A descriptive, retrospective single-centre study of air-leak syndrome in intensive care unit patients with COVID-19

Acute respiratory failure is the predominant presentation of intensive care unit (ICU) patients with COVID-19, and lung protective strategies are recommended to mitigate additional respiratory complications such as air-leak syndrome. The aim of this study is to investigate the prevalence, type, and timing of air-leak syndrome with regards to associated factors and patient outcome in patients with COVID-19 in ICUs at a large Swedish emergency hospital.

This retrospective study included all adult patients admitted to an ICU for COVID-19-related respiratory failure at Södersjukhuset between March 6, 2020, and June 6, 2021. Primary outcomes were proportion of patients diagnosed with air-leak syndrome and its different types of manifestations, and timing of diagnoses in relation to ICU admission and initiation of invasive ventilation. Secondary outcomes included the highest level of respiratory support prior to the diagnosis of air-leak syndrome, patient characteristics and treatment variables associated with air-leak syndrome, and 90-day mortality for patients with air-leak syndrome compared to those without.

Out of a total of 669 patients, 81 (12%) were diagnosed with air-leak syndrome. Air-leak syndrome manifested as pneumomediastinum (PMD) (n = 58, 72%), pneumothorax (PTX) (n = 43, 53%), subcutaneous emphysema (SCE) (n = 28, 35%) and pneumatocele (PC) (n = 4, 4.9%). Air-leak syndrome was diagnosed at a median of 14 days (IQR 6-22) after ICU admission and 12 days (IQR 6-19) following the initiation of invasive ventilation. The highest respiratory support prior to diagnosis was invasive ventilation (IV) in 64 patients (79%), non-invasive ventilation in two patients (2.5%), and low- or high-flow oxygen in 15 patients (19%). Multiple logistic regression showed that pulmonary disease at baseline (OR 1.87, 95% CI 1.07-3.25), a lower body mass index (OR 0.95, 95% CI 0.9-0.99), admission later compared with earlier in the pandemic (OR 3.89, 95% CI 2.14-7.08), and IV (OR 3.92, 95% CI 2.07-7.44) were associated with an increased risk of air-leak syndrome. Compared with patients not diagnosed with air-leak syndrome, patients with air-leaks had a higher mortality at 90 days after ICU admission, 46% versus 26% (p <.001). However, the mortality rate differed with different air-leak manifestations, 47% for PMD, 47% for PTX, 50% for the combination of both PMD and PTX and 0% in patients with only SCE and/or PC, respectively.

In 669 ICU patients with COVID-19, 12% had one or more manifestations of air-leak syndrome. Notably, PMD, rather than PTX, was the most common manifestation, suggesting a potentially distinctive feature of COVID-19-related air-leak syndrome. Further research is needed to determine whether COVID-19 involves different pathophysiological or iatrogenic mechanisms compared with other critical respiratory conditions.

Clinicaltrials.gov, identifying number, NCT05877443.

This single-centre cohort study of air leakage into soft tissue in ventilated COVID cases presents findings for associated factors and clinical manifestations, including with different COVID-19 periods and treatments.