Simple methods for improving the value of oximetry in the study of pulmonary oxygen uptake.

The purpose of this paper is to show how an ear oximeter can be used to measure the oxygenation of the blood by the lungs with much greater discrimination than is generally appreciated. There are three stages: the first is to calibrate the oximeter by a practicable method; the second is to use the instrument to construct an oxygen dissociation curve; and the third is to minimize an important source of variability (the oxygen capacity of the blood) by dealing with differences in saturation instead of absolute saturation. In normal subjects at sea level the arterial 02 saturation is in the range 92 to 98%. Four sources of variability contribute to this wide range of arterial saturation and its consequent lack of discrimination as a means of assessing respiratory function. First there is variability in the Po2 of the alveolar air (PAo2). This is due chiefly to variation in alveolar ventilation but also to variation in barometric pressure and in the respiratory exchange ratio or respiratory quotient. Secondly, there is variability in the difference or gradient of 02 pressure between the alveolar air and the arterial blood. This is chiefly due to variation in ventilation: perfusion relationships but also to diffusion, chemical reactions, and 'true' shunts. Thirdly, there is variability in the relation between arterial Po2 and saturation-that is, variation in the position and slope of the 02 dissociation curve. Fourthly, there is analytical variability. Some of this is, of course, random, but much of it is systematic and stems from variability in the estimate of the 02 'capacity' of the blood which causes error because the 02 saturation is (02 content 0 °2 capacity) x 100. Of these four sources of variability the first can now be readily eliminated by estimating arterial