This article discusses the theoretical and practical aspects of using bedside ultrasonography for pneumothorax. Although the Focused Assessment with Sonography in Trauma (FAST) examination has been part of Advanced Trauma Life Support (ATLS) for over a decade,  the addition of pneumothorax evaluation to this protocol to create the extended FAST (e-FAST) examination is a relatively new development.
Over the past 15 years, the use of bedside ultrasonography in the emergency department (ED) has revolutionized patient care.  Trauma patients, who are particularly vulnerable to the complications of delayed diagnostic studies, have benefited tremendously from rapid ultrasound-guided evaluation at the bedside. [3, 4, 5] In 2011, the Eastern Association for the Surgery of Trauma updated their guidelines, giving a level 2 recommendation for the use of thoracic ultrasound to diagnose pneumothorax. 
The traditional initial evaluation of patients with a suspected traumatic pneumothorax was with chest radiography, typically performed with the patient in the supine position so as to preserve cervical spine immobilization. However, this method is grossly inadequate for detecting pneumothoraces, with sensitivities as low as 36–48% in some studies.  Given the low sensitivity of portable chest x-ray, ultrasound has also been suggested as an alternative to portable chest x-ray in the detection of traumatic pneumothorax. In a retrospective review performed at a level I trauma center, ultrasound was shown to predict the absence of pneumothorax with 93.8% sensitivity and a negative predictive value of 99.9% when compared to chest x-ray. One pneumothorax was missed on ultrasound that was visualized on chest x-ray; however, this was a small, low anterior pneumothorax that was not clinically significant. All patients in this study who required tube thoracostomy placement had pneumothoraces that were detected on ultrasound. These findings suggest that for clinically significant pneumothoraces, ultrasound is a feasible alternative or replacement for portable chest x-ray, as it is more readily available and has higher sensitivity. 
In a retrospective study, the diagnostic accuracy of the e-FAST was compared to that of multidetector computed tomography (MDCT) and of invasive interventions. With the use of CT scanning, 87 pneumothoraces were detected among the 763 lung fields studies; with the use of ultrasound, 67 of 87 pneumothoraces were detected. Of the 20 missed, 17 were considered mild and not life-threatening. The diagnostic performance of ultrasound in this study was found to have a sensitivity of 77% and a specificity of 99.8%. Although ultrasonography may not be as diagnostically accurate as CT scanning, CT is limited in unstable patients, and this limitation can result in delayed diagnosis and treatment. 
Although bedside sonography was first implemented by emergency medicine physicians, it is now used by multiple specialties, including radiologists,  surgeons, and adult and pediatric intensivists. [11, 12] Furthermore, in addition to its traditional use in the ED, bedside ultrasound is starting to be used to confirm central line placement  and to monitor postoperative patients. [14, 15] Studies have also shown that bedside ultrasonography is a valuable tool in diagnosing iatrogenic pneumothoraces.  In the past few years, there has been an international effort to standardize the application of lung ultrasound, and in 2011, the International Liaison Committee on Lung Ultrasound (ILC-LUS) published a set of evidence-based recommendations for lung ultrasound, including the application of ultrasound for the detection of pneumothorax. 
Ultrasound use has also been suggested as a feasible diagnostic tool in the prehospital setting. In a transport vehicle, ambient noise and movement may limit physical exam and delay diagnosis. In a prospective, observational study, nonphysician aeromedical providers were trained to perform and interpret thoracic ultrasound in the detection of pneumothoraces in adult trauma patients and adult medical patients requiring endotracheal intubation. Twenty pneumothoraces and one mainstem intubation were identified, of which 16 were correctly identified and the mainstem intubation was incorrectly diagnosed. Prehospital ultrasound had a sensitivity of 68% (95% confidence interval (CI) 46-85%), a specificity of 96% (95% CI 90-98%), and an overall accuracy of 91% (95% CI 85-95%). In comparison, ED ultrasound had a sensitivity of 84% (95% CI 62-94%), a specificity of 98% (95% CI 93-99%), and an accuracy of 96% (95% CI 90-98%). 
It is critical to maintain a high suspicion for a pneumothorax in all trauma patients with a significant mechanism, as well as in nontrauma patients with a history or physical findings suggestive of the diagnosis.
