Incidence and risk factors of COVID-19 associated pneumothorax

BACKGROUND: Pneumothorax has been increasingly observed among patients with coronavirus disease-2019 (COVID-19) pneumonia, specifically in those patients who develop acute respiratory distress syndrome (ARDS). In this study, we sought to determine the incidence and potential risk factors of pneumothorax in critically ill adults with COVID-19. METHOD: This retrospective cohort study included adult patients with laboratory-confirmed SARS-CoV-2 infection admitted to one of the adult intensive care units of a tertiary, academic teaching hospital from May 2020 through May 2021. RESULTS: Among 334 COVID-19 cases requiring ICU admission, the incidence of pneumothorax was 10% (33 patients). Patients who experienced pneumothorax more frequently required vasopressor support (28/33 [84%] vs. 191/301 [63%] P = 0.04), were more likely to be proned (25/33 [75%] vs. 111/301 [36%], P<0.001), and the presence of pneumothorax was associated with prolonged duration of mechanical ventilation; 21 (1–97) versus 7 (1–79) days, p<0.001 as well as prolonged hospital length of stay (29 [9–133] vs. 15 [1–90] days, P<0.001), but mortality was not significantly different between groups. Importantly, when we performed a Cox proportional hazard ratio (HR) model of multivariate parameters, we found that administration of tocilizumab significantly increased the risk of developing pneumothorax (HR = 10.7; CI [3.6–32], P<0.001). CONCLUSION: Among 334 critically ill patients with COVID-19, the incidence of pneumothorax was 10%. Presence of pneumothorax was associated with prolonged duration of mechanical ventilation and length of hospital stay. Strikingly, receipt of tocilizumab was associated with an increased risk of developing pneumothorax.