ABSTRACT
AIM: To determine the incidence of positive fluoroscopic oesophagography in patients presenting with spontaneous or blunt traumatic pneumomediastinum. MATERIALS AND METHODS: Retrospective chart review was performed on patients who underwent fluoroscopic oesophagography for spontaneous or blunt traumatic pneumomediastinum between 2001-2019. Patients were excluded for history of oesophageal surgery, penetrating trauma, oesophageal cancer, or tracheal/oesophageal instrumentation. RESULTS: Two hundred and fifty-two patients met the inclusion criteria; 170 presented with spontaneous pneumomediastinum and 82 presented with blunt traumatic pneumomediastinum. Fluoroscopic oesophagography was positive in eight patients with spontaneous pneumomediastinum, for a positivity rate of 4.7% (8/170). There was one false-negative case in a patient who presented with spontaneous pneumomediastinum and was found to have a non-full-thickness oesophageal injury on endoscopy. Fluoroscopic oesophagography was negative in all patients with blunt traumatic pneumomediastinum (0/82). The sensitivity and specificity of fluoroscopic oesophagography were 88.9% (8/9) and 100% (243/243), respectively. Oesophageal injury was more common in patients with spontaneous pneumomediastinum and a pleural effusion (5/11, 45.4%) than in patients with spontaneous pneumomediastinum and no pleural effusion (4/159, 2.5%, p<0.001). CONCLUSION: The present findings do not support routine oesophagography in patients with blunt traumatic pneumomediastinum. Conversely, a positivity rate of 4.7% in patients with spontaneous pneumomediastinum suggests oesophagography may be warranted in this population, particularly if an associated pleural effusion is present.