
==== Front
Sci Rep
Sci Rep
Scientific Reports
2045-2322
Nature Publishing Group UK London

39256438
69266
10.1038/s41598-024-69266-x
Article
Comparison of the risk of pneumothorax in COVID-19 and seasonal influenza
Song Myung Jin 1
Kang Minsun 2
Song Kyoung-Ho 1
Kim Hong Bin 1
Kim Eu Suk 1
Jung Jaehun 2
Lim Sung Yoon nucleon727@snu.ac.kr

1
1 https://ror.org/00cb3km46 grid.412480.b 0000 0004 0647 3378 Department of Internal Medicine, Seoul National University College of Medicine and Seoul National University Bundang Hospital, 82 Gumi-ro, Bundang-gu, Seongnam-si, 13620 Republic of Korea
2 https://ror.org/03ryywt80 grid.256155.0 0000 0004 0647 2973 Department of Preventive Medicine, Gachon University College of Medicine, 38 Dokjeom-ro 3-beon-gil, Namdong-gu, Incheon, 21565 Republic of Korea
10 9 2024
10 9 2024
2024
14 210772 6 2024
2 8 2024
© The Author(s) 2024
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Limited evidence exists regarding the link between coronavirus disease 2019 (COVID-19) and pneumothorax. Therefore, we aimed to evaluate the occurrence rate of pneumothorax in hospitalized patients with COVID-19 and compare the risk of pneumothorax between patients with COVID-19 and influenza. This retrospective cohort study used patient data from the National Health Insurance Service of South Korea. Patients diagnosed with COVID-19 (December 2019 to December 2021) and influenza (January 2019 to December 2021) who required hospitalization and respiratory support were included. We identified 46,460 patients with COVID-19 and 6,117 with influenza. The occurrence rate of pneumothorax was 0.74% in patients with COVID-19. In an inverse probability of treatment weighting matched cohort, the Cox proportional hazards regression model showed that COVID-19 was not associated with an increased risk of pneumothorax compared to influenza (hazard ratio, 1.22; 95% confidence interval, 0.75–1.99). However, the risk of pneumothorax associated with COVID-19 compared to influenza was significantly higher in patients without chronic lung disease than in those with (P for heterogeneity = 0.037). In conclusion, COVID-19, compared with influenza, is not associated with an increased risk of pneumothorax; however, it is associated with an increased risk in patients without chronic lung disease.

Keywords

COVID-19
Influenza
Pneumonia
Pneumothorax
Chronic lung disease
Subject terms

Diseases
Medical research
Risk factors
http://dx.doi.org/10.13039/501100003725 National Research Foundation of Korea NRF-2021R1A5A2030333 Jung Jaehun issue-copyright-statement© Springer Nature Limited 2024
==== Body
pmcIntroduction

Pneumothorax describes the presence of air in the pleural space. Pneumothorax is more common in patients with underlying respiratory disorders such as chronic obstructive pulmonary disease (COPD), emphysema, and lung infections, resulting from communication between the damaged alveolar spaces and the pleura1.

As of March 2023, the coronavirus disease 2019 (COVID-19) pandemic resulted in over 670 million cases and approximately 6.8 million deaths worldwide2. During the COVID-19 pandemic, pneumothorax has been reported as a frequent complication in COVID-19 pneumonia, which is associated with an increased risk of mortality3–5. The occurrence rate of pneumothorax in patients with COVID-19 varies considerably across studies, ranging from 5–24% in mechanically ventilated patients6 and 0.7–1% in hospitalized patients7,8.

Although several studies have highlighted frequent pneumothorax in patients with COVID-19, comparisons with relevant controls are limited and have reported inconsistent results9–13. Additionally, multiple cases of pneumothorax have been documented in patients with influenza, a virus that, like COVID-19, primarily causes respiratory tract infections14. Consequently, it remains unclear whether the increased susceptibility to pneumothorax in patients with COVID-19 reflects a specific trait of COVID-19 or is merely a consequence of the very high prevalence of pneumonia during the short period of the pandemic.

To address this unresolved issue, we leveraged nationwide claims data from South Korea. We selected influenza pneumonia as a harmonized control group to compare with COVID-19. Our aim was to assess the occurrence rate of pneumothorax in patients hospitalized with COVID-19 pneumonia and determine whether COVID-19 pneumonia influences the risk of pneumothorax compared to influenza pneumonia.

Results

Study population and baseline characteristics

During the study period, 325,990 patients were hospitalized for COVID-19, of whom 46,477 (14.3%) received respiratory support. After excluding 17 patients who did not meet inclusion criteria, the remaining eligible 46,460 patients formed the COVID-19 group (mean [standard deviation, SD] age, 62.5 [17.0] years; 24,597 males [52.9%]). A total of 90,330 patients with influenza were hospitalized during the study period, and 6,118 received respiratory support. After excluding one patient who did not meet the inclusion criteria, the remaining 6,117 patients formed the influenza group (mean [SD] age, 73.0 [15.3] years; 2,801 males [45.8%]) (Fig. 1 and Table 1).Figure 1 Study flow. COVID-19, coronavirus disease 2019; IPTW, inverse probability of treatment weighted.

Table 1 Baseline characteristics of overall and inverse probability of treatment weighted cohorts.

