Pulse oximetry - methodological conciderations : Experimental, in vitro and human investigations

Sammanfattning: Pulse oximeters monitor non-invasively the haemoglobin oxygen saturation. Light of two specific wavelengths, red and infrared light, is shined through a tissue, usually the finger, and the intensity of the light transmitted is measured. Oxyhaemoglobin absorbs the light differently when compared to reduced haemoglobin. The light transmission is measured on the pulse beat. It is therefore possible to calculate and display the arterial oxygen saturation. Pulse oximeters are primatily calibrated on healthy humans during transient mild hypoxia. Severe hypoxia is not extensively studied. Furthermore, the influence of alterations in blood flow and haematocrit has not been sufficiently investigated. The aims of this study were therefore to evaluate the accuracy of pulse oximetry during severe hypoxia and assess the influence of haematocrit and blood flow changes on the pulse oximeter readings. An animal- and an in vitro model have been developed, and to be able to assess the influences of blood flow a pulse oximeter probe was attached to the foot of volunteers during limb elevation. In addition, a developed prototype of a fibre optic sensor, measuring heart and respiratory rates using pulse oximeter technique, was compared with visual observation, capnography.impedance plethysmography and an acoustic sensor. Results from the animal model showed a reduced pulse oximeter accuracy when the invasive oxygen saturation was below approximately 85 %, and that the light absorption characteristics for the two wavelengths used in pulse oximetry (660 nm and 940 nm) were similar for rabbit- and human haemoglobin. After haemodilution, an increase in the pulse oximeter readings was found, resulting in an increased accuracy when the invasive oxygen saturation was over 85 %. The results from the in vitro model showed a strong dependency of the pulse oximeter accuracy on the haematocrit value and indicated that the pulse oximeter mainly measures the oxygen saturation in smaller pulsating vessels. During limb elevation in humans, resulting in reduced blood flow, the quotient between red and infrared light was changed resulting in a falsely low pulse oximeter reading. The new fibre optic sensor correlated well with the other methods monitoring respiratory rate. During apnoea the fibre optic sensor detected the event correctly. In conclusion, pulse oximetry correlated well with invasively measured oxygen haemoglobin saturation values when SaO2 > 85 %. Overestimation by the pulse oximeter was found during severe hypoxia. A dependence on the haematocrit value was found in studies using an in vitro- and an animal model. During reduction of blood perfusion in humans, the pulse oximeter readings varied with changes in blood flow. The new fibre optic sensor technique to measure respiratory rate, correlated well with other techniques measuring respiratory rate. The fibre optic technique may have several clinical advantages.