[Transesophageal arterial oxygen saturation monitoring: an experimental study].

Objective: To assess the sensitivity and accuracy of a novel transesophageal approach to monitoring the descending aortic oximetry (SeO2).

Methods: Nine dogs were involved in the experimental study. After the induction of anaesthesia, the carrier of the oximetry probe (Nellcor-D20, USA) was inserted into the lower segment of esophagus to monitor SeO2, and the probe was "locked" in position of post-descending aorta after the opening of thoracic cavity. Another probe was pasted on the surface of lingual mucous membrane. The readings and figures of SeO2 and surface of lingual mucosa oximetry (SmO2) were observed continuously and recorded simultaneously. Vital signs were monitored with pulse oxygen saturation (SpO2), invasive blood pressure by femoral artery, HR, EKG, PetCO2, T, FiO2. The SaO2 of blood gas analysis by femoral artery was used as the "gold standard" to calculated the relative and absolute deviations of SeO2 and SmO2. The changes of SeO2 and SaO2 were compared in case of acute hypoxia when values of SmO2 dropped to 90%, 80%, 70%, 60% and the patient was re-ventilated with 100% oxygen.

Results: (1) SeO2, SmO2 and SaO2 were 100% when the patients were ventilated with 100% oxygen. During hypoxia, the descent of SeO2 from 100% to 90% was (91.03+/-20.23) s (P<0.001) earlier than that of SmO2. And after re-supply of pure oxygen, the ascent of SeO2 was (25.9+/-6.0) s (P<0.05) earlier than that of SmO2. (2) SaO2 was well related with SeO2 and SmO2 (R2: 0.9884 and 0.9296) respectively. The relative and absolute deviations of SeO2 were 1.6% and 1.3%, while those of SmO2 were 7.6% and 6.1% from arterial blood samples SaO2. (3) There were no significant differences in MAP, HR, ECG, PetCO2 and T.

Conclusion: This study showed that SeO2 monitoring is sensitive. It could accurately reflect the arterial oxygen saturation not only in normal condition but also during hypoxia and the re-ventilation with 100% oxygen. SeO2 responds faster and is closer to SaO2, compared with SmO2 measurements. This may be an alternative method in the cases where the monitoring of peripheral SpO2 is difficult.