vs –0.26 (1.12) g dl^–1 for values >10 g dl^–1, but accuracy decreased after colloid administration (R²=0.25).

SpHb correlated moderately with Hb_satlab with a slight underestimation in both phases in patients undergoing major hepatic resection. Accuracy increased for lower Hb_satlab values but decreased in the presence of colloid solution. Further improvements are necessary to improve device accuracy under these conditions, so that SpHb might become a sensitive screening device for clinically significant anaemia.