Although mechanical ventilation (MV) is considered mainstay of therapy for respiratory failure, it can be accompanied by various adverse effects. Pneumothorax (PTX) is one of the most critical complications occurred during positive pressure ventilation.[1] It is associated with elevated airway pressure in the presence of underlying lung diseases like acute respiratory distress syndrome (ARDS).[2]

Reported incidence rate of PTX ranged from 4% to 15% in patients receiving mechanical ventilation. PTX can develop as a result of thoracentesis, central line insertion, bronchoscopy and tracheotomy, in addition to ventilator-induced barotraumas.[3–5]

When its location is not typical, diagnosis of PTX in patients with MV can be difficult with chest radiography alone. Decreased mental status and underlying lung diseases also complicate early diagnosis based on clinical findings.[6] In patients with MV, PTX promotes the entry of large amounts of air into the pleural space as a result of continued positive pressure, leading to tension PTX and subsequent cardiovascular collapse and death.[7] The aim of this article was to identify clinical characteristics in critically ill patients who developed PTX during MV.

In patients admitted to intensive care unit (ICU), clinical signs and symptoms, laboratory test results, radiographic findings were prospectively recorded into our electronic medical record system during their hospital stay from December 2011 through March 2013 (13 ICU beds). Data entry was performed by an internal medicine resident and reviewed by an internal medicine staff in the ICU. Based on prospectively entered data, patients who developed PTX were screened and their medical records were retrospectively analyzed. Patients were excluded if they were aged below 18 or maintained on MV following surgery or neurological injury and if they already had a chest tube inserted due to PTX before being admitted to the ICU. Patients’ data were separately analyzed if their PTX was associated with iatrogenic causes.

A total of 326 (208 men and 118 women with the mean age of 65.3 ± 8.74) were hospitalized during the study period, and their clinical characteristics were summarized in Table 1. On admission, APACHE II and SOFA scores were 18.1 ± 9.01 and 8.1 ± 4.58, respectively. Diabetes and hypertension were the most common underlying diseases. Of 326 patients, 89 (27.3%) met the diagnostic criteria for acute lung injury (ALI)/ARDS and 45 (13.8%) had chronic obstructive pulmonary disease (COPD).[8] The average length of stay in the ICU was 24.5 ± 69.1 days, and 84.6% patients had hospital stay of two weeks or less. A total of 75 patients (23%) died in the ICU while 156 patients (47.4%) were discharged to survive.

In the ICU, 15 patients (4.6%) developed PTX. Of those, four patients had iatrogenic PTX following thoracentesis in one case and central line insertion in three cases. The remaining 11 patients (3.3%) (6 men and 5 women with the mean age of 68.3 ± 9.12) appeared to have MV-associated PTX. Diagnosis was made by chest radiograph in seven cases (63.6%) and by chest CT scan alone in two cases (18.2%). As for underlying lung diseases, four patients had either COPD or idiopathic pulmonary fibrosis (two for each condition), and seven (63.6%) met diagnostic criteria for ALI/ARDS. During the study period, PTX occurred in two patients (4.7%) out of 45 patients with COPD and seven patients (7.8%) out of 89 patients with ALI/ARDS. Those nine patients were placed on assisted MV mode when PTX occurred, showing average positive end-expiratory pressure of 7.9 ± 4.6 cmH₂O and tidal volume of 361 ± 63.7 ml (6.8 ± 1.47 ml/kg). Time duration between application of MV and the onset of PTX was 11 ± 15.5 (1–55) days. At the time of diagnosis of PTX, five patients (45.5%) exhibited systolic blood pressure of less than 90mmHg, and seven patients (63.6%) showed pulse rate of more than 120 beat/min. Of 11 patients with PTX, five died, and two of them (18.2%) died of PTX. Chest tube insertion was performed for 10 patients (90.9%), and a large bore catheter was percutaneously inserted into the remaining one patient (Table 2). One patient removed a chest tube during MV and two patients removed a chest tube after extubation. Two patients were discharged with MV and chest tube inserted.

Two patients whose death was likely caused by PTX had the following chest radiograph results and CT findings during hospitalization (Figs.1, 2). One of two patients developed tension PTX during treatment of ventilation-associated pneumonia (VAP), calling for immediate chest tube insertion. However, the patient’s underlying organ failure exacerbated, and the patient died 3 days after the onset of PTX. The other patient was placed on MV for more than two weeks due to VAP, but acute hypoxia and hypotension occurred and got worse, and PTX was detected on chest X-ray in 5 hours. A chest tube was inserted 2 hours after the diagnosis of PTX. However the patient’s condition rapidly deteriorated and died about 12 hours later.

