Pneumothorax in COVID-19: A Negative Prognostic Indicator for In-hospital Mortality

We aimed to evaluate differences in mortality between Coronavirus 2019 (COVID-19) patients without pneumothorax, with pneumothorax, and with pneumothorax requiring intervention. Secondarily, we ascertained whether size or method of intervention for pneumothorax provides insight into the poor prognosis associated with PTX.

Data from adult patients from a large integrated health system constituting 21 medical centers whose first hospitalization for COVID-19 occurred between January 3, 2020 and December 31, 2022 were collected and divided into those with PTX and those without. Those with PTX were further subdivided into intervention (PTX_I), defined as pigtail catheter or tube thoracostomy placement, or no intervention (PTX_O). Baseline characteristics, comorbidities, vaccination status, intensive care unit (ICU) admission, and ventilatory assistance were compared among groups. Multivariable logistic regression analysis was performed to evaluate variable associations with in-hospital mortality. Propensity score analysis was conducted to adjust for potentially confounding factors between those receiving intervention versus no intervention.

We identified 38,107 patients. Incidence rate of PTX was 1.2% (487/38,107). Among patients with PTX, 78.8% (384/487) received intervention. Length of stay (LOS) was the highest in the PTX_I group, second highest in the PTX_O group, and lowest in the group without PTX (31.0 d, 22.2 d, and 6.8 d, respectively, P < 0.01). ICU admission rate was the highest in the PTX_I group, lower in the PTX_O group, and lowest in the group without PTX (353/384 or 91.9%, 73/103 or 70.9%, and 6324/37,620 or 16.8%, respectively, P < 0.001). In-hospital mortality rate was highest in the PTX_I group, lower PTX_O group, and lowest in the group without PTX (68.5% and 263/384, 63.1% and 65/103, 8.4% and 3143/37,620, respectively, P < 0.001). While 97% of patients with large pneumothoraces received intervention, only 50% of patients with small pneumothoraces received intervention. Small pneumothoraces were associated with worse in-hospital mortality (P = 0.04), but this difference was not seen in medium or large pneumothoraces. Large bore thoracostomy tube (>20 Fr) placement was associated with a higher rate of in-hospital mortality (adjusted odds ratio 1.96, 95% confidence interval 1.14-3.40, P = 0.016). Logistic regression analysis of all COVID-19 patients revealed that when adjusting for age, sex, comorbidities, obesity, vaccination status, sepsis, ventilatory assistance, and ICU admission, PTX was an independent predictor (adjusted odds ratio = 3.86, 95% confidence interval : 2.99-5.00; P < 0.001) of in-hospital mortality. Propensity score analysis showed that pneumothorax with intervention was associated with lower mortality rate compared with no intervention (P < 0.01).

Pneumothorax is an important negative prognostic indicator in COVID-19 as its presence is associated with increased in-hospital mortality. Intervention for pneumothorax is associated with improved in-hospital mortality. Further investigation into type of intervention for COVID-19 PTX, namely pigtail catheter or large bore thoracostomy tube, is warranted to clarify the association of larger bore tubes with higher in-hospital mortality.