Throughout the article, the authors focused on base excess (BE), which was used for the characterization of the three groups (ie, normal daytime Paco2 ≤ 6 kPa associated with BE < 2 mmol/L [non-OHS] or BE ≥ 2 mmol/L [early OHS], and elevated daytime Paco2 > 6 kPa [OHS]). Overall, the authors highlighted BE in early OHS rather than arterial standard [HCO3−]. Nevertheless, results show a strong overlap of arterial standard [HCO3−] and BE values between early-OHS and OHS groups. It should be noted that arterial standard [HCO3−] obtained from the ABL90FLEX blood-gas analyzer (Radiometer Medical ApS) is not measured but calculated according to complex formulas including correction factors for hemoglobin and oxygen saturation and that BE is calculated from arterial standard [HCO3−]. Previous analytical comparisons of blood-gas analyzers have shown that BE values strongly differ between devices, especially due to the difference between calculation formulas, and that the reproducibility of BE measurements was not as good as those for arterial standard [HCO3−].2 Moreover, as referenced on the Westgard QC website,3 within- and between-subject biologic BE variability is much larger than that of bicarbonate (76.4% and 43.2% compared with 4.0% and 4.8%, respectively), making the use of BE in early OHS somewhat questionable.