Variables	Before IPTW, No. (%)	After IPTW, No. (%)	
COVID-19 (n=46,460)	Influenza (n=6,117)	SMD	COVID-19 (n=52,697)	Influenza (n=49,295)	SMD	
Age, year (mean, SD)	62.5 (17.0)	72.95 (15.3)	0.65	63.8 (18.3)	64.7 (49.4)	0.03	
Sex, male	24,597 (52.9)	2801 (45.8)	0.14	27,426 (52.0)	25,799 (52.3)	0.01	
Residence			0.65			0.07	
 Seoul (capital city)	15,867 (34.2)	843 (13.8)		16,702 (31.7)	14,334 (29.1)		
 Other metropolitan city	22,686 (48.8)	2676 (43.8)		25,390 (48.2)	24,139 (49.0)		
 Other area	7,907 (17.0)	2598 (42.5)		10,605 (20.1)	10,822 (22.0)		
Income level			0.20			0.06	
 Q1 (Lowest 20%)	7,461 (16.1)	927 (15.2)		8,430 (16.0)	8,341 (16.9)		
 Q2 (21–50%)	9,706 (20.9)	1008 (16.5)		10,725 (20.4)	9,972 (20.2)		
 Q3 (51–80%)	12,521 (27.0)	1546 (25.3)		14,077 (26.7)	13,150 (26.7)		
 Q4 (Highest 20%)	12,635 (27.2)	1817 (29.7)		14,464 (27.5)	12,537 (25.4)		
 Unknown	4,137 (8.9)	819 (13.4)		5,001 (9.5)	5,295 (10.7)		
Comorbidities	
 Diabetes	15,703 (33.8)	2745 (44.9)	0.23	18,535 (35.2)	17,538 (35.6)	0.01	
 Hypertension	22,493 (48.4)	4014 (65.6)	0.35	26,643 (50.6)	25,748 (52.2)	0.03	
 Myocardial infarction	943 (2.0)	250 (4.1)	0.12	1,213 (2.3)	1,363 (2.8)	0.03	
 Congestive heart failure	4,092 (8.8)	1269 (20.8)	0.34	5,462 (10.4)	6,034 (12.2)	0.06	
 Cerebrovascular disease	6,855 (14.8)	1707 (27.9)	0.33	8,698 (16.5)	9,648 (19.6)	0.08	
 Chronic pulmonary Disease	12,173 (26.2)	3865 (63.2)	0.80	16,172.1 (30.7)	16,590 (33.7)	0.06	
 Chronic liver disease	6,173 (13.3)	893 (14.6)	0.04	7,076 (13.4)	6,543 (13.3)	0.00	
 Chronic renal disease	3,633 (7.8)	768 (12.6)	0.16	4,450 (8.4)	4,792 (9.7)	0.04	
Charlson comorbidity index			0.47			0.11	
 0–2	37,165 (80.0)	3591 (58.7)		40,706 (77.3)	35,920 (72.9)		
 3–5	7,583 (16.3)	2082 (34.0)		9,781 (18.6)	10,521 (21.3)		
 >6	1,712 (3.7)	444 (7.3)		2,210 (4.2)	2,854 (5.8)		
Type of hospital			0.08			0.00	
 Tertiary hospital	40,691 (87.6)	5182 (84.7)		46,015 (87.3)	43,010 (87.3)		
 General hospital	5,769 (12.4)	935 (15.3)		6,682 (12.7)	6,285 (12.8)		
Use of corticosteroid	13,630 (29.3)	1040 (17.0)	0.30	14,630 (27.8)	11,278 (22.9)	0.11	
Highest level of respiratory support during hospital admission	
 Conventional Oxygen therapy	36,292 (78.1)	5444 (89.0)	0.30	41,834 (79.4)	37,617 (76.3)	0.07	
 High Flow Nasal Cannula	6,804 (14.6)	322 (5.3)	0.32	7,131 (13.5)	7,056 (14.3)	0.02	
 Mechanical ventilation	3,009 (6.5)	336 (5.5)	0.04	3,361 (6.4)	4,027 (8.2)	0.07	
 ECMO	355 (0.8)	15 (0.3)	0.07	371 (0.7)	595 (1.2)	0.05	
Data are presented as number (%) unless otherwise specified.

COVID-19, coronavirus disease; ECMO, extracorporeal membrane oxygenation; IPTW, inverse probability of treatment weighted; SMD, standardized mean difference.

The COVID-19 group consisted of younger individuals with fewer comorbidities than the influenza group. Notably, the prevalence of chronic lung disease in the COVID-19 group was lower compared to the influenza group. Corticosteroid use was more common in the COVID-19 group. Conventional oxygen therapy was less frequent as the highest level of respiratory support during hospitalization, while high-flow nasal cannulation was more frequent in patients with COVID-19 than in those with influenza. After adjusting for inverse probability of treatment weighting (IPTW), the cohort showed a good ability to balance the measured confounders between the COVID-19 and influenza groups, with most standardized mean differences (SMDs) < 0.1 after weighting, except for two variables: Charlson comorbidity index (CCI) and the use of corticosteroids (Table 1).

Development of pneumothorax

Before IPTW matching, the occurrence rate of pneumothorax was 0.74% (343/46,460) in patients with COVID-19 and 0.49% (30/6,117) in those with influenza. The occurrence rates of pneumothorax based on the level of respiratory support are presented in Supplementary Table S1.

After IPTW matching, the occurrence rate of pneumothorax was significantly higher in the COVID-19 group compared to the influenza group (0.72% [378/52,697] vs. 0.58% [286/49,295]; P = 0.007). Pneumothorax was more prevalent in the COVID-19 group than in the influenza group at all support levels (Table 2). Table 2 Primary outcome and outcomes of interest in IPTW-adjusted cohort.

Variables	COVID-19	Influenza	P value	
In total IPTW-adjusted cohort	n = 52,697	n = 49,295		
 Incidence of pneumothorax	378 (0.72)	286 (0.58)	0.007	
 Incidence of pneumothorax according to the respiratory support	
  Conventional Oxygen therapy	25/41834 (0.06)	43/37617 (0.11)	0.009	
  High Flow Nasal Cannula	51/7131 (0.72)	24/7056 (0.34)	0.002	
  Mechanical ventilation	238/3361 (7.08)	203/4027 (5.04)	<0.001	
  ECMO	65/371 (17.52)	16/595 (2.69)	<0.001	
 In-hospital mortality	3744 (7.1%)	2428 (4.9%)	<0.001	
 Length of hospital stay, day, median (IQR)	11 (3–16)	6 (3–10)	<0.001	
Among those who developed pneumothorax	n = 378	n = 286		
 Duration of any respiratory support before pneumothorax, day, median (IQR)	0 (0–1)	0 (0–1)	0.112	
Data are presented as number (%) unless otherwise specified.

