Decision Making in Emergency Critical Care

SECTION 3 - Critical Care Ultrasonography

7
Pulmonary Ultrasonography

Feras Khan and Anne-Sophie Beraud

BACKGROUND

Over the past decade, bedside point-of-care ultrasonography, or ultrasound (US), has become an indispensable tool in critical care and emergency medicine. It is an efficient and effective diagnostic aid myriad conditions and has improved procedure safety in both the emergency department (ED) and intensive care units (ICUs).1 The American College of Emergency Physicians (ACEP) recommends that all emergency medicine residents train to proficiency in emergency US.2 Lung US the subject of this chapter, is fast becoming an integral component of point-of-care US for both intensivists and emergency physicians. First developed in European ICUs, lung US has proven to be highly useful in detecting disease processes including pneumonia, pneumothorax, pleural effusions, and pulmonary edema.3 With recent advances in technology, point-of-care US can now be performed at bedside with relatively small devices. This allows physicians to make decisions quickly and safely—without having to transport the patient out of a monitored setting—and has helped minimize computed tomography (CT) use and associated patient exposure to ionizing radiation. In 2012, the first evidence-based guidelines for point-of-care lung US were published in order to standardize definitions for a variety of lung pathologies.4

PROBE SELECTION, TECHNICAL EQUIPMENT, AND SCANNING TECHNIQUE

Transducer Selection

Lung US can be performed with three types of US transducers: linear (usually used for vascular access or nerve blocks), phased array (“cardiac”), or convex (“abdominal”). Because of its high frequency (7.5 to 10 MHz), the linear probe is preferred for analyzing superficial anatomy such as the pleura as well as individual rib interspaces. The linear probe, however, does not allow deep penetration to visualize deeper structures, such as the lungs themselves; better suited for this are the phased-array (2 to 8 MHz) and the convex probes (3.5 MHz).

Imaging Modalities

The US transducer generates US waves that are reflected back to the transducer. These returning waves generate a signal that is determined by the difference in the acoustic impedance of the tissues encountered.5 There are two US modes commonly used for lung imaging. The first, B-mode (brightness mode), generates a 2D image. The second, M-mode (motion mode), displays images in relation to elapsed time (one axis showing the depth of the image-producing interface and the other showing time) (Fig. 7.1). M-mode allows recording of motion of the interface toward and away from the transducer. The use of each mode is discussed in detail in the sections below. When performing a bedside lung US exam, all preset filters should be turned off to allow lung artifacts to appear. The probe indicator points cephalad in all exams.

FIGURE 7.1 Seashore sign. This image is taken from the 3rd intercostal space in the midclavicular line using a convex probe in M-mode. The granular appearance of the lung creates the “seashore” sign. The arrow indicates the pleural line.

Imaging Technique

Prior to use, both machine and probe should be thoroughly cleaned with disinfectant to limit contamination and nosocomial infection spread.6 The patient is typically imaged in the supine position. For patients in a critical care setting, it may be difficult to obtain true posterior views. In these patients, a protocol using the anterior and lateral chest walls has been described (Fig. 7.2) that images two interspaces (2nd and 5th) along the midclavicular line and at the midaxillary line.7 This approach allows the clinician to quickly assess eight lung zones. For a more thorough examination in stable patients, the probe should be advanced longitudinally and transversely along the 2nd, 3rd, and 4th and 5th intercostal interspaces.

FIGURE 7.2 Lung zones. The four lung zones that are shown on this patient indicate areas that should be included when performing an ultrasound examination.

Training Requirements

No consensus exists regarding the number of supervised US exams needed for a clinician to achieve proficiency in lung US. The ACEP has proposed that 25 to 50 studies be reviewed by a qualified ultrasonographer in order to demonstrate competence in a specific exam (e.g., pulmonary, cardiac).

NORMAL SONOGRAPHIC THORACIC ANATOMY

The lung parenchyma is normally filled with air, which has very low acoustic impedance and is therefore not detected on ultrasonography. Pulmonary disease processes result in changes to the air–fluid interface in the lung. These changes generate unique US patterns, or artifacts, which can help identify a variety of conditions including pleural effusions, pneumothorax, pneumonia, and alveolar–interstitial syndrome (AIS).

