The pulse oxygen saturation: Inspired oxygen pressure (SpO₂:P₁O₂) diagram: Application in the ambulatory assessment of pulmonary vascular disease

Background: The inspired oxygen to pulse saturation (SpO₂:P ₁O₂) curve mirrors the oxyhaemoglobin dissociation curve but is shifted to the right by the difference between inspired and alveolar oxygen fractions. Inspection of the curve has been used to discern changes in shunt. 

Aims: We aimed to demonstrate the SpO₂: P₁O₂ curve's quantification using readily available software, and its clinical utility in assessing ambulatory patients. 

Methods: Six normal and seven hereditary haemorrhagic telangiectasia (HHT) clinic patients have been studied. After measuring barometric pressure, seated subjects were monitored with finger probe pulse oximetry while breathing increasing fractions of humidified oxygen through a Venturi mask until either their SpO₂ or P₁O₂ was 100%. Resulting SpO₂ and P₁O₂ values were plotted, fitted, assessed for goodness-of-fit and compared using the free, open source nonlinear regression extension package drc for the R programming language whose accuracy was also tested against two NIST reference sigmoidal curve datasets. 

Results: All reference, individual and collected normal results were well fitted by the drc software with highly insignificant goodness-of-fit F tests. The laboratory normal curve was identically fitted by symmetrical Hill and asymmetrical Weibull functions. Although normal and negative HHT screening patient curves did not differ, there were significant differences between normal, pre-treatment (shunt fraction 9.44%), and post-treatment HHT curves. 

Conclusion: Nonlinear regression analysis of the SpO₂:P₁O₂ curve permits valid, simple, safe, noninvasive, sensitive and cheap estimation of shunt in ambulatory patients.