COVID-19, coronavirus disease 2019; ECMO, extracorporeal membrane oxygenation; IPTW, inverse probability of treatment weighted.

However, the univariable Cox proportional hazard model, which accounts for the time elapsed until the pneumothorax occurrence, found no significant association between COVID-19 and the risk of pneumothorax (hazard ratio [HR], 1.22; 95% confidence interval [CI], 0.75–1.99, P = 0.414, Table 3). Among the patients who developed pneumothorax, the median duration of respiratory support before the development of pneumothorax was 0 days (interquartile range [IQR], 0–1) in both groups, showing that respiratory support and the development of pneumothorax occurred on the same day in most patients. Table 3 Cox proportional hazard model for pneumothorax in the IPTW-adjusted cohort.

Variables	Univariable analysis	
HR (95% CI)	P value	
COVID-19 (vs Influenza)	1.22 (0.75–1.99)	0.414	
Age, year	1.02 (1.00–1.03)	0.043	
Sex, male	1.60 (1.06–2.42)	0.027	
Residence	
  Other metropolitan city (vs. capital city)	0.91 (0.56–1.47)	0.701	
  Other area (vs. capital city)	1.14 (0.70–1.85)	0.605	
Income level	
  Q2 (vs Q1)	1.33 (0.73–2.44)	0.352	
  Q3 (vs Q1)	0.93 (0.57–1.49)	0.751	
  Q4 (vs Q1)	1.255 (0.754–2.09)	0.382	
  Unknown (vs Q1)	1.84 (0.70–4.82)	0.217	
Comorbidities	
  Diabetes	1.47 (0.97–2.22)	0.072	
  Hypertension	1.42 (0.89–2.28)	0.140	
  Myocardial infarction	0.95 (0.53–1.70)	0.857	
  Congestive heart failure	1.46 (0.86–2.49)	0.166	
  Cerebrovascular disease	1.03 (0.66–1.59)	0.908	
  Chronic pulmonary Disease	1.47 (0.97–2.24)	0.069	
  Chronic liver disease	0.92 (0.63–1.35)	0.672	
Chronic renal disease	1.17 (0.78–1.76)	0.451	
Charlson comorbidity index	
  3–5 (vs 0–2)	1.15 (0.77–1.71)	0.496	
  >6 (vs 0–2)	1.25 (0.67–2.34)	0.479	
Type of hospital	
  General hospital (vs tertiary hospital)	0.52 (0.28–0.96)	0.037	
  Use of corticosteroid	1.06 (0.67–1.67)	0.800	
Type of Respiratory support	
  High Flow Nasal Cannula (vs COT)	0.76 (0.44–1.32)	0.331	
  Mechanical ventilation (vs COT)	24.13 (15.99–36.41)	<.001	
  ECMO (vs COT)	12.84 (5.91–27.90)	<.001	
CI, Confidence interval; COT, Conventional Oxygen therapy; COVID-19, coronavirus disease 2019; ECMO, extracorporeal membrane oxygenation.

Subgroup analyses

The associations of pneumothorax with COVID-19 compared to influenza were consistent across all subgroups but showed significant heterogeneity based on the presence of chronic lung disease (Fig. 2). The risk of pneumothorax associated with COVID-19, compared to that associated with influenza, was significantly higher in patients without chronic lung disease than in those with chronic lung disease (HR, 2.51; 95% CI, 1.25–5.04 vs. 1.05, 95% CI, 0.58–1.93, P for heterogeneity = 0.037).Figure 2 Subgroup analysis for pneumothorax in IPTW-adjusted cohort. IPTW, inverse probability of treatment weighted; ECMO, extracorporeal membrane oxygenation; CCI, Charlson comorbidity index.

In-hospital mortality

Before IPTW matching, in-hospital mortality was 5.00% (2,323/46,460) in patients with COVID-19 and 4.07% (249/6,117) in those with influenza (Supplementary Table S1). After IPTW matching, in-hospital mortality was significantly higher in the COVID-19 group compared to the influenza group (7.1% [3,744/52,697] vs. 4.9% [2,428/49,295]; P < 0.001) (Table 2 and Supplementary Figure S1). The univariable Cox proportional hazards model showed no association between COVID-19 and in-hospital mortality compared to influenza (HR, 0.92 [95% CI, 0.75–1.14]; P = 0.460, Supplementary Table S2).

Delta variant and development of pneumothorax in COVID-19 patients

To assess the association between the delta variant and the development of pneumothorax in patients with COVID-19, an unweighted Cox proportional hazards analysis was conducted on the initial study population with COVID-19 before IPTW matching (n = 46,460). The multivariable Cox proportional hazards model, controlling for confounding factors influencing pneumothorax, found no significant association between the delta variant and the development of pneumothorax (adjusted HR, 0.91 [95% CI, 0.72–1.14]; P = 0.403, Supplementary Table S3).

Discussion

In this retrospective cohort study using nationwide claims data from South Korea, the occurrence rate of pneumothorax, which required closed thoracostomy in hospitalized COVID-19 patients needing respiratory support, was 0.74%. In the IPTW-matched cohort, the rate of pneumothorax was significantly higher in COVID-19 patients compared to those with influenza. However, the Cox proportional hazards regression model indicated that COVID-19 was not linked to an increased risk of pneumothorax compared to influenza. In a subgroup analysis, COVID-19 was associated with a higher risk of pneumothorax in patients without chronic lung disease, compared to influenza.