Lung Sliding

The lung can be visualized in the intercostal spaces, which delineate US windows between each rib. The parietal pleura lines the thoracic wall, covers the superior surface of the diaphragm, and separates the pleural cavity from the mediastinum. The visceral pleura covers the surface of the lung. The pleural space is a virtual space between the parietal and the visceral pleura.

In the intercostal spaces, the pleural line is situated below the subcutaneous tissue, about 0.5 to 2 cm from the skin depending on chest wall thickness. It is a horizontal and thin structure that appears intensely hyperechoic on US imaging (Fig. 7.3). In normal healthy lungs, US imaging will demonstrate “lung sliding” of parietal pleura against the visceral pleura during the respiration.

FIGURE 7.3 A-lines/pleural line. This image uses a vascular probe and shows a typical A-line (long arrow) as well as the pleural line (short arrow).

A-Lines

“A-lines” are hyperechoic horizontal artifacts seen in healthy lungs and represent repetition artifact generated by the pleural line (Fig. 7.3). Importantly, A-lines can also be seen in patients with pneumothorax as described below. In M-mode, healthy lungs will demonstrate a “seashore” sign (Fig. 7.1). This description captures the “wavelike” pattern produced by normal pleural line movement coupled with the “sandy beach” or granular appearance generated by lung parenchyma.

B-Lines

A “B-line” is a reverberation artifact with the following properties:8

1.A vertical comet-tail artifact

2.Arises from the pleural line

3.Well defined

4.Hyperechoic

5.Long (does not fade)

6.Erases A-lines

7.Moves with lung sliding

One or two B-lines may be seen in the dependent lung zones in the normal lungs.9 A large number of B-lines are pathologic (Fig. 7.4A and B) and will be described below in the AIS section. These artifacts are also called “comet tails.”

FIGURE 7.4 A: B-lines: alveolar–interstitial syndrome. This image shows three B-lines (arrows) in an interspace characteristic of AIS. There is a varying degree of thickness on each B-line. Image used courtesy of Dr. Darrell Sutijono. B: B-lines. This image shows 7 B-lines (arrows). Image used courtesy of Dr. Liz Turner.

CLINICAL CONDITIONS

Pneumothorax

A pneumothorax can be traumatic or nontraumatic in etiology. A large pneumothorax, especially if causing hemodynamic compromise, may require emergent treatment. Chest radiography (CXR) is the imaging modality most commonly used to evaluate pneumothorax; it has, however, been repeatedly demonstrated to be poorly sensitive (36% to 48%) for this condition.1013 CT remains the gold standard for the diagnosis of pneumothorax, but is time consuming and requires that the patient be transported out of the acute care setting.

Recent studies have demonstrated bedside lung US to have similar sensitivity to CT for the detection of pneumothorax,10 making it ideal for the evaluation of hemodynamically unstable or ventilated patients in whom there is concern for lung collapse. A number of lung US findings exist that can help confirm or exclude pneumothorax. The presence of lung sliding has a negative predictive value for pneumothorax of close to 99%.14 The lung sliding examination should be performed with the patient in a supine position allowing air to rise to the most anterior part of the chest and should be evaluated at several points on the anterior and lateral chest wall. The presence of B-lines also rules out pneumothorax with a negative predictive value of 98% to 100%.1517 If a pneumothorax exists, a “stratosphere” or “bar-code” sign will replace the normal “seashore” sign seen using M-mode. The “stratosphere” sign is caused by air interrupting the normal pleural line reflection (Fig. 7.5). A “lung point,” which represents the interface of normal lung next to an area of pneumothorax (Fig. 7.6A and B), may also be observed and is the most specific indicator for this condition.18 Lung point is best visualized using M-mode with the probe held in the middle of the interspace transecting the lung—lung sliding will be seen on the part of the pleural line with intact lung and then will disappear in the area of pneumothorax. Finally, the absence of the “lung pulse” has also been described as a sign of pneumothorax.18 The lung pulse refers to the rhythmic movement of the visceral and parietal pleural in step with the heart rate that is seen in normal healthy lungs. A combination of absent lung sliding and the presence of A-lines results in a sensitivity of 95% and a specificity of 94% for pneumothorax.16 Guidelines recommend imaging at least four zones on each lung field to identify these findings.

