Show Summary Details
Page of

Management of pneumothorax and bronchial fistulae 

Management of pneumothorax and bronchial fistulae
Management of pneumothorax and bronchial fistulae

Wissam Abouzgheib

and Raquel Nahra

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2022. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 18 May 2022

Key points

  • Observation is adequate in small pneumothorax.

  • Small bore chest tubes are as effective as large ones and should be used initially.

  • Ultrasound plays an important role in ruling out a pneumothorax.

  • Suction is not always need in an air leak.

  • One-way endobronchial valves may be used to treat prolonged air leaks.

Definition and classification of pneumothorax

Pneumothorax refers to the presence of air in the pleural space. The classification of pneumothorax includes spontaneous, traumatic, or iatrogenic. Spontaneous pneumothorax occurs without obvious cause, either primary without evidence of underlying lung disease or secondary with apparent underlying lung disease, often chronic obstructive pulmonary disease (COPD). Traumatic pneumothorax occurs after a blunt or penetrating trauma to the chest. Iatrogenic pneumothorax occurs after a diagnostic or therapeutic intervention, such as transthoracic lung biopsy, central line placement, or barotrauma due to mechanical ventilation. The incidence of pneumothorax, in patients receiving mechanical ventilation, ranges between 7 and 14% [1]‌. Patients with acute lung injury or ARDS are at high risk.

The clinical presentation ranges from asymptomatic to respiratory failure and prolonged bronchopleural fistulas (BPFs). Prompt diagnosis and management are crucial, especially in symptomatic and critically-ill patients requiring mechanical ventilation. This chapter discusses the diagnosis and management of pneumothorax and BPFs in general with a focus on critically-ill patients


Clinical manifestations and presentations of pneumothorax are widely variable. Sometimes found incidentally on routine chest imaging, the presence of a pneumothorax is often clinically suspected with the appropriate clinical setting. Spontaneous pneumothorax is usually sought in patients between 20 and 40 years old presenting with sudden onset of dyspnoea and pleuritic chest pain. Secondary pneumothorax in patients with known underlying chronic lung diseases, presenting with more severe symptoms and iatrogenic or traumatic pneumothorax in patients with diagnostic/therapeutic interventions or preceding trauma. Symptoms vary, a small pneumothorax can be asymptomatic and self-limited, whereas a large pneumothorax can cause hypoventilation, hypoxaemia, and/or haemodynamic instability.

Tension pneumothorax represents a surgical emergency and requires emergent intervention. It may lead to respiratory failure requiring mechanical ventilation. It may also complicate pre-existing respiratory failure on positive pressure ventilation.

In intubated and sedated patients, a pneumothorax should be suspected with sudden and unexplained worsening respiratory failure, increased oxygen requirements, haemodynamic instability, and a sudden rise in peak and plateau pressures. It is ideally diagnosed based on clinical presentation, risk factors, and physical examination, and not by imaging followed by immediate emergent decompression.

The first-line imaging modality used to identify a pneumothorax is chest radiography.

A typical finding is the displacement of the white visceral pleural line from the chest wall on an upright chest X-ray. Contralateral shift of the trachea and mediastinum may also be present in spontaneous pneumothorax, but not necessarily suggestive of tension. The underlying lung parenchyma should be examined for the presence of underlying lung disease that would suggest a secondary spontaneous pneumothorax. In bedridden or ICU patients, care should be exercised in order to differentiate visceral pleural line from skin folds. Skin folds frequently extend beyond the rib cage, while blood vessels and lung parenchyma often extend beyond the skin fold. Their attenuation profile is also different, forming a negative black Mach band instead of the white visceral pleural line (Fig. 124.1)

Fig. 124.1 (Left) White visceral pleural line in pneumothorax. (Right) Black Mach band in skin fold.

Fig. 124.1 (Left) White visceral pleural line in pneumothorax. (Right) Black Mach band in skin fold.

Computed tomography (CT) is reserved for complicated or unclear situations. However, CT scans are more accurate in determining size of pneumothorax when compared to chest radiography [2]‌.