Multiple studies have reported that COVID-19 may increase the susceptibility to alveolar rupture, leading to pneumothorax, pneumomediastinum, or subcutaneous emphysema. However, studies reporting the occurrence rate of barotrauma, including pneumothorax, pneumomediastinum, and subcutaneous emphysema, have focused solely on patients with COVID-19, lacking a relevant control group. Studies have reported a higher occurrence rate of pneumothorax compared to the known 5–8% in non-COVID-19 acute respiratory disease syndrome (ARDS) 15–17. Moreover, only a few studies have directly compared the occurrence rate of barotrauma between COVID-19 patients and controls9,10,13,18. Three studies utilized non-COVID-19 ARDS cases from the pre-pandemic period as a control group for COVID-19 ARDS9,10,13 and one study utilized patients without COVID-19 who visited the emergency department, irrespective of the reason for the visit as a control group for patients who were confirmed to have COVID-19 in the emergency department18. Three studies found that the occurrence rate of barotrauma was significantly higher in patients with COVID-19 compared to those without COVID-199,10,18. In contrast, one study reported no significant difference in the occurrence rate of barotrauma between the two groups13. Thus, the results of previous studies are, inconsistent. Furthermore, the non-COVID-19 patient group used as a control was notably heterogeneous. This raises questions about the appropriateness of control groups in such studies. Therefore, we selected patients who required respiratory support due to influenza as a control group to evaluate whether COVID-19 increases the risk of pneumothorax.

Our results showed that COVID-19 was not linked to a higher risk of pneumothorax compared to influenza among general hospitalizations requiring respiratory support. However, in patients without pre-existing chronic lung disease, COVID-19 was associated with an increased risk of developing pneumothorax. Chronic lung diseases, such as COPD, asthma, and emphysema, carry a high risk of spontaneous pneumothorax due to underlying alveolar damage19. While COVID-19 does not elevate the risk of pneumothorax in patients with pre-existing chronic lung disease, who already have an elevated risk, it does increase the risk in those without chronic lung disease. Consistent with our results, a prospective multicenter observational study of COVID-19 patients found that none of the 21 patients with spontaneous pneumomediastinum had COPD, whereas among 528 patients with spontaneous pneumomediastinum, 38 (7.2%) had COPD20. Additionally, Cai et al. found that chronic bronchitis and COPD were more prevalent in the group without pneumothorax or pneumomediastinum compared to the group with pneumothorax or pneumomediastinum group (4 (7.1%) vs. 13 (15.3%), P < 0.001)21.

The increased risk of pneumothorax in COVID-19 patients can be attributed to self-inflicted lung injury from repeated excessive inspiratory efforts, which raise transpulmonary pressure more in COVID-19 patients than in non-COVID patients20,22–24. The exact pathophysiological mechanism behind the increased risk of pneumothorax in COVID-19 patients without chronic lung disease remains uncertain. However, it is plausible that patients without chronic lung disease, who generally have stronger inspiratory efforts, are more susceptible to this phenomenon in the context of COVID-19.

In general, the occurrence of pneumothorax is lower in patients without chronic lung disease; however, this study demonstrated that in COVID-19 patients without chronic lung disease, the risk of pneumothorax is higher compared to those with influenza. Medical professionals should be aware of this elevated risk and engage in vigilant monitoring for the early detection and management of pneumothorax, especially in COVID-19 patients without chronic lung disease. Additionally, greater care should be given to implementing lung-protective ventilation strategies in this subgroup to prevent the occurrence of pneumothorax.

In hospitalized patients with COVID-19 who require supplemental oxygen, systemic corticosteroid therapy improves clinical outcomes, presumably by mitigating systemic inflammatory responses, as recommended by the guidelines25,26. Conversely, in patients with influenza, corticosteroid use is known to contribute to increased mortality by hindering viral clearance27. Nevertheless, 17% of patients with influenza in our study were treated with corticosteroids before IPTW matching. Given the significant number of individuals with chronic lung disease (26.2% vs. 63.2%) in the influenza group, corticosteroids were presumably used to treat exacerbations of underlying lung diseases, such as asthma or chronic obstructive pulmonary disease. Due to their role in the inflammatory and healing processes, corticosteroids are associated with an increased risk of pneumothorax and prolonged air leak after developing pneumothorax28,29. Considering corticosteroids as drugs related to pneumothorax occurrence, we conducted a subgroup analysis based on corticosteroid use and found no significant interaction between COVID-19 and the development of pneumothorax according to corticosteroid use.

This study had several strengths. First, utilizing South Korea’s insurance claims data enabled the enrolment of all patients hospitalized for COVID-19 or influenza who required respiratory support during the study period. This comprehensive approach ensures a broad and representative sample size. Second, we significantly reduced confounding factors by selecting influenza-infected patients as the control group for COVID-19 patients requiring oxygen treatment, thereby enhancing comparability between the groups.

Nevertheless, this study also had some limitations. First, pneumothorax was defined based on the performance of closed thoracostomy. Cases in which a closed thoracostomy was not required were not considered; therefore, the incidence might have been underestimated. However, clinically significant cases involving air leaks that affect the prognosis typically require a closed thoracostomy. Thus, our definition encompassed the most clinically relevant instances of pneumothorax. Second, our study did not have information on ventilator settings, which could have contributed to an air leak by increasing the stress on the alveoli. However, previous studies have noted that most patients with COVID-19 pneumonia who develop pneumothorax experience this complication despite receiving lung protective ventilation9,30,31. Third, potential intermediate factors in the causal pathway between exposure and outcome, such as maximal respiratory support and corticosteroid use, were included in the IPTW matching, which may introduce bias. Fourth, the respiratory support level used as a variable in IPTW matching was classified into four categories, and the low proportion of patients receiving higher levels of respiratory support could lead to overfitting. Finally, the absence of radiologic data and specific measures of respiratory dynamics in our analysis represents a limitation. Further research should focus on collecting detailed imaging findings and quantifying respiratory parameters to elucidate the mechanisms leading to pneumothorax in individuals with COVID-19.