FIGURE 7.5 Stratosphere sign. This image uses a linear probe in M-mode and shows the “stratosphere sign,” indicating a pneumothorax. Also known as the “bar-code sign.”

FIGURE 7.6 A: Lung point. This image uses a linear probe at the fourth right-sided intercostal space and shows the exact point at which a pneumothorax begins (arrow). Part (y) of the image shows normal sliding lung, while part (x) demonstrates absence of lung sliding consistent with a pneumothorax. Image used courtesy of Dr. Liz Turner. B: Lung point in M-mode. This image shows the alternating seashore (y) and stratosphere sign (x) in a patient with a pneumothorax demonstrating “lung point.” Image used courtesy of Dr. Darrell Sutijono.

In trauma, lung US has become a part of the Focused Assessment with Sonography for Trauma (FAST) as described in the extended FAST (E-FAST) protocol.19 In this study of 225 trauma patients, a trained attending trauma surgeon using a 5- to 10-MHz linear transducer performed all US examinations. The protocol required imaging over the anteromedial chest at the second interspace at the midclavicular line and at the anterolateral chest wall near the 4th or 5th intercostal space at the midaxillary line. The absence of lung sliding and B-lines (comet tails) corresponded to an US diagnosis of pneumothorax. Lung US was found to be more sensitive than CXR alone (48.8% vs. 20.9%) with similar specificities (99.6% and 98.7% respectively). Compared with a composite standard (CXR, chest and abdomen CT, clinical course, and clinical interventions), the sensitivity of E-FAST was 58.9% with a specificity of 99.1%. The low sensitivity of US in this study was attributed to the high rate of occult pneumothorax or partial pneumothorax. Nevertheless, the study highlights the importance of lung US as an integral part of trauma assessment and the need to incorporate lung US into the FAST protocol (E-FAST). In a 2012 systematic review of eight primarily trauma studies with a total of 1,048 patients, lung US was found to have superior sensitivity to CXR (90.9% vs. 50.2%) but similar specificity (98.2% vs. 99.4%) for detection of pneumothorax.20

Alveolar–Interstitial Syndrome

AIS describes a group of conditions—including pulmonary edema, interstitial pneumonia, and pulmonary fibrosis—that demonstrate similar findings on lung US.9 Specifically, the normal air–fluid interface responsible for the artifacts seen on US imaging is shifted toward the fluid side. Cardiogenic pulmonary edema is the most common source of this change and is characterized on US by the presence of multiple B-lines (Fig. 7.4A and B). B-lines correspond to interlobular septal thickening on CT imaging, which denotes pulmonary vascular congestion.9 B-lines are thought to be reverberation artifacts produced as the US beam strikes these congested areas.

To identify these findings, the US should be in B-mode, and at least eight lung zones should be imaged. A lung zone is considered “positive” when three or more B-lines are present.4 Two or more positive zones bilaterally are required to meet the US definition of AIS (it is not uncommon to have one or two B-lines in normal patients in dependent lung areas).4 Bilateral, diffuse B-lines have been demonstrated to have a specificity of 95% and a sensitivity of 97% for the diagnosis of pulmonary edema.9 In this study, AIS was confirmed by CXR in 86 of 92 patients who had diffuse B-lines in all lung fields. In another study of 300 ED patients presenting with shortness of breath, 77 had radiologic evidence of diffuse AIS detected by lung US with a sensitivity and specificity of 85.7% and 97.7%.21

The ability of lung US to predict the presence of pulmonary edema has been compared to extra-vascular lung water (EVLW) calculations by the PiCCO system (Pulse index Contour Continuous Cardiac Output, Pulsion Medical Systems, Germany) and to pulmonary artery catheter (PAC)–derived wedge pressure.22 Although only 20 patients were enrolled in this study, positive linear correlations were found between a total B-line score and EVLW (r = 0.42) and PAC wedge pressure (r = 0.48). It should be noted that patients with lung disease were excluded from this study and that conditions such as pulmonary fibrosis or acute respiratory distress syndrome (ARDS) can present with B-lines as well.