Pneumothorax should be considered in the following clinical situations:
Penetrating trauma to the thorax and abdomen
Blunt trauma to the thorax or abdomen
Blunt trauma with a significant mechanism (eg, a high-speed motor vehicle accident or a major fall)
Post-procedures involving the thorax (eg, thoracentesis, internal jugular or subclavian central line placement)
The following signs are suggestive of pneumothorax:
Decreased or absent lung sounds, especially if unilateral
Resistance to ventilation (eg, patients who are "hard to bag")
Tachycardia (a nonspecific but common finding)
Jugular venous distention
Tracheal deviation (a late finding in tension pneumothorax)
The following symptoms are suggestive of pneumothorax:
There are no specific contraindications for bedside ultrasonography in the ED; clinical judgment should be used to determine its appropriate application. Patients with significant pulmonary disease and previous pneumothoraces may have altered findings on ultrasound examinations. In these special circumstances, patients should undergo definitive imaging with thoracic CT scanning when stable.
Successful application of ultrasonography to the detection of pneumothorax requires a solid understanding of the basic anatomy and pathophysiology of the pulmonary system. Both lungs are divided into lobes. The gross functional subunits of each lung are called segments. The right lung comprises 10 segments: 3 in the right upper lobe (apical, anterior and medial), 2 in the right middle lobe (medial and lateral), and 5 in the right lower lobe (superior, medial, anterior, lateral, and posterior). The left lung comprises 8 segments: 4 in the left upper lobe (apicoposterior, anterior, superior lingula, and inferior lingula) and 4 in the left lower lobe (superior, anteromedial, lateral, and posterior). For more information about the relevant anatomy, see Lung Anatomy.
Normally, the lungs are covered by a continuous serous membrane that folds back on itself to create 2 layers: the visceral pleura, which covers the lungs and adjoining structures, and the parietal pleura, which is attached to the chest wall. The pleural cavity is the "potential" space between these closely apposed layers and is normally filled with a small amount of pleural fluid, which allows for normal lung sliding during respiration. In normal circumstances, negative pressure prevents air from entering the pleural space.
A pneumothorax occurs when air infiltrates the pleural cavity, either as a result of traumatic injury to the chest wall or spontaneously (spontaneous pneumothorax is historically believed to be caused by the rupture of pleural blebs). Although there are many risk factors that predispose patients to the development of a pneumothorax, most patients presenting to the ED have pneumothorax secondary to trauma.
A particularly dangerous complication, the dreaded tension pneumothorax, can develop when an injury to the lung parenchyma or bronchus acts as a 1-way valve, allowing air to enter the pleural cavity but preventing it from escaping. A tension pneumothorax can develop rapidly and is greatly exacerbated by positive-pressure ventilation, posing a great danger to intubated patients. For all of these reasons, rapid detection of pneumothoraces in trauma patients is critical, and bedside ultrasonography is a fast, reliable means of accomplishing this task.
The following considerations should be kept in mind in the performance of the procedure (see Technique):
Small pneumothoraces tend to show the lung point in the anterior chest, whereas larger pneumothoraces have their transition areas on the lateral chest 
Subcutaneous emphysema and pleural effusions may obscure visualization, and patients with these conditions should be examined with thoracic CT whenever possible
In intubated patients, the lack of pleural sliding and the absence of comet-tail artifacts do not always indicate a pneumothorax; mainstem intubation and poor ventilation should be considered in the differential (with mainstem intubation, most frequently on the right side, the other lung collapses completely, so that no lung sliding will be seen on the collapsed side)
It is often helpful to compare the 2 sides of the chest to each other; however, pneumothoraces may also occur bilaterally
The use of ultrasonography to detect pneumothoraces was first described in 1987.  Since then, numerous studies have been published describing this application of ultrasonography and documenting its significantly better sensitivity and specificity in comparison with chest radiography. [21, 22, 23] According to the review by the ILC-LUS, lung ultrasound may be more effective at ruling out pneumothorax than ruling in the diagnosis, as compared to supine anterior chest radiography.  Although ultrasonography is widely used in emergency and critical care settings for rapid diagnosis of potentially life-threatening conditions, its utility and reliability remain controversial among some specialists. 
In a 2009 analysis of all published papers on the topic, the overall sensitivity of transthoracic ultrasonography for the diagnosis of pneumothorax ranged from 58.9% to 100%, and the specificity ranged from 94% to 100%.  A standardized technique for performing the procedure and specific sonographic signs indicating a pneumothorax have been described and validated in the literature, and these have been incorporated into the core curriculum in emergency medicine. 
A portable ultrasound machine with a high-frequency (5-10 MHz) linear probe is typically used. A curvilinear, low-frequency probe (2-5 MHz) is employed by some sonographers; although it is more difficult to use accurately, it allows the entire E-FAST examination to be performed without any switching of probes. Most studies evaluating the sensitivity of bedside ultrasonography in detecting pneumothoraces have used high-frequency linear probes, which are the standard recommendation for novice practitioners.