In conclusion, COVID-19, compared to influenza, was not associated with an increased risk of pneumothorax among hospitalized patients requiring respiratory support in general. However, our findings suggest that, in patients without pre-existing chronic lung disease, COVID-19 is associated with an increased risk of pneumothorax compared with influenza.

Methods

Study design and population

This retrospective cohort study used claims data provided by the National Health Insurance Service (NHIS) in South Korea. South Korea has implemented mandatory nationwide health insurance that covers 97% of its residents, with premiums adjusted based on income or property value. The remaining 3% residents who cannot afford premiums are covered by government-funded medical aid32. Therefore, using NHIS data, we analyzed almost the entire South Korean population.

We defined patients with COVID-19 as those assigned with the International Classification of Diseases, 10th revision (ICD-10) code U07.1 between December 2019 and December 2021. Patients with influenza were defined as those diagnosed with influenza using ICD-10 codes J09–J11 between January 2019 and December 2021. All consecutive adult patients (age ≥ 19 years) who were hospitalized for COVID-19 or influenza for at least 2 days with any respiratory support were included. Individuals who experienced pneumothorax within 30 days before COVID-19 or influenza infection were excluded (Fig. 1).

Outcomes and variables

Our primary outcome was the development of pneumothorax requiring closed thoracostomy. Although the diagnosis of pneumothorax determined by the ICD codes is likely underreported, reimbursable procedures are precisely identifiable in the claims data. Therefore, we confined the outcomes to cases of pneumothorax that necessitated closed thoracostomy, which is a reimbursable procedure to treat pneumothorax. Data on in-hospital mortality, length of hospital stay, and duration of any respiratory support before pneumothorax were also collected as outcomes of interest.

The following covariates were extracted: demographic information (age and sex), socioeconomic status (income level and residence), CCI calculated using the registered ICD-10 codes from 1 year before the diagnosis of COVID-19 or influenza, type of hospital, corticosteroid use, and level of respiratory support. We delineated two distinct periods to assess the impact of the COVID-19 variants: before the dominance of the delta variant in Korea (June 14, 2021) and thereafter (June 15, 2021).

Statistical analysis

Baseline characteristics are presented as means (SD) or medians (IQR) for continuous variables and as numbers (percentages) for categorical variables. We used IPTW analysis to adjust for potential confounding factors between COVID-19 and influenza pneumonia, which has an advantage over propensity score matching in that it retains more individuals (thus increasing its power) and estimates hazards with less bias33,34. The IPTW was calculated by considering the reciprocal probability of being in a group through logistic regression using the following covariates: age, sex, residence, income, comorbidities, CCI, type of hospital, use of corticosteroids, and level of respiratory support. Weight stabilization was used to reduce the impact of extreme weights. We assessed the balance of covariates before and after IPTW matching using absolute SMD and specified an SMD of ≤ 0.1 as a negligible difference between the two groups35.

Weighted Cox proportional hazards regression was used to estimate the association between COVID-19, compared with influenza, and the development of pneumothorax and in-hospital mortality. We examined the proportional hazards assumption using Schoenfeld residuals for each covariate and did not observe any relevant violations. As weighting creates a pseudo-population containing replications of individuals, we used a robust sandwich variance estimator to correctly estimate the variance and CIs for the effect36.

We performed a subgroup analysis to evaluate the association between COVID-19 (vs. influenza) and the development of pneumothorax in subgroups defined by sex, age (< 65 or ≥ 65 years), chronic lung disease, CCI (0–2, ≥ 3), use of corticosteroids, and the various respiratory support modalities. Heterogeneity was tested across the subgroups. We used unweighted Cox proportional hazard regression to analyze the data of patients with COVID-19 before IPTW adjustment to evaluate how the delta variant of COVID-19 affected the likelihood of pneumothorax.

All reported p-values were 2-tailed, and p < 0.05 was considered statistically significant. Data analyses were performed using SAS statistical software (version 9.4; SAS Institute Inc.), and plots were generated using R software 4.0.0 (R Foundation for Statistical Computing).

Ethics

This study was approved by the Institutional Review Board of Seoul National University Bundang Hospital (number: X-2305-826-903) on April 24, 2023. The requirement for informed consent was waived by the Seoul National University Bundang Hospital Institutional Review Board because the data analyses were performed retrospectively using anonymized data from South Korean NHIS database. This study was conducted in accordance with the ethical principles stated by the 1975 Declaration of Helsinki.

Supplementary Information

Supplementary Information.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-69266-x.

Author contributions

MJS: conceptualization, methodology, writing—original draft, writing—review & editing. MSK: conceptualization, methodology, formal analysis, investigation, data curation, visualization. KHS: methodology, investigation, writing—review & editing. HBK: methodology, investigation, writing—review & editing. ESK: conceptualization, writing—review & editing, supervision. JHJ: conceptualization, writing—review & editing, supervision, funding acquisition. SYL: conceptualization, writing—review & editing, supervision.

Funding

Jaehun Jung was supported by National Research Foundation of Korea (Grant No. NRF-2021R1A5A2030333). The funders had no role in the design and conduct of the study; collection, analysis, and interpretation of data; and writing the manuscript.

Data availability

The data supporting the findings of this study are available from the National Health Insurance Service. However, there are restrictions on the availability of these data, which were used under license for the current study and are not publicly accessible. Data can be made available from the corresponding authors upon reasonable request and with permission from the National Health Insurance Service (https://nhiss.nhis.or.kr/bd/ab/bdaba000eng.do).

Competing interests

The authors declare that they have no competing interests. Jaehun Jung was supported by National Research Foundation of Korea (Grant No. NRF-2021R1A5A2030333).

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Myung Jin Song and Minsun Kang.

