PURPOSE: Review the clinical hallmarks and management of spontaneous pneumomediastinum, and analyze how it differs trauma-related pneumomediastinum.
METHODS: A retrospective analysis was performed of all patients with a finding of pneumomediastinum, over a 10-year period. Penetrating thoracic injuries were excluded. Two groups were statistically compared: Trauma-related pneumomediastinum and spontaneous pneumomediastinum.
RESULTS: Pneumomediastinum was found in seventy-four patients. The suspected causes were blunt thoracic trauma (46%), barotrauma (16%), forceful emesis (8%), forceful coughing (5%), asthma (5%), esophageal perforation (5%) and others (15%).The sensitivity of chest-X-ray (CXR) to diagnose pneumomediastinum was 26%. Spontaneous pneumomediastinum was equally seen in males and females and was associated with pneumothorax in 11% and subcutaneous emphysema in 44%. The mortality was 23%.Trauma-related pneumomediastinum, as compared to spontaneous pneumomediastinum, was less likely to be diagnosed by CXR (16% vs 55%, p<0.05), more likely to develop a pneumothorax (66% vs 11%, p<0.05), more likely to have an associated pleural effusion (22% vs 0%, p<0.05) and had a longer hospital stay (12 vs 4 days, p<0.05).Esophagogram and bronchoscopy were performed in 62% of the patients, and were invariably negative in the absence of overt signs of esophageal or tracheobronchial injury. Follow-up was obtained in 75% and no recurrence was identified.
CONCLUSION: Chest radiography has a low yield in the diagnosis of pneumomediastinum, with an even lower usefulness in the trauma population. Patients with spontaneous pneumomediastinum have a shorter hospital stay and are less likely to develop pneumothoraces and pleural effusions. Esophagogram and bronchoscopy may be overutilized in this patient population.
CLINICAL IMPLICATIONS: Chest radiography understimates the true incidence of pneumomediastinum; however, is the main means for its diagnosis. Pleural effusions and pneumothoraces, if present, portend a more serious underlying injury and are unusual in spontaneous pneumomediastinum. Esophogram and bronchoscopy may be safely forgone in spontaneous pnemomediastinum with low likelihood of esophageal perforation.
DISCLOSURE: Manuel Caceres, None.