Reliability and validity of non-invasive determined haemoglobin mass and blood volumes

Introduction: The carbon monoxide (CO) rebreathing method used for the determination of haemoglobin mass (Hbmass) is associated with blood sample analysis (in this study: Radiometer ABL800). As an alternative hereto the aim of the present study was to evaluate the use of a portable and non-invasive CO pulse oximeter (Rad-57). Method: With simultaneous determination of CO in the circulation by ABL800 (%HbCO) and Rad-57 (SpCO), Hbmass and blood volume (BV) were determined in duplicates in 24 volunteers. Percentage of typical errors (%TE) within methods and linear correlations between the two procedures were computed. Results: Hbmass (Rad-57 = 798 ± 230 g; ABL800 = 781 ± 192 g) and BV (Rad-57 = 5700 ± 1373 ml; ABL800 = 5581 ± 1096 ml) were similar between methods. However, the %TE for Hbmass was higher (P<0·001) for Rad-57 (5·84 ± 5·29%) than for ABL800 (1·35 ± 1·13%). Similarly, the %TE for BV was higher (P<0·001) for Rad-57 (6·06 ± 5·76%) than for ABL800 (1·48 ± 1·25%). Lower (P<0·05) correlation coefficients between the methods were found when Hbmass > 905 g and BV > 6193 ml. Conclusion: Assessment of SpCO by Rad-57 resulted in considerably less precise determinations of Hbmass and BV, especially for high values. Thus, non-invasive assessment of Hbmass and BV cannot be recommended for scientific purposes, but may nonetheless be useful in clinical settings.