These authors jointly supervised this work: Eu Suk Kim, Jaehun Jung and Sung Yoon Lim.
==== Refs
References

1. Noppen M De Keukeleire T Pneumothorax Respiration 2008 76 121 127 10.1159/000135932 18708734
Noppen, M. & De Keukeleire, T. Pneumothorax. Respiration 76, 121–127. 10.1159/000135932 (2008).18708734 10.1159/000135932
2. Johns Hopkins University & Medicine: COVID-19 Map - Johns Hopkins Coronavirus Resource Center. https://coronavirus.jhu.edu/map.html. (Accessed 20 Mar, 2024).
3. Bonato M Pneumothorax and/or Pneumomediastinum worsens the prognosis of COVID-19 patients with severe acute respiratory failure: A multicenter retrospective case-control study in the North-East of Italy J. Clin. Med. 2021 10 5 10.3390/jcm10214835
Bonato, M. et al. Pneumothorax and/or Pneumomediastinum worsens the prognosis of COVID-19 patients with severe acute respiratory failure: A multicenter retrospective case-control study in the North-East of Italy. J. Clin. Med. 10, 5. 10.3390/jcm10214835 (2021).10.3390/jcm10214835
4. Akram J Epidemiological and outcome analysis of COVID-19-associated pneumothorax: Multicentre retrospective critical care experience from Qatar BMJ Open 2022 12 e053398 10.1136/bmjopen-2021-053398 35190427
Akram, J. et al. Epidemiological and outcome analysis of COVID-19-associated pneumothorax: Multicentre retrospective critical care experience from Qatar. BMJ Open 12, e053398. 10.1136/bmjopen-2021-053398 (2022).35190427 10.1136/bmjopen-2021-053398
5. Ragnoli B Pneumothorax in hospitalized COVID-19 patients with severe respiratory failure: Risk factors and outcome Respir Med. 2023 211 107194 10.1016/j.rmed.2023.107194 36889518
Ragnoli, B. et al. Pneumothorax in hospitalized COVID-19 patients with severe respiratory failure: Risk factors and outcome. Respir Med. 211, 107194. 10.1016/j.rmed.2023.107194 (2023).36889518 10.1016/j.rmed.2023.107194
6. Belletti A Barotrauma in coronavirus disease 2019 patients undergoing invasive mechanical ventilation: A systematic literature review Crit. Care Med. 2022 50 491 500 10.1097/ccm.0000000000005283 34637421
Belletti, A. et al. Barotrauma in coronavirus disease 2019 patients undergoing invasive mechanical ventilation: A systematic literature review. Crit. Care Med. 50, 491–500. 10.1097/ccm.0000000000005283 (2022).34637421 10.1097/ccm.0000000000005283
7. Zantah M Dominguez Castillo E Townsend R Dikengil F Criner GJ Pneumothorax in COVID-19 disease-incidence and clinical characteristics Respir Res. 2020 21 236 10.1186/s12931-020-01504-y 32938445
Zantah, M., Dominguez Castillo, E., Townsend, R., Dikengil, F. & Criner, G. J. Pneumothorax in COVID-19 disease-incidence and clinical characteristics. Respir Res. 21, 236. 10.1186/s12931-020-01504-y (2020).32938445 10.1186/s12931-020-01504-y
8. Marciniak SJ COVID-19 pneumothorax in the UK: A prospective observational study using the ISARIC WHO clinical characterisation protocol Eur. Respir. J. 2021 10.1183/13993003.00929-2021 34083401
Marciniak, S. J. et al. COVID-19 pneumothorax in the UK: A prospective observational study using the ISARIC WHO clinical characterisation protocol. Eur. Respir. J.10.1183/13993003.00929-2021 (2021).34083401 10.1183/13993003.00929-2021
9. Lemmers DHL Pneumomediastinum and subcutaneous emphysema in COVID-19: barotrauma or lung frailty? ERJ. Open Res. 2020 10.1183/23120541.00385-2020 33257914
Lemmers, D. H. L. et al. Pneumomediastinum and subcutaneous emphysema in COVID-19: barotrauma or lung frailty?. ERJ. Open Res.10.1183/23120541.00385-2020 (2020).33257914 10.1183/23120541.00385-2020
10. McGuinness G Increased incidence of barotrauma in patients with COVID-19 on invasive mechanical ventilation Radiology 2020 297 E252 e262 10.1148/radiol.2020202352 32614258
McGuinness, G. et al. Increased incidence of barotrauma in patients with COVID-19 on invasive mechanical ventilation. Radiology 297, E252-e262. 10.1148/radiol.2020202352 (2020).32614258 10.1148/radiol.2020202352
11. Fiacchini G Evaluation of the incidence and potential mechanisms of tracheal complications in patients with COVID-19 JAMA Otolaryngol. Head Neck Surg. 2021 147 70 76 10.1001/jamaoto.2020.4148 33211087
Fiacchini, G. et al. Evaluation of the incidence and potential mechanisms of tracheal complications in patients with COVID-19. JAMA Otolaryngol. Head Neck Surg. 147, 70–76. 10.1001/jamaoto.2020.4148 (2021).33211087 10.1001/jamaoto.2020.4148
12. Protti A Barotrauma in mechanically ventilated patients with Coronavirus disease 2019: A survey of 38 hospitals in Lombardy, Italy Minerva Anestesiol 2021 87 193 198 10.23736/s0375-9393.20.15002-8 33325217
Protti, A. et al. Barotrauma in mechanically ventilated patients with Coronavirus disease 2019: A survey of 38 hospitals in Lombardy, Italy. Minerva Anestesiol 87, 193–198. 10.23736/s0375-9393.20.15002-8 (2021).33325217 10.23736/s0375-9393.20.15002-8
13. Knox DB Brunhoeber A Peltan ID Brown SM Lanspa MJ Comparison of radiographic pneumothorax and pneumomediastinum in COVID-19 vs. non-COVID-19 acute respiratory distress syndrome Intens. Care Med. 2022 48 1648 1651 10.1007/s00134-022-06816-9
