Predictors and Impact of Pneumothorax on Patients of Chronic Obstructive Pulmonary Disease Exacerbation: A Study from National Inpatient Sample


Internal Medicine

Document Type

Conference Proceeding

Publication Title

American Journal of Respiratory and Critical Care Medicine

Conference Name

ATS 2019 International Conference

Conference Date



Rationale: Patients with chronic obstructive pulmonary disease (COPD) have limited cardiopulmonary reserve. COPD is a common cause of secondary spontaneous pneumothorax and cases of COPD exacerbation can get complicated by pneumothorax (secondary spontaneous/iatrogenic). We aimed to determine prevalence of pneumothorax in these patients and to identify predictors of development of pneumothorax and mortality. Methods: We analyzed discharge data from the National Inpatient Sample 2012-2014, Healthcare Cost and Utilization Project. We reviewed discharges (age >18) with primary diagnosis of COPD exacerbation and identified patients with pneumothorax using the International Classification of Diseases, Ninth Revision, Clinical Modification code (491.21 and 512.0, 512.1, 512.82, 512.89), respectively. Multivariable logistic regression adjusting for age, gender and comorbidities was used to analyze adjusted odds ratio (OR) Results: A total of 1,377,785 discharges with a diagnosis of COPD exacerbation were identified. Of the total cohort 3,060 patients developed pneumothorax with a prevalence of 0.22%. Mean age was 68.6 years (standard deviation =11.8). On comparing COPD exacerbation patients with pneumothorax to those without pneumothorax, pneumothorax patients were more likely to be male (55% p <0.001), more likely to have acute kidney injury (AKI) (16.3% vs 8% p <0.001), sepsis (7.2% vs 1% p <0.001), were more likely to require noninvasive (10.9% vs 6.3% p <0.001) and invasive (31.2% vs2.2% p <0.001) ventilation. They were also, more likely to have hospital stay of >8 days (44.3% vs 7.5%). Interestingly, patients with pneumothorax were less obese (4.9 % vs 7.4% p <0.001) and less morbidly obese (2.9% vs 7.3% p<0.001) as compared to patients without pneumothorax. Unadjusted mortality in pneumothorax group was 17.3 % compared to 1.2 % (p <0.001) in those without pneumothorax. On multivariate regression analysis, development of pneumothorax had statistically significant increased risk of mortality with OR of 3.51 (95% CI: 2.65-4.65, p <0.001). Multivariate logistic regression for predictors of pneumothorax showed that patients with cancer (OR: 3.36, 95% CI: 2.70-4.18 p<0.001) and those requiring mechanical ventilation (OR:18.32, 95% CI 15.09-22.24 p<0.001) had higher odds of developing pneumothorax. On the other hand, non-invasive ventilation, age, female sex, sepsis, AKI, and obesity were not associated with development of pneumothorax. Conclusion: Our study showed that COPD exacerbation patients who developed pneumothorax had increased risk of mortality and increased length of hospital stay. Patients who required mechanical ventilation and those with cancer had increased odds of developing pneumothorax. Non-invasive ventilation was not associated with the development of pneumothorax.





Publication Date



American Thoracic Society


Thematic Poster Session

B64. COPD: Mechanism And Treatment

Abstract A3877