Pulse oximetry is a poor predictor of hypoxemia in stable children with sickle cell disease

Objective: To evaluate the accuracy of the pulse oximeter to detect hypoxemia in patients with sickle cell disease in an ambulatory care setting. Study Design: Simultaneous measurements of PaO₂, arterial oxygen saturation by co-oximetry (criterion standard), and pulse oximetry were performed in 21 children with sickle cell disease during 22 outpatient visits. The bias and precision of the pulse oximeter compared with measured arterial oxygen saturation by co-oximetry were determined. The sensitivity, specificity, and positive and negative predictive values of the pulse oximeter to detect hypoxemia (PaO₂ <70 mm Hg) were also calculated. Results: The mean difference between pulse oximetry and measured oxygen saturation (bias) was 5.0% and the SD (precision) was 5.3. Twenty-one patients had a PaO₂ greater than 70 mm Hg; 7 of these (33%) were predicted to be hypoxic by pulse oximetry with values less than 93%, for a specificity to detect normoxia of 67%. Conclusion: Making treatment decisions based on pulse oximetry data alone in patients with sickle cell disease who are not acutely ill may be inappropriate.