This article was to describe PTX as one of the most serious complications in critically ill patients receiving positive pressure MV. In this article, we demonstrate that PTX can be the main cause of death although its incidence rate is not high.

In this article, MV-associated PTX occurred in 11 patients, showing an incidence rate of nearly 3.4%. In the past, the incidence of PTX used to be high at 20% in patients with MV in the ICU.[2,3,9,10] However, recent studies reported the incidence rate of 3% in mechanical ventilated patients.[11,12] Lowered incidence of PTX was likely due to lung protective ventilation strategies with low tidal volumes that are widely performed in patients with ARDS in clinical practices. Low tidal volumes used for other types of respiratory disease are also likely to contribute to lowering the incidence of PTX with efficacy in preventing a sudden rise in airway pressure.[13–15] In addition, advanced skills for procedures such as thoracentesis and central venous catheter insertion and increasing application of ultrasound in ICU procedures appear to have reduced the incidence of lung injury.

The risk of PTX is closely associated with underlying lung diseases. In this article, PTX occurred in the patients mechanically ventilated due to ARDS, which is the most common risk factor for MV-associated PTX.[2] Chronic obstructive airway disease is also known to have a moderate level of risk of PTX.[2] In Anzeuto et al.[11] study,154 patients developed PTX during MV showed that 10% of total episodes of PTX occurred in chronic interstitial lung disease (ILD) group, followed by the ARDS group (6.5%), the asthma group (6.3%) and the COPD group (2.9%). In our study, we found similar results that the patients who developed PTX had ARDS, COPD and ILD as underlying diseases.

In this article, 11 patients developed PTX, which was the main cause of death for two patients. One of patients demonstrated exacerbated underlying organ failure before dying a few days after the onset of PTX. The other patient experienced acute hypotension and shock and progressive multi-organ failure after PTX. If PTX develops during MV, continued positive pressure turns intrapleural pressure positive, which put pressure on structures in the mediastinum. As a result, tension PTX can develop with hypotension and shock.[7] Because tension PTX can cause death by cardiovascular compromise, immediate diagnosis and intervention are crucial to prevent the progression from MV-associated PTX to tension PTX.[16] However, not every case of MV-associated PTX progresses to tension PTX. Death resulted from MV-associated tension PTX only in a few cases of PTX. An earlier study reported 13 cases of PTX in 100 patients with ALI and that only one patient died of PTX while the remaining patients had other cause of death than PTX, claiming that organ failure was more responsible for death.[17]

Diagnosis of PTX can be difficult in patients receiving MV. In this article, PTX was diagnosed with chest radiographs in most cases while two cases were diagnosed only by CT scan. Because of the limitations of chest radiograph as a diagnostic tool in the ICU, the role of ultrasound scan is growing in the diagnosis of lung parenchyma or pleural diseases.[18] In Xirouchaki et al.[19] study, chest radiographic findings were prospectively compared with ultrasound findings in patients who were slated to undergo CT scan in the ICU. They found ultrasonography was more effective in the diagnosis of PTX.

Immediate chest tube insertion is an important step in the treatment of PTX in the ICU to remove intrapleural air and eventually prevent tension PTX. The patient with tension PTX requires intervention immediately before diagnosing PTX on chest radiography. When tension PTX is suspected with clinical findings including sudden hypoxia, hypertension, tachycardia and increased airway pressure, a 14–16 gauge needle should be inserted into the pleural space through the second intercostal space before insertion of a chest tube.[7] Chest tube insertion represents an essential therapeutic procedure for PTX developed during MV. A large bore tube is commonly used. However, the British Thoracic Society recommends small-sized tubes for PTX management.[20] The most common disadvantage of small-bore tube is that it functions less effectively in draining air or liquid, although intermittent irrigations can overcome such an issue.[7] The use of small-bore chest tubes for 62 mechanically ventilated patients was successful in 70% of cases without specific adverse effects in an earlier study.[12]

This article underlines the importance of appropriate diagnosis and chest tube insertion for critically ill patients who developed PTX during MV because MV-associated PTX may progress to a tension PTX and eventually lead to death.