Background: Tracheoesophageal fistulae (TEF) are difficult to detect and require a high index of suspicion. We hypothesized that capnography to identify a spike in end-tidal carbon dioxide (etCO2) during esophagoscopy with carbon dioxide (CO2) insufflation would facilitate TEF diagnosis because of gas passage from the esophagus to the trachea.
Methods: Medical records of 42 consecutive cases of recurrent, acquired, or missed congenital TEF diagnosed between January 2015 and November 2019 that underwent esophagoscopy with CO2 insufflation were reviewed. A control cohort of 97 similarly endoscopically evaluated patients with surgical confirmation of absence of recurrent TEF (eg, patients undergoing posterior tracheopexy) was also collected. All patients underwent pre-operative esophagoscopy, bronchoscopy, and capnography; diagnostic abilities of various combinations of modalities for TEF identification were calculated.
Results: Statistical analysis identified a maximum intra-esophagoscopy end-tidal CO2 level of 68 mmHg as the optimal discriminator between cases and controls, though in practice, we anecdotally find that recurrent TEFs typically permit rapid rise ≥90 mmHg. Increasing numbers of diagnostic modalities increased diagnostic sensitivity to detect recurrent TEF; the highest diagnostic sensitivity for TEF identification was achieved by the combination of intra-esophagoscopy fluoroscopy with bronchoscopy and capnography ≥68 mmHg (sensitivity = 88.1%). There were multiple cases of TEF (N = 7 for etCO2 ≥68 mmHg, N = 3 for etCO2 ≥90 mmHg) identified by capnography that were missed by esophagoscopy. There were 5 (for etCO2 ≥68 mmHg) or 6 (for etCO2 ≥90 mmHg) cases of recurrent TEF that were missed by all nonsurgical methods.
Conclusion: Attention to etCO2 during esophagoscopy with CO2 insufflation represents a simple, novel way to detect TEF. Identification of TEF remains challenging, though combinations of diagnostic modalities improve diagnostic sensitivity.