Lung US has also been used to monitor improvement in patients with varying degrees of pulmonary congestion/edema. In a study of 40 patients undergoing routine dialysis, B-lines were recorded pre- and postdialysis.23 In 34 out of 40 patients, the number of B-lines underwent statistically significant reduction from predialysis to postdialysis. The study suggests that quantification of B-lines could potentially be used to complement daily patient weights in monitoring improvement in pulmonary congestion/edema. Lung US has also proven useful in measuring B-line improvement in acute decompensated heart failure.22

Pneumonia/Lung Consolidation

Pneumonia is a common diagnosis in both ED and ICU patients. Using US, lung consolidations have been described as a subpleural area with tissue-like hypoechoic texture8 (Fig. 7.7A and B) and can resemble the US appearance of the liver, a pattern called “hepatization.” Other US findings in patients with pneumonia include air bronchograms, comet-tail reverberation artifacts in a localized area, and a vascular pattern within the consolidation. Hyperechoic, linear, tubular artifacts within an isoechoic region suggest atelectasis.

FIGURE 7.7 A: Pneumonia. This image shows a hyperechoic area (arrow) corresponding to an air bronchogram with pneumonia (x). The lung begins to resemble the liver (y) on US, a pattern termed “hepatization.” There is also a pleural effusion (z). Image used courtesy of Dr. Liz Turner. B: Pneumonia. This image shows a hyperechoic area (arrow) that correlates to air bronchograms and pneumonia. The liver (x) and lung (y) are visible. Image used courtesy of Dr. Darrell Sutijono.

In a prospective study of 65 ICU patients, US was found to have a sensitivity of 90% and a specificity of 98% for pneumonia when compared to alveolar consolidation on CT.24 It should be noted that the ultrasonographers in this study were highly experienced and performed a thorough lung examination on each patient. In a separate study, lung US was used in 49 patients presenting to an ED with signs and symptoms of pneumonia.25 All patients received both an US examination and a CXR. If the CXR was negative for pneumonia, and the US positive, a confirmatory CT was performed. Thirty-two out of 49 patients were confirmed to have pneumonia, with US outperforming CXR in diagnostic accuracy, 96.9% versus 75%, respectively. Limitations of the study included its nonblinded design, the small number of patients studied, and the variable experience of the ultrasonographers. Due to the design of the study, false negatives of US may have been missed.

Pleural Effusion

Detection of Pleural Effusion

Lung US is well validated as a tool for the detection of pleural fluid. Using B-mode, the probe is positioned along the mid- to posterior axillary line on the lateral aspect of the chest wall. The diaphragm should be identified as well as the liver (Fig. 7.8). Pleural fluid appears as an anechoic area superior to the diaphragm. The “sinusoid sign,” which demonstrates variation in the interpleural distance with each respiratory cycle, has also been used as an indicator of pleural effusion.26 A systematic review of four lung US studies, using chest CT as a gold standard, demonstrated US to have a mean sensitivity of 93% and specificity of 96% for detecting pleural effusion.27

FIGURE 7.8 Pleural effusion. This image uses a convex probe in the midaxillary line and shows a pleural effusion (y). The collapsed lung is seen under the effusion (x) and the diaphragm can be seen (arrow) with the liver underneath.

In trauma patients, lung US has been used to rapidly detect hemothorax. In one study, 61 trauma patients underwent a standard FAST examination with two additional views used to evaluate the thoracic cavity laterally; the sensitivity and specificity of US for hemothorax were 92% and 100%, respectively.28

Quantification of Pleural Effusion

The amount of pleural fluid can also be quantified by US. A traditional posteroanterior CXR can identify effusions as small as 175 mL.29 US has been able to detect as little as 20 mL of pleural fluid.30 In a study of patients with known pleural effusions, 81 US examinations were performed to quantify amount of pleural fluid.31 Patients were examined in a supine position with mild trunk elevation of 15 degrees and with the probe placed in the posterior axillary line perpendicular to the body axis. The maximal distance between the visceral and parietal pleural (Sep) in end-expiration was measured, which allowed calculation of the estimated volume (V). The volume estimated by US was compared to the volume of fluid obtained after thoracentesis:

A positive correlation was seen with both Sep and V (r = 0.72 and r2 = 0.52, respectively). The mean prediction of error of V was 158.4 ± 160.5 mL. No complications were noted during US-guided thoracentesis in this study.