Some investigators have reported using ultrasound to diagnose or rule out a pneumothorax [3]‌, particularly in patients where it is needed urgently at the bedside, such as in ICU or for a trauma patient. The use of bedside ultrasonography has emerged in the past few years as the modality of choice in intensive care units (ICUs), where ultrasonography trained physicians are available. Bedside ultrasonography offers several advantages over chest radiography or CT scans, including rapid availability, lack of radiation, real time interpretation, and lower cost. It also offers the ability to immediately rule out a pneumothorax after an invasive procedure or in the midst of a clinical deterioration. Two easily identifiable signs are needed to rule out pneumothorax. The sliding lung sign presents as shimmering of the visceral pleura when it moves relative to the parietal pleura during the respiratory cycle. The Comet tail artefacts (also called B-lines) are several echogenic lines that originate at the visceral–parietal pleural interface and extend to the bottom of the sonographic image (Fig. 124.2). The sensitivity and specificity of ultrasound for pneumothorax range from 86 to 98%, which is superior to supine chest radiograph (sensitivity 28 to 75%) [4,5]. Both techniques demonstrated high specificity.

Fig. 124.2 Ultrasound. (Left) No pneumothorax. (Right) Pneumothorax: absence of pleural sliding or B-lines.

Fig. 124.2 Ultrasound. (Left) No pneumothorax. (Right) Pneumothorax: absence of pleural sliding or B-lines.

Management of pneumothorax

Several parameters need to be considered in the management of a pneumothorax, such as type, size, severity of symptoms, first episode or recurrence, and patient preference. The British Thoracic Society guidelines define a pneumothorax as small if the distance from chest wall to the visceral pleural line is less than 2 cm or large if the distance is 2 cm or greater [6]‌. Some clinicians prefer 3 cm laterally and 4 cm apically as the threshold to distinguish small and large pneumothoraces.

Conservative approach: oxygen and observation

The first episode of spontaneous pneumothorax warrants a conservative approach when faced with small size pneumothorax and symptoms are absent or mild. Normal air re-absorption from the pleural space occurs at a rate of 1.25 % and it increases three- to four-fold when high flow (non-intubated patient) or high FiO2 (intubated patient) oxygen supplementation is instituted. Patients managed conservatively should be observed in the emergency department for 3–6 hours and discharged home if repeat chest imaging shows no progression.

Manual needle/catheter aspiration

Simple manual aspiration using an intercostal needle or small catheter is indicated in non-complicated patients presenting with a first episode spontaneous pneumothorax, large pneumothorax, or associated with symptoms. It consists of inserting a needle or catheter in the pleural space, and aspirating the pleural air followed by removal of the needle or catheter. The overall resolution rate in spontaneous pneumothorax ranges between 59–83% of patients [7]‌.

Tube thoracostomy: insertion and management

When simple aspiration is unsuccessful to keep the lung inflated, or when air leak is large or persistent, then a tube thoracostomy is indicated. Prior reflection on the magnitude of the underlying air leak, size of the pneumothorax, associated co-morbidities, and concurrent therapies may help when choosing tube thoracostomy size. There is no evidence that large tubes (20–24 F) are any better than small tubes (10–16 F) in the management of pneumothoraces. The initial use of large (20–24 F) intercostal tubes is not recommended, although it may become necessary to replace a small chest tube with a larger one if there is a large air leak preventing complete re-inflation of the lung [6]‌.

The presence and degree of an air leak after chest tube placement, is also important and should be assessed daily to determine whether the tube can be removed. The number of chambers that are bubbling in a wet, suction-controlled, closed drainage provides a semi-quantitative measure of the severity of the leak. It does not indicate flow in any precise manner, but can provide an indication of any day-to-day increases or decreases in the degree of air leak.

Our recommended approach is, for patients who are not at risk for a large air leak, a smaller catheter (8–14 Fr) can be used in combination with a water seal or unidirectional flutter valve (i.e., Heimlich valve), which allows the patient to be mobile. These patients can be followed the next day with chest X-ray after clamping the chest tube for several hours for chest tube removal.

For patients with moderate size pneumothorax or possible moderate air leak, a 16–24 Fr chest tube is usually sufficient to maintain evacuation of the pleural space. The tube may initially be connected to a water seal, but if lung re-expansion does not occur then application of suction (–20 cmH2O) is needed. Suction helps juxtapose the visceral and parietal pleura, which in theory leads to mechanical pressure on the bronchopleural fistula and potentially promotes healing.

A large surgical chest tube >24 Fr is needed in patients with large air leak, in mechanically-ventilated patients, and when smaller size tubes fail to re-expand the lung.

Small bore drains are as effective for air drainage as large bore drains and are more comfortable for patients. If there is associated blood, a large bore drain will be required. There are no large randomized, controlled trials directly comparing small and large bore drains.

The most common position for chest tube insertion is in the mid-axillary line, through the ‘safe triangle’. This position minimizes risk to underlying structures, such as the viscera and internal mammary artery. A more posterior position may be chosen if suggested by the presence of a loculated collection. While this is relatively safe, it is not the preferred site, as it is more uncomfortable for the patient to lie on after insertion and there is more risk of the drain kinking.

