In order to recognize a pneumothorax, it is first important to note the following normal anatomic structures and their sonographic appearance (video 1).
Video 1: Normal anatomy
Figure 1: Probe positionThe most common initial test for respiratory distress is chest x-ray. Recently, growing evidence has shown that PoCUS can reliably detect thoracic pathology with equal if not better sensitivity than chest x-ray (CXR). In the case of pneumothorax PoCUS is significantly more sensitive than CXR. It also has the advantage of being performed at the bedside more quickly than CXR. Computed tomography offers little advantage over PoCUS, is impractical and comes with the cost of significant radiation.
Traditionally it was thought that ultrasound was not a useful modality for investigating lung pathology because air scatters ultrasound waves. There has been a growing body of evidence in recent years contradicting this practice. In fact, ultrasound images of normal lung show predictable artifacts that are disturbed in disease states and can be readily differentiated by ultrasound.
A recent meta-analysis comparing PoCUS to computed tomography in adults with pneumothorax found that in the hands of trained users PoCUS had a sensitivity of 91% and specificity of 98%. The same study found supine CXR had a sensitivity of only 50% [1]. While upright CXR may be more sensitive, it has never been formally compared to CT and the few studies available have found a sensitivity of 83-84% [2,3]. While these are mainly adult studies, PoCUS maintains its superior test characteristics in the neonatal population with novice sonographers who received one hour of training with sensitivities and specificities nearing 100% suggesting it is a useful tool throughout the lifespan [4,5,6].
Ultrasound has the additional benefits of being free from ionizing radiation and quicker to perform at the bedside. In a study of PoCUS versus CXR for neonatal pneumothorax ultrasound took on average 5 minutes versus 19 minutes with portable x-ray [4].
PoCUS is safe, timely and effective for the diagnosis of pneumothorax.
Supine CXR showing pneumothoraxPoCUS:
Sensitivity 91%
Specificity 98%
Supine CXR:
Sensitivity 50%
Specificity 99%
PoCUS vs CXR in trained users [1]
1. Grimberg et al. Diagnostic accuracy of sonography for pleural effusion: systematic review. Sao Paulo Med J (2010) 128;2: 90-95.
2. Asthon-Cleary, DT. Is thoracic ultrasound a viable alternative to conventional imaging in the critical care setting? BJA (2013) doi:10.1093/bja/aet076
3. Kurian et al. Comparison of Ultrasound and CT in the Evaluation of Pneumonia Complicated by Parapneumonic Effusion in Children. AJR (2009) DOI:10.2214/AJR.09.2791
4. Calder and Owens. Imaging of parapneumonic pleural effusions and empyema in children. Ped Rad (2009) DOI: 10.1007/s00247-008-1133-1
5. Volpicelli et al. International evidence-based recommendations for point-of-care lung ultrasound. Intensive Care Med (2012) DOI:10.1007/s00134-012-2513-4
6. Islam et al. The diagnosis and management of empyema in children: a comprehensive review from the APSA Outcomes and Clinical Trials Committee. J Ped Surg (2012) DOI: 10.1016/j.jpedsurg.2012.07.047
7. Peris et al. The Use of Point-of-Care Bedside Lung Ultrasound Significantly Reduces the Number of Radiographs and Computed Tomography Scans in Critically Ill Patients. Anesth&Analg (2010) DOI: 10.1213/ANE.0b013e3181e7cc42
8. Jones BP, Tay ET, Elikashvili I, Sanders JE, Paul AZ, Nelson BP, Spina LA, Tsung JW, Feasibility and Safety of Substituting Lung Ultrasound for Chest X-ray When Diagnosing Pneumonia in Children: A Randomized Controlled Trial, CHEST (2016), doi: 10.1016/j.chest.2016.02.643
While definitively characterizing effusions is out of the scope of PoCUS it is useful to be aware of how different types of effusion appear on ultrasound. As fluid becomes more complex (clotted blood, pus, adhesions) It becomes more echogenic. Fibrin stranding and loculations appear as echogenic lines of material traversing the fluid (figure 8). [3] If these are seen on PoCUS further imaging or consultation may be required to better characterize the effusion.
Figure 8: Complicated Effusion
While there have been many studies showing complex methods of estimating pleural effusion volumes, we will not review each as this is beyond the scope of PoCUS. Generally speaking, effusions are considered small when the anechoic fluid visualized is limited to the costodiaphragmatic recess and does not extend beyond the dome of the diaphragm. The decision to drain an effusion should be made based on clinical factors in the presence of an effusion, not on the size itself. Additionally, in pediatrics there are no validated, simple ways of estimating volume.
There are three characteristic features of a pleural effusion on ultrasound:
As fluid collects in the pleural space it can be directly visualized by ultrasound as an anechoic area. The presence of a pleural effusion results in the loss of normal lung artifact: mirror image of the liver if viewed via the abdomen or of A and B lines if viewed via the thorax. The appearance of the fluid itself can vary depending on the type of effusion.
Figure 6: Anechoic space
Given the fluid in a pleural effusion is not collecting in the lung itself but in the potential space between the chest wall and lung it has distinct anatomic boundaries including (figure 7):
Figure 7: Anatomic boundaries
It is important to note these anatomic landmarks to ensure the location of the fluid, especially prior to performing procedures.
Fluid is an efficient transmitter of ultrasound waves and thus a pleural effusion allows visualization of structures not normally seen. The diaphragm can be visualized with ease and in large effusions can be seen through its entire course. Additionally, the vertebral bodies previously obscured by aerated lung will be visible from the posterolateral approach. This is known as the “spine sign” (video 3).
Video 3: Spine Sign
Fluid collected in the pleural space is free-flowing in the absence of loculations and adhesions. It subject to the forces of a moving diaphragm, heart and lung. When viewing an effusion via ultrasound it one can note how the fluid conforms to the shape of moving structures (video 4).
Of note, since fluid separates the visceral and parietal pleura normal lung sliding and pleural motion will be eliminated at the level of the effusion as well.
Video 4: Dynamic Movement
The normal lung is air-filled. Because air scatters ultrasound waves, we use the presence of normal artefacts to assess the lung. Depending on the window used, the lung’s appearance may vary. When viewed directly, the lung refracts US waves, but reverberation artefact creates A lines and occasionally B-lines (figure 2).
Figure 3: A-lines
Figure 4: B lines
When viewed via the abdomen ultrasound waves are reflected by the diaphragm and create a mirror image effect: the lung appears as a liver-like organ above the diaphragm (figure 5). This artifact moves with respiration as the diaphragm moves up and down.
Figure 5: Mirror Artifact
In the absence of pathology, aerated lung fills the costophrenic angle and obscures the diaphragm in the nearfield. As the diaphragm descends with respiration, the air-filled lung crosses the screen and further obscures the diaphragm and structures behind. This is referred to as the curtain sign (video 2).
Video 2: Curtain sign
In order to recognize pleural effusions, it is first important to note the following normal anatomic structures and their sonographic appearance at the costophrenic interface.
Video 1: Normal anatomy at posterior axillary line