Characterization of Pleural Effusion

US may also be able to help identify subtypes of pleural effusion (transudate or exudate). Pleural fluid patterns are characterized as anechoic, complex nonseptated, and complex septated.32 In a study of 320 patients undergoing both thoracentesis and lung US, transudates were anechoic in appearance (the type of effusion was determined by both chemical analysis of pleural fluid and clinical evaluation (i.e., evaluation of ascites, peripheral edema)).33 Complex septated or nonseptated effusions were always found to be exudates. At this time, however, US should not be substituted for thoracentesis and definitive chemical evaluation of pleural fluid.

Ultrasound-Guided Thoracentesis

US-guided thoracentesis is both safe and efficient. A study of 67 patients with pleural effusion compared US-guided to blind thoracentesis; the use of US prevented organ puncture in 10% of cases and increased identification of accurate puncture sites by 26%.34 When performing an US-guided thoracentesis, the clinician should place the patient in a supine position and use a convex probe in the midaxillary line to detect the effusion. Prior to insertion of a needle, relevant anatomic landmarks should also be identified, including the rib space, diaphragm, and depth of effusion. The needle should be directed superior to the rib to avoid the neurovascular bundle.

Acute Respiratory Distress Syndrome/Acute Lung Injury

Using traditional CXR, ARDS can appear similar to AIS, cardiogenic pulmonary edema, and pulmonary fibrosis. Recently, attempts have been made to identify a lung US pattern unique to ARDS. A recent study compared the lung US findings of 58 patients, in which 18 met criteria for acute lung injury (ALI)/ARDS (based on the American–European Consensus Conference diagnostic criteria) and 40 had acute pulmonary edema.35 In ALI or ARDS, the lung examination showed areas that were spared of B-lines, while in cardiogenic pulmonary edema, the distribution of B-lines was more diffuse. ALI/ARDS patients also had more posterior lung consolidations with typical air bronchogram findings and a pleural line with reduced “sliding” and thickened and coarser appearance. A recently published guideline for lung US in ARDS lists the following associated findings4:

1.Anterior subpleural consolidation

2.Absence or reduced lung sliding

3.Sparred areas of normal parenchyma

4.Pleural line abnormalities

5.Nonhomogenous distribution of B-lines

To date, these findings have not been validated in a prospective study and should not replace the traditional diagnostic approach to APE or ARDS/ALI.

BEDSIDE LUNG ULTRASOUND IN EMERGENCY (BLUE) PROTOCOL

A recent major study used a lung US-based algorithm (Fig. 7.9) to categorize shortness of breath in patients presenting to the ICU and compared results to final ICU diagnosis.7 Rare causes or uncertain diagnosis were excluded from the study (<2%). Six lung zones were analyzed for A- or B-lines, lung sliding, and alveolar consolidation. An US of both lower extremities was also performed for deep venous thrombosis.

FIGURE 7.9 Blue Protocol Algorithm. PLAPS, posterior/lateral alveolar or pleural syndrome. From Lichtenstein D, Meziere G. Relevance of lung US in the diagnosis of acute respiratory failure: the BLUE protocol. Chest. 2008;134:117–125.

A predominantly A-line pattern was seen in patients with chronic obstructive lung disease (89% sensitivity and 97% specificity). Multiple anterior diffuse B-lines with lung sliding were seen in patients with pulmonary edema (97% sensitivity and 95% specificity). A normal lung examination and deep venous thrombosis on lower extremity US indicated pulmonary embolism (81% sensitivity and 99% specificity). Lack of lung sliding plus A-lines and a lung point indicated pneumothorax (81% sensitivity and 100% specificity). Anterior and posterior consolidations, anterior asymmetric interstitial patterns, or anterior diffuse B-lines with abolished lung sliding indicated pneumonia (89% sensitivity and 94% specificity). The postero-lateral alveolar and/or pleural syndrome (PLAPs) is an entity seen on US that usually indicates pneumonia and was used in the BLUE protocol algorithm. The postero-lateral segment is found on the lower lateral part of the chest wall and is positive if there is evidence of effusion and areas of consolidation. These patterns correctly identified the final diagnosis in 90.5% of cases. It should be noted that 41 patients were excluded from the study for the following reasons: multiple diagnoses, no final diagnosis, and “rare” causes such as interstitial lung disease or massive pleural effusion.

