Eighteen patients with pneumomediastinum secondary to COVID-19 at a tertiary medical center

ABSTRACT

INTRODUCTION: COVID-19 is associated with significant pulmonary morbidity. We present a series of patients with pneumomediastinum (PM) at a regional, COVID-only designated hospital. Diagnosis, management, and outcomes were evaluated. METHODS: This retrospective study surveyed COVID-19 patients admitted to a regionally designated COVID-only hospital from March through May 2020. Patients diagnosed with PM were classified into two groups isolated PM (PMO) and PM with concomitant pneumothorax (CMBND). Demographics, ventilatory parameters, and outcomes are included. RESULTS: Of 772 patients evaluated, 18 patients were diagnosed with PM (2%) one-third presented with PMO while two-thirds demonstrated CMBND. Demographic data between groups were comparable with respect to gender, age, history of lung disease, hypertension, and need for mechanical ventilation. Diabetes was more prevalent in the CMBND group. Mean time to diagnosis of PM following admission was 10 days, and 6.5 days following intubation. Indication for evaluation leading to the diagnosis of PM were incidental (44%), post-intubation chest x-ray (22%), clinical deterioration (22%), and subcutaneous emphysema (11%). At intubation, median PEEP (range) among PMO vs. CMBND was 14 (12-15) vs. 14 (10-18) cm H2O. Median PEEP (range) at diagnosis among PMO vs. CMBND was 12 (12-15) vs. 10 (8-21) cm H2O. At intubation, the median PIP (range) among PMO vs. CMBND was 33 (24-38) vs. 32 (23-55) cm H2O. The median PIP (range) at diagnosis among PMO vs. CMBND was 32 (31-45) vs. 38 (24-60) cm H2O. Median tidal volume and FiO2 for both groups were similar while observed mortality was equal (83%). CONCLUSIONS: There is no significant difference in demographic characteristics, mortality or ventilatory parameters between PMO and CMBND. PEEP dynamics were not predictive of impending PM;however, PIP increases did portend development of PM in a large subset of patients. PIP increases of greater than 40% over intubation baseline observed for a median of 8 days were associated with the development of PM. Careful attention to progressive PIP increases over several days should enhance suspicion for PM. The relationship between PIP dynamics and COVID-19 associated PM warrants further investigation.