Knox, D. B., Brunhoeber, A., Peltan, I. D., Brown, S. M. & Lanspa, M. J. Comparison of radiographic pneumothorax and pneumomediastinum in COVID-19 vs. non-COVID-19 acute respiratory distress syndrome. Intens. Care Med. 48, 1648–1651. 10.1007/s00134-022-06816-9 (2022).10.1007/s00134-022-06816-9
14. Guo HH Sweeney RT Regula D Leung AN Best cases from the AFIP: Fatal 2009 influenza A (H1N1) infection, complicated by acute respiratory distress syndrome and pulmonary interstitial emphysema Radiographics 2010 30 327 333 10.1148/rg.302095213 20068001
Guo, H. H., Sweeney, R. T., Regula, D. & Leung, A. N. Best cases from the AFIP: Fatal 2009 influenza A (H1N1) infection, complicated by acute respiratory distress syndrome and pulmonary interstitial emphysema. Radiographics 30, 327–333. 10.1148/rg.302095213 (2010).20068001 10.1148/rg.302095213
15. Fahmy OH Is microthrombosis the main pathology in coronavirus disease 2019 severity?-A systematic review of the postmortem pathologic findings Crit. Care Explor. 2021 3 e0427 10.1097/cce.0000000000000427 34036278
Fahmy, O. H. et al. Is microthrombosis the main pathology in coronavirus disease 2019 severity?-A systematic review of the postmortem pathologic findings. Crit. Care Explor. 3, e0427. 10.1097/cce.0000000000000427 (2021).34036278 10.1097/cce.0000000000000427
16. Villar J Dexamethasone treatment for the acute respiratory distress syndrome: A multicentre, randomised controlled trial Lancet Respir. Med. 2020 8 267 276 10.1016/s2213-2600(19)30417-5 32043986
Villar, J. et al. Dexamethasone treatment for the acute respiratory distress syndrome: A multicentre, randomised controlled trial. Lancet Respir. Med. 8, 267–276. 10.1016/s2213-2600(19)30417-5 (2020).32043986 10.1016/s2213-2600(19)30417-5
17. Santa Cruz R Villarejo F Irrazabal C Ciapponi A High versus low positive end-expiratory pressure (PEEP) levels for mechanically ventilated adult patients with acute lung injury and acute respiratory distress syndrome Cochrane Database Syst Rev 2021 3 Cd009098 10.1002/14651858.CD009098.pub3 33784416
Santa Cruz, R., Villarejo, F., Irrazabal, C. & Ciapponi, A. High versus low positive end-expiratory pressure (PEEP) levels for mechanically ventilated adult patients with acute lung injury and acute respiratory distress syndrome. Cochrane Database Syst Rev 3, Cd009098. 10.1002/14651858.CD009098.pub3 (2021).33784416 10.1002/14651858.CD009098.pub3
18. Miró Ò Frequency, risk factors, clinical characteristics, and outcomes of spontaneous pneumothorax in patients with coronavirus disease 2019: A case-control, Emergency Medicine-Based Multicenter Study Chest 2021 159 1241 1255 10.1016/j.chest.2020.11.013 33227276
Miró, Ò. et al. Frequency, risk factors, clinical characteristics, and outcomes of spontaneous pneumothorax in patients with coronavirus disease 2019: A case-control, Emergency Medicine-Based Multicenter Study. Chest 159, 1241–1255. 10.1016/j.chest.2020.11.013 (2021).33227276 10.1016/j.chest.2020.11.013
19. Hallifax RJ Goldacre R Landray MJ Rahman NM Goldacre MJ Trends in the incidence and recurrence of inpatient-treated spontaneous pneumothorax, 1968–2016 Jama 2018 320 1471 1480 10.1001/jama.2018.14299 30304427
Hallifax, R. J., Goldacre, R., Landray, M. J., Rahman, N. M. & Goldacre, M. J. Trends in the incidence and recurrence of inpatient-treated spontaneous pneumothorax, 1968–2016. Jama 320, 1471–1480. 10.1001/jama.2018.14299 (2018).30304427 10.1001/jama.2018.14299
20. Elabbadi A Spontaneous pneumomediastinum: A surrogate of P-SILI in critically ill COVID-19 patients Crit Care 2022 26 350 10.1186/s13054-022-04228-1 36371306
Elabbadi, A. et al. Spontaneous pneumomediastinum: A surrogate of P-SILI in critically ill COVID-19 patients. Crit Care 26, 350. 10.1186/s13054-022-04228-1 (2022).36371306 10.1186/s13054-022-04228-1
21. Cai Z Patient self-inflicted lung injury associated pneumothorax/pneumomediastinum is a risk factor for worse outcomes of severe COVID-19: A case-control study Sci Rep 2024 14 15437 10.1038/s41598-024-66229-0 38965293
Cai, Z. et al. Patient self-inflicted lung injury associated pneumothorax/pneumomediastinum is a risk factor for worse outcomes of severe COVID-19: A case-control study. Sci Rep 14, 15437. 10.1038/s41598-024-66229-0 (2024).38965293 10.1038/s41598-024-66229-0
22. Esnault P High respiratory drive and excessive respiratory efforts predict relapse of respiratory failure in critically ill patients with COVID-19 Am. J. Respir Crit. Care Med. 2020 202 1173 1178 10.1164/rccm.202005-1582LE 32755309
Esnault, P. et al. High respiratory drive and excessive respiratory efforts predict relapse of respiratory failure in critically ill patients with COVID-19. Am. J. Respir Crit. Care Med. 202, 1173–1178. 10.1164/rccm.202005-1582LE (2020).32755309 10.1164/rccm.202005-1582LE