For apical and large pneumothoraces extending to the apex, an antero-apical approach is favoured. It requires minimal positioning and rotation of a critically-ill patient. The second intercostal space in the mid-clavicular line is often chosen, two finger breadths from the lateral sternal border. The internal mammary vessels are at risk, but bedside ultrasound may be very helpful in choosing the optimal location, while avoiding vascular structures.

Once the lung has completely expanded and the air leak has resolved, the chest tube removal process begins. Applied suction should be discontinued first [7]‌. The chest tube should be clamped 6–12 hours after the last evidence of an air leak and a chest radiograph be performed 24 hours after the last evidence of an air leak [7]. If repeat chest X-ray shows resolution of pleural air, then chest tube removal is warranted.

In the case of clinical deterioration, such as worsening dyspnoea, hypoxaemia, subcutaneous emphysema, or signs of tension pneumothorax, the chest tube should be immediately unclamped and connected to suction.

When an air leak persists for more than 3–5 days then air leak management and recurrence prevention should be considered [8]‌.

Bronchopleural fistula

A bronchopleural fistula (BPF) is a communication between the pleural space and the bronchial tree. It represents a significant clinical problem and generally occur after a spontaneous pneumothorax caused by underlying lung diseases, chest trauma, lung resection, or biopsy. The post-operative complication of pulmonary resection is the most common cause. It is also associated with increased costs and high morbidity and mortality [9]‌.

When significant, a BPF may contribute to respiratory failure by increasing work of breathing. However, even when minimal, they can contribute to other complications related to prolonged bed rest or inactivity, such as atelectasis, deconditioning, nosocomial infections, and deep venous thrombosis [10].

The severity of the air leak is generally stratified depending on its frequency (intermittent or continuous) or its occurrence during respiratory cycles (inspiratory, expiratory, or both). Occasionally, it may be severe, leading to detectable differences in the inspired and expired tidal volumes in patients on mechanical ventilation.

When one or more large BPF occurs in a patient with acute respiratory distress syndrome (ARDS), management is extremely challenging. Ventilation is usually not a big problem since air exiting through the BPF has significant CO2 and has contributed to CO2 elimination. Oxygenation is the main problem, since it is difficult to maintain adequate positive end expiratory pressure (PEEP). In attempts to minimize leak and facilitate healing of the BPF, high frequency oscillatory ventilation may be useful.

Single lung ventilation may also be considered where large BPF’s are present in the mechanically-ventilated patient. The strategy here would be to allow ventilation of the side without the BPF with more conventional ventilator settings, while using settings that minimize tidal volume and airway pressure on the side of the leak.

The initial management of a BPF is conservative for the first 3–5 days with tube thoracostomy, together with treatment of the underlying lung diseases. In patients receiving mechanical ventilation, reducing the air leak through ventilator adjustments to minimize the tidal volume and plateau airway pressure may also help. Ventilator settings should be adjusted to minimize both tidal volume and plateau airway pressure, which may lead to lesser alveolar distension and lower transpulmonary pressure gradient. Applying the least amount of chest tube suction necessary may also be beneficial, as well as weaning and extubating the patient as soon as possible. These strategies are not supported by high quality evidence, but elimination or reduction of contributing factors makes it less likely that a new pneumothorax will develop, or that the existing BPF will worsen and become physiologically significant.

When prolonged, current guidelines recommend early surgical consultation for possible thoracotomy or video-assisted thoracoscopic surgery [7,8], although the American College of Chest Physicians favours the latter [7]‌.

Surgical correction of BPFs has been associated with high success rate (80–95%) and low mortality rate. However, in critically-ill patients or those with poor performance status, surgical interventions may be too risky. Bronchoscopic management is a less invasive alternative in such situations and may be conducted at the bedside in the ICU. Various devices and sealing compounds have been tried and used with flexible bronchoscopy in the past, but with limited success [11,12,13], such as instillation of fibrin, acrylic glue, and tissue glue [14]. Also deployment of self-expandable metallic stents or watanabe spigots have been utilized [12].

One-way endobronchial valves, initially developed for bronchoscopic reduction surgery, became available more than a decade ago and have been successfully used in case series for the treatment of prolonged air leaks [15,16]. The airway leading the air leak is initially identified using intermittent balloon occlusion. The valve is then introduced to the desired airway loaded on a delivery catheter through the working channel of the bronchoscope. The valve is umbrella-shaped when deployed in the airway. It fully expands during inspiration and blocks the airflow distally, while it slightly collapses and allows for mucus clearance during expiration (Fig. 124.3). The valve is easily removable bronchoscopically several weeks after resolution of air leak.

