Abstract
The aim of this study was to determine whether it is possible using ear-oximetry to prescribe the correct oxygen flow rates during exercise in chronic obstructive pulmonary disease (COPD) patients on long-term oxygen therapy (LTOT). Twenty COPD patients on LTOT, with exercise desaturation breathing oxygen at resting flow rates, performed a series of 6-min treadmill walking tests, with a progressive increase in oxygen flows until oxygen saturation measured by ear- or pulse-oximetry (Sp,O2) was above 90%. The exercise studies were repeated the next day, saturation being measured both noninvasively by ear-oximetry (Sp,O2) and invasively by CO-oximeter (Sa,O2). The exercise studies continued until both Sa,O2 and Sp,O2 were above 90%. Reproducibility and agreement of the results were analysed according to Bland and Altman. Sp,O2 was significantly lower than Sa,O2 by, on average, 0.7% (p < 0.004). Sp,O2 reproducibility between the two days was good. The invasive and noninvasive oxygen flow prescriptions agreed in only 10 subjects; in six subjects ear-oximetry over-estimated the oxygen supply (p < 0.0005), whilst in four subjects it underestimated (p < 0.01). Contingency table analysis with coded raw data for the values of the sixth minute (that of the deepest desaturation) showed poor agreement between CO- and pulse-oximetry (Chi-squared p < 0.003). However, theoretically, if the Sp,O2 target had been raised to 93%, there would have been hardly any underestimations of Sa,O2 p = NS). We concluded that noninvasive measurement of oxygen saturation is not adequate for estimating arterial saturation in chronic obstructive pulmonary disease. We suggest, as a working solution, that a new cut-off limit of 93% oxygen saturation measured by pulse oximetry should be used as the value below which exercise-induced desaturation should be corrected in order to allow oxygen to be properly prescribed during activities of daily life.