Sedation is typically not required for bedside ultrasonography. The patient should be placed in a supine position (see the image below). If it is clinically necessary, the patient may also be examined while sitting upright, but this positioning may reduce the sensitivity of ultrasonography for detecting small pneumothoraces.
Ultrasonographic evaluation for pneumothorax
Examination should begin with the patient in a supine position and at the most superior portion of the chest, which should correspond to the least gravitationally dependent area of the thorax. This is usually in the third or fourth intercostal space in the mid-clavicular line. Although there is no absolute standard, examination of 2 or 3 interspaces is generally recommended for a complete examination. Increasing the number of interspaces evaluated on thoracic ultrasound will increase the sensitivity of the exam. 
The ribs are identified; these will appear hyperechoic, and their acoustic shadows will appear as hypoechoic rays extending from the ribs. The interspace between the 2 ribs is used as a fixed anatomic landmark during the examination. Next, the pleural line is identified; this is a hyperechoic line found at the inferior border of the space between the 2 ribs.
Ultrasonographic findings suggestive of a pneumothorax have been thoroughly researched and are well described in the literature. [27, 28, 29] The following is a summary of the most routinely used diagnostic signs.
Findings suggestive of pneumothorax
The presence of a pneumothorax is characterized by the following findings: (1) the absence of pleural (lung) sliding; (2) the absence of so-called comet-tail artifacts, also referred to as B-lines; (3) the absence of a lung pulse; and (4) the presence of one or more lung points. [30, 17] The so-called lung point is a somewhat recently described sign that, although difficult to identify, is pathognomonic for a pneumothorax and can be used to measure the size of the pneumothorax.
Absence of pleural sliding and lung pulse
In normal patients, the pleural line represents both the parietal and visceral layers of the pleura, and back-and-forth sliding of that line is easily visualized during the respiratory cycle. The lung pulse refers to the subtle movement of the visceral upon the parietal pleural along with cardiac beats.  In the presence of a pneumothorax, air accumulates between the 2 layers and blocks transmission of sound waves, so that the sliding is not visualized. This phenomenon can be seen in real time in the 2-D mode but is more easily visualized by viewing a still image in M-mode (motion mode; see the video below).
The appearance of normal lung has been described as the seashore sign (see the first image below). This term refers to the change in appearance between soft tissue and lung, divided by the pleural line, a change resembling that between sand and sea waves. In the presence of a pneumothorax, this demarcation is lost, and the appearance on M-mode imaging is described as the stratosphere sign (see the second image below).
Absence of comet tails
Comet tails are artifacts that are thought to be created when ultrasound waves bounce off the interface between the apposing visceral and parietal layers of the pleura. They appear as hypoechoic vertical raylike projections off the pleural line and are parallel to the rib shadows previously noted (see the image below).
The presence of air in the pleural space inhibits the propagation of sound waves, preventing the appearance of comet tails. Comet-tail artifacts may be hard to visualize, requiring considerable patience; the transducer should be kept in a fixed location on the chest as the dynamic lung is observed sliding throughout the expiratory cycle. The presence of comet tails is 60% specific for the absence of pneumothorax. Combined with the absence of lung sliding, the absence of comet tails has a negative predictive value of 100% and a specificity of 96.5%.
The lung point is a more recently described sign that is pathognomonic for the presence of a pneumothorax.  The lung point is the actual point at which the normal lung pattern (ie, lung sliding and comet-tail artifacts) is replaced by a pattern consistent with a pneumothorax (ie, no lung sliding and no comet-tail artifacts).
The lung point is a dynamic sign and, like the comet tails, can be visualized only by keeping the transducer in a fixed position and watching the pleura throughout the respiratory cycle. This point should be sought by longitudinally scanning the anterior, lateral, and posterior positions of the chest wall. Although it is the most specific sign of pneumothorax, it is also the hardest to visualize and may require an experienced operator to locate. Finding both transition zones (from normal lung to pneumothorax and then back again) allows calculation of pneumothorax size.
Complications of Procedure
Bedside ultrasonography for pneumothorax has no complications, aside from incorrect diagnosis of the presence or absence of a pneumothorax without correct identification of associated signs (or the lack thereof). If the ultrasound study is inconclusive, a definitive study with thoracic CT scanning is recommended; this remains the criterion standard for identifying pneumothoraces.