Fig. 124.3 Endobronchial valve. Placement. Left, bronchoscope introduced to culprit airway. Middle, preloaded valve into delivery catheter introduced through the scope channel. Right, deployed valve diverting air going towards broncho-pleural fistula.

Fig. 124.3 Endobronchial valve. Placement. Left, bronchoscope introduced to culprit airway. Middle, preloaded valve into delivery catheter introduced through the scope channel. Right, deployed valve diverting air going towards broncho-pleural fistula.

Spiration® Valve System images used courtesy of Spiration, Inc.


Pneumothorax and bronchopleural fistula are frequently encountered in clinical practice. In critically-ill patients, they could represent a serious problem leading to significant co-morbidities. Rapid recognition and intervention are key elements. Bedside ultrasound in the ICU is a new diagnostic tool that facilitates the early diagnosis and treatment.

Surgical correction of BPFs remains the gold standard treatment, but in critically-ill patients, several bronchoscopic interventions may be offered. One-way endobronchial valves placement in the leaky airway is a very promising new strategy in poor surgical candidates.


The authors acknowledge R. Phillip Dellinger MD, Camden, USA for his guidance in the preparation of this manuscript.


1. de Lassence A, Timsit JF, Tafflet M, et al. (2006). Pneumothorax in the intensive care unit: incidence, risk factors, and outcome. Anesthesiology, 104, 5–13.Find this resource:

2. Engdahl O, Toft T, and Boe J. (1993). Chest radiograph—a poor method for determining the size of a pneumothorax. Chest, 103, 26–9.Find this resource:

3. Mandavia DP and Joseph A. (2004). Bedside echocardiography in chest trauma. Emergency Medical Clinics of North America, 22, 601–19.Find this resource:

4. Wilkerson RG and Stone MB. (2010). Sensitivity of bedside ultrasound and supine anteroposterior chest radiographs for the identification of pneumothorax after blunt trauma. Academic Emergency Medicine, 17, 11–17.Find this resource:

5. Lichtenstein DA. (2009). Ultrasound examination of the lungs in the intensive care unit. Pediatric Critical Care Medicine, 10, 693–8.Find this resource:

6. Henry M, Arnold T, and Harvey J. (2003). BTS guidelines for the management of spontaneous pneumothorax. Thorax, 58(Suppl. 2), ii39–52.Find this resource:

7. Baumann MH, Strange C, Heffner JE, et al. (2001). Management of spontaneous pneumothorax: an American College of Chest Physicians Delphi consensus statement. Chest, 119, 590–602.Find this resource:

8. MacDuff A, Arnold A, and Harvey J. (2010). Management of spontaneous pneumothorax: British Thoracic Society Pleural Disease Guideline 2010. Thorax, 65(Suppl. 2), ii18–31.Find this resource:

9. Shrager JB, DeCamp MM, and Murthy SC. (2009). Intraoperative and postoperative management of air leaks in patients with emphysema. Thoracic Surgery Clinics, 19, 223–31, ix.Find this resource:

10. Cerfolio RJ. (2002). Advances in thoracostomy tube management. Surgical Clinics of North America, 82, 833–48, vii.Find this resource:

11. Takaoka K, Inoue S, and Ohira S. (2002). Central bronchopleural fistulas closed by bronchoscopic injection of absolute ethanol. Chest, 122, 374–8.Find this resource:

12. Watanabe S, Watanabe T, and Urayama H. (2003). Endobronchial occlusion method of bronchopleural fistula with metallic coils and glue. Thoracic Cardiovascular Surgery, 51, 106–8.Find this resource:

13. Varoli F, Roviaro G, Grignani F, et al. (1998). Endoscopic treatment of bronchopleural fistulas. Annals of Thoracic Surgery, 65, 807–9.Find this resource:

14. Baumann WR, Ulmer JL, Ambrose PG, et al. (1997). Closure of a bronchopleural fistula using decalcified human spongiosa and a fibrin sealant. Annals of Thoracic Surgery, 64, 230–3.Find this resource:

15. Travaline JM, McKenna RJ, Jr, De Giacomo T, et al. (2009). Treatment of persistent pulmonary air leaks using endobronchial valves. Chest, 136, 355–60.Find this resource:

16. Gillespie CT, Sterman DH, Cerfolio RJ, et al. (2001). Endobronchial valve treatment for prolonged air leaks of the lung: a case series. Annals of Thoracic Surgery, 91, 270–3.Find this resource: