Comparison of end-tidal carbon dioxide, oxygen saturation and clinical signs for the detection of oesophageal intubation.

The reliability of various methods for detecting oesophageal intubation was assessed by means of a single blind study in rats. Both oesophagus and trachea were simultaneously intubated. The presence or absence of various clinical signs was noted during tracheal or oesophageal ventilation and arterial blood gases and end-tidal CO2 were measured. Oesophageal ventilation for one minute was associated with significant decreases (P less than 0.001) in pH, PaO2 and SaO2 and a significant (P less than 0.001) increase in PaCO2. Although mean PaO2 decreased by 70 per cent and mean SaO2 decreased by 31 per cent, 43 percent of rats failed to demonstrate a decrease in SaO2 below 85 per cent. Oxygen saturation was the least reliable method for detecting oesophageal intubation (sensitivity = 0.5, specificity = 0.9, positive predictive value (PPV) = 0.8). Chest movement was the most reliable clinical sign for detecting oesophageal intubation (sensitivity = 0.9, specificity = 1.0, PPV = 1.0). Oesophageal rattle was the second most reliable clinical sign (PPV = 0.9). Moisture condensation in the tracheal tube (PPV = 1.0) and abdominal distension (PPV = 0.9) were judged to be the least reliable because each had a high false negative rate of 0.3. The most reliable method for the early detection of oesophageal intubation in rats was end-tidal, CO2 (sensitivity 1.0, specificity = 1.0, PPV = 1.0). In addition, end-tidal CO2 when used in conjunction with the four clinical signs improved the reliability of these signs.