Accuracy and precision of buccal pulse oximetry

We sought to describe the accuracy and precision of buccal pulse oximetry (SbpO(2)) compared with arterial oxygen saturation (SaO(2)) and pulse oximetry (SpO(2)) in healthy adults at normoxemia and under 3 induced hypoxemic conditions.

In this prospective, correlational study, SbpO(2), SaO(2), and SpO(2) values were recorded at normoxemia and at three hypoxemic conditions (SpO(2)=90%, 80%, and 70%) for 53 healthy, nonsmoking adults who were without cardiac or pulmonary disease, baseline hypoxemia, peripheral edema, dyshemoglobinemia, and fever. Bland-Altman analyses were used to assess agreement and precision between SbpO(2) and SaO(2) measures and between SbpO(2) and SpO(2) measures. Data were adjusted to account for a lag time between buccal and finger sites.

When comparing SbpO(2) and SaO(2) values, mean differences of -1.8%, .3%, 2.4%, and 2.6% were evident at the normoxemia, 90%, 80%, and 70% levels, respectively. When comparing SbpO(2) and SpO(2) values, the mean differences were -1.4%, .1%, 3.3%, and 4.7% at the normoxemia, 90%, 80%, and 70% levels, respectively. The SbpO(2) and SaO(2) values met a priori precision criteria (1.6%; 95% confidence limit, -4.9% to 1.3%) at normoxemia. The SbpO(2) and SpO(2) values met precision criteria at normoxemia (1.5%; 95% confidence limit, -4.4% to 1.5%) and 90% (1.9%; 95% confidence limit, -3.6% to 3.8%) conditions, but exceeded precision criteria at the other tested conditions. On average, SpO(2) lagged 21 seconds behind SbpO(2).

Buccal oximetry is an inaccurate and imprecise method of assessing SpO(2) when oxygen saturation is <90%. The divergence between SbpO(2) and both SaO(2) or SpO(2) values increased as hypoxemia worsened. The buccal method overestimated oxygen saturation in proportion to the degree of hypoxemia. Such overestimates may lead nurses to conclude falsely that a patient's arterial oxygen saturation is acceptable when further assessment or intervention is warranted.