Occurrence and outcomes of acquired tracheoesophageal fistula in mechanically ventilated adults in a five year single center study

Purpose Acquired tracheoesophageal fistula (TEF) can complicate invasive ventilation, but its occurrence across an entire intensive care population is poorly characterized. Treatment-focused series may not capture patients who remain critically ill and cannot undergo definitive repair. We aimed to quantify the observed institutional occurrence of TEF and describe the clinical course and outcomes of affected patients. Methods We retrospectively reviewed adults with endoscopically confirmed acquired TEF diagnosed in a tertiary intensive care unit between 2021 and 2025. Annual ventilation records were merged and deduplicated to identify 3,442 unique invasively ventilated patients. Clinical and anatomical findings, treatment, and all-cause in-hospital mortality were summarized descriptively. Exact 95% confidence intervals (CIs) were calculated for selected proportions. Results TEF was identified in 20 patients, corresponding to an observed five-year institutional proportion of 0.58% (95% CI, 0.36–0.90%), or 5.81 per 1,000 ventilated patients. Median age was 70 years, and median time from intensive care admission to TEF diagnosis was 46 days (interquartile range, 33–65.5). Median total ventilation duration was 60.5 days. Tracheostomy preceded TEF diagnosis in 14 patients (70%). Ventilator-associated pneumonia was documented in 19 (95%) and septic shock in all 20. Median fistula diameter was 2.0 cm. Five patients underwent esophageal metallic stenting; closure was not documented, and none underwent definitive surgical repair. All 20 died during hospitalization (100%; 95% CI, 83.2–100%). Conclusions Acquired TEF was detected in approximately one in 172 invasively ventilated patients and occurred in a cohort with an exceptionally poor hospital outcome. Clinically triggered ascertainment may underestimate its occurrence. These findings support sustained diagnostic vigilance during prolonged airway instrumentation, while neither a higher-than-expected incidence nor mortality attributable specifically to TEF can be established. Clinical Trial Or Systematic Review Registration Not applicable. This retrospective observational study was not registered in a public registry.