• anti-doping;
  • analytical error;
  • carboxyhemoglobin

The assessment of total hemoglobin mass (tHb-mass) with the optimized carbon monoxide-rebreathing procedure (oCOR) is discussed as a promising method to detect blood doping. The method requires repeated measurements of the carboxyhemoglobin fraction (%HbCO) using spectrophotometers (CO oximeters). In order to determine whether %HbCO measurements with different spectrophotometers yield similar tHb-masses, the results of 57 tHb-mass calculations from simultaneous %HbCO measurements with two different spectrophotometers (RapidLab, OSM3) were analyzed. For the comparison of longitudinal tHb-mass alterations (ΔtHb-mass), 3 tHb-mass measurements were obtained at 6-month intervals (33–37 subjects). Because of significant differences in %HbCO measurements, the limits of agreement for tHb-massOSM3 and tHb-massRapidLab were 11.2% (95% reference range −6.8 to +15.6%) and the correlation of ΔtHb-masses as determined with the two spectrophotometers over two time intervals was weak (r: 0.28–0.66). In only about 70% of all ΔtHb-mass estimations did ΔtHb-massOSM3 and ΔtHb-massRapidLab show the same direction of change. Apparently, the analytical variation in tHb-mass determination with oCOR increases considerably with the use of different spectrophotometers. Therefore, agreement on the use of one spectrophotometer that accurately measures low %HbCO values is needed if oCOR should be used in an anti-doping setting.