Pneumothorax and barotrauma in invasively ventilated patients with COVID-19.

Dear Editor,

We read with great interest the article by Chopra et al. recently published on Respiratory Medicine [1].

In their multicenter study, the Authors present incidence, clinical characteristics and outcome of pneumothorax in critically ill patients with coronavirus disease 2019 (COVID-19). They found that pneumothorax rate among invasively ventilated COVID-19 patients is 80/594 (13 %), and mortality rate for patients who developed pneumothorax is 50/80 (62 %). On the contrary, among 160 randomly selected invasively ventilated COVID-19 patients without pneumothorax, mortality was 78/160 (49 %).

We recently published a systematic review on rate of barotrauma among invasively ventilated COVID-19 patients [2]. Our pooled analysis included data from 13 studies and 1814 patients [[3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]] and we found that pooled estimate of pneumothorax rate was 10.7 % (95 % confidence interval [CI] = 6.7 %–14.7 %), while overall rate of barotrauma event was 16.1 % (95 % CI = 11.8 %–20.4 %). In addition, we found an overall mortality rate for COVID-19 patients who developed barotrauma of 61.6 % (95 % CI = 50.2 %–73.0 %), as compared with a mortality of 49.5 % (95 % CI = 41.1 %–58.0 %) for COVID-19 patients who did not develop barotrauma.

We are pleased to read that our findings are further reinforced by another study, that confirms that pneumothorax rate among mechanically ventilated COVID-19 patients is between 10 and 15 %. Indeed, after including the study by Chopra et al. in an updated meta-analysis, the pooled pneumothorax rate is 11.1 % (95 % CI = 7.4 %–14.8 %) (Fig. 1 ), while the overall barotrauma rate is 15.4 % (95 % CI = 11.7 %–19 %) (Fig. 2 ). Furthermore, the study by Chopra et al. also confirms a mortality rate above 60 % for COVID-19 patients with barotrauma (updated pooled estimate = 61.4 %; 95 % CI = 52.1 %–70.7 %) (Fig. 3 ), as compared with a mortality lower than 50 % for patients who did not develop barotrauma (updated pooled estimate = 49.5 %; 95 % CI = 42.8 %–52.3 %) (Fig. 4 ).

Fig. 1

Forest plot for pneumothorax development in invasively ventilated COVID-19 patients.

Fig. 2

Forest plot for barotrauma development in invasively ventilated COVID-19 patients.

Fig. 3

Forest plot for longest follow-up mortality in invasively ventilated COVID-19 patients who developed barotrauma.

Fig. 4

Forest plot for longest follow-up mortality in invasively ventilated COVID-19 patients who did not developed barotrauma.

Collectively, these data confirm that barotrauma occur frequently in COVID-19 patients requiring mechanical ventilation, and is associated with a worse outcome and a very high mortality risk. Interestingly, the study by Chopra et al. is one of the few studies that identified worse lung mechanics at start of mechanical ventilation as a risk factor for development of barotrauma [2]. Most of previously published studies reported no significant differences among mechanical ventilation settings/parameters between patients who developed barotrauma and those who did not. They also found a trend towards lower age and higher use of steroids in pneumothorax patients, which were also suggested by other Authors [2,16].

Interestingly, in a recent study by our group, we identified Macklin-like radiological sign [17,18] detected on chest computed tomography (CT) scan as potential predictor of subsequent development of barotrauma about 12 days in advance [3].

Considering the high mortality rate associated with development of barotrauma in COVID-19 patients, and the ongoing debate on optimal timing of intubation in these patients [[19], [20], [21]], we believe that it might be justified to avoid intubation in patients with Macklin-like radiological sign on chest CT, and prefer early support with alternative techniques including awake prone positioning and extracorporeal membrane oxygenation [[22], [23], [24], [25]].

Funding

None.

Declaration of competing interest

None.