23. Tonelli R Inspiratory effort and lung mechanics in spontaneously breathing patients with acute respiratory failure due to COVID-19: A matched control study Am J Respir Crit Care Med 2021 204 725 728 10.1164/rccm.202104-1029LE 34214009
Tonelli, R. et al. Inspiratory effort and lung mechanics in spontaneously breathing patients with acute respiratory failure due to COVID-19: A matched control study. Am J Respir Crit Care Med 204, 725–728. 10.1164/rccm.202104-1029LE (2021).34214009 10.1164/rccm.202104-1029LE
24. Weaver L High risk of patient self-inflicted lung injury in COVID-19 with frequently encountered spontaneous breathing patterns: a computational modelling study Ann Intensive Care 2021 11 109 10.1186/s13613-021-00904-7 34255207
Weaver, L. et al. High risk of patient self-inflicted lung injury in COVID-19 with frequently encountered spontaneous breathing patterns: a computational modelling study. Ann Intensive Care 11, 109. 10.1186/s13613-021-00904-7 (2021).34255207 10.1186/s13613-021-00904-7
25. Horby P Dexamethasone in hospitalized patients with covid-19 N Engl J Med 2021 384 693 704 10.1056/NEJMoa2021436 32678530
Horby, P. et al. Dexamethasone in hospitalized patients with covid-19. N Engl J Med 384, 693–704. 10.1056/NEJMoa2021436 (2021).32678530 10.1056/NEJMoa2021436
26. National Institutes of Health: COVID-19 Treatment Guidelines: Systemic Corticosteroids. https://www.covid19treatmentguidelines.nih.gov/therapies/immunomodulators/systemic-corticosteroids/ (accessed 20 Mar, 2024).
27. Ni YN Chen G Sun J Liang BM Liang ZA The effect of corticosteroids on mortality of patients with influenza pneumonia: A systematic review and meta-analysis Crit. Care 2019 23 99 10.1186/s13054-019-2395-8 30917856
Ni, Y. N., Chen, G., Sun, J., Liang, B. M. & Liang, Z. A. The effect of corticosteroids on mortality of patients with influenza pneumonia: A systematic review and meta-analysis. Crit. Care 23, 99. 10.1186/s13054-019-2395-8 (2019).30917856 10.1186/s13054-019-2395-8
28. Eastridge CE Hamman JL Pneumothorax complicated by chronic steroid treatment Am. J. Surg. 1973 126 784 787 10.1016/s0002-9610(73)80071-6 4758799
Eastridge, C. E. & Hamman, J. L. Pneumothorax complicated by chronic steroid treatment. Am. J. Surg. 126, 784–787. 10.1016/s0002-9610(73)80071-6 (1973).4758799 10.1016/s0002-9610(73)80071-6
29. Nishimoto K Pneumothorax in connective tissue disease-associated interstitial lung disease PLoS ONE 2020 15 e0235624 10.1371/journal.pone.0235624 32634173
Nishimoto, K. et al. Pneumothorax in connective tissue disease-associated interstitial lung disease. PLoS ONE 15, e0235624. 10.1371/journal.pone.0235624 (2020).32634173 10.1371/journal.pone.0235624
30. Belletti A Predictors of pneumothorax/pneumomediastinum in mechanically ventilated covid-19 patients J. Cardiothorac. Vasc. Anesth. 2021 35 3642 3651 10.1053/j.jvca.2021.02.008 33678544
Belletti, A. et al. Predictors of pneumothorax/pneumomediastinum in mechanically ventilated covid-19 patients. J. Cardiothorac. Vasc. Anesth. 35, 3642–3651. 10.1053/j.jvca.2021.02.008 (2021).33678544 10.1053/j.jvca.2021.02.008
31. Kahn MR Watson RL Thetford JT Wong JI Kamangar N High incidence of barotrauma in patients with severe coronavirus disease 2019 J. Intensive Care Med. 2021 36 646 654 10.1177/0885066621989959 33722090
Kahn, M. R., Watson, R. L., Thetford, J. T., Wong, J. I. & Kamangar, N. High incidence of barotrauma in patients with severe coronavirus disease 2019. J. Intensive Care Med. 36, 646–654. 10.1177/0885066621989959 (2021).33722090 10.1177/0885066621989959
32. National Health Insurance Service: 2021 National Health Insurance statistics. https://www.hira.or.kr/bbsDummy.do?pgmid=HIRAA020045020000&brdScnBltNo=4&brdBltNo=2314&pageIndex=1&pageIndex2=1. (Accessed 20 Mar, 2024).
33. Austin PC Stuart EA Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies Stat Med 2015 34 3661 3679 10.1002/sim.6607 26238958
Austin, P. C. & Stuart, E. A. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med 34, 3661–3679. 10.1002/sim.6607 (2015).26238958 10.1002/sim.6607
34. Chesnaye NC An introduction to inverse probability of treatment weighting in observational research Clin. Kidney J. 2022 15 14 20 10.1093/ckj/sfab158 35035932
Chesnaye, N. C. et al. An introduction to inverse probability of treatment weighting in observational research. Clin. Kidney J. 15, 14–20. 10.1093/ckj/sfab158 (2022).35035932 10.1093/ckj/sfab158
35. Austin PC Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples Stat. Med. 2009 28 3083 3107 10.1002/sim.3697 19757444
Austin, P. C. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat. Med. 28, 3083–3107. 10.1002/sim.3697 (2009).19757444 10.1002/sim.3697
36. Austin PC Variance estimation when using inverse probability of treatment weighting (IPTW) with survival analysis Stat. Med. 2016 35 5642 5655 10.1002/sim.7084 27549016
Austin, P. C. Variance estimation when using inverse probability of treatment weighting (IPTW) with survival analysis. Stat. Med. 35, 5642–5655. 10.1002/sim.7084 (2016).27549016 10.1002/sim.7084