Limitations

Although bedside US is easy to perform in most patients, certain scenarios pose challenges. Obese patients with thick chest walls generate suboptimal images and limit artifact formation. Inadequate imaging may also occur in patients with subcutaneous emphysema or chest tubes and in trauma or postsurgical patients with large dressings in place. Adequate training is also essential in allowing a sonographer to recognize findings with confidence.37

CONCLUSION

The use of lung US has increased dramatically since its introduction in the 1980s. Portable US machines permit safe, cost-effective, and rapid detection of a variety of lung pathologies at bedside, while minimizing the need to transport patients away from the critical care setting. Lung US may help patients avoid harmful ionizing radiation exposure associated with repetitive CXR or CT.36 Potential future uses of lung US include predicting successful extubation from the ventilator, evaluating recruitment maneuvers in mechanically ventilated patients, and differentiating ARDS from typical AIS patterns.

LITERATURE TABLE

REFERENCES

1.National Institute of Clinical Excellence. Final Appraisal Determination: Ultrasound Locating Devices for Placing Central Venous Catheters. National Institute of Clinical Excellence; 2002.

2.American College of Emergency Physicians. Policy statement. Emergency ultrasound guidelines. Ann Emerg Med. 2009;53:550–570.

3.Lichtenstein D, Axler O. Intensive use of general ultrasound in the intensive care unit. Prospective study of 150 consecutive patients. Intensive Care Med. 1993;19(6):353–355.

4.Volpicelli G, Elbarbary M, Blaivas M, et al. International evidence-based recommendations for point-of-care lung ultrasound. Intensive Care Med 2012;38:577–591.

5.Chan VWS. Ultrasound Imaging for Regional Anesthesia. 2nd ed. Toronto, ON: Toronto Printing Company; 2009.

6.Fowler C, McCracken D. US Probes: risk of cross infection and ways to reduce it-comparison of cleaning methods. Radiology. 1999;213:299–300.

7.Lichtenstein D, Meziere G. Relevance of lung ultrasound in the diagnosis of acute respiratory failure: the BLUE protocol. Chest. 2008;134:117–125.

8.Lichtenstein DA. Whole Body Ultrasonography in the Critically Ill. Berlin, Germany: Springer-Verlag; 2010.

9.Lichtenstein D, Meziere G, Biderman P, et al. The comet-tail artifact. An ultrasound sign of alveolar-interstitial syndrome. Am J Respir Crit Care Med. 1997;156:1640–1646.

10.Rowan KR, Kirkpatrick AW, Liu D, et al. Traumatic pneumothorax detection with thoracic US: correlation with chest radiography and CT- initial experience. Radiology. 2002;225:210–214.

11.Neff MA, Monk JS, Peters K, et al. Detection of occult pneumothoraces on abdominal computed tomographic scans in trauma patients. J Trauma. 2000;49:281–285.

12.Tocino IM, Miller MH, Frederick PR, et al. CT detection of occult pneumothoraces in head trauma. Am J Roentgenol. 1984;143:987–990.

13.Rhea JT, Novelline RA, Lawrason J, et al. The frequency and significance of thoracic injuries detected on abdominal CT scans of multiple trauma patients. J Trauma. 1989;29:502–505.

14.Blaivas M, Lyon M, Duggal S. A prospective comparison of supine chest radiography and bedside ultrasound for the diagnosis of traumatic pneumothorax. Acad Emerg Med. 2005;12:844–849.

15.De Luca C, Valentino M, Rimondi M, et al. Use of chest sonography in acute-care radiology. J Ultrasound. 2008;11:125–134.

16.Lichtenstein DA, Meziere G, Lascols N, et al. Ultrasound diagnosis of occult pneumothorax. Crit Care Med. 2005;33:1231–1238.

17.Soldati G, Testa A, Pgnataro G, et al. The ultrasonographic deep sulcus sign in traumatic pneumothorax. Ultrasound Med Biol. 2006;32:1157–1163.

18.Lichtenstein D, Meziere G, Biderman P, et al. The “Lung Point”: an ultrasound sign specific to pneumothorax. Intensive Care Med. 2000;26:1434–1440.

19.Kirkpatrick AW, Sirois M, Laupland KB, et al. Hand-held thoracic sonography for detecting post-traumatic pneumothoraces: the extended focused assessment with sonography for trauma (EFAST). J Trauma. 2004;57:288–295.

20.Alrajhi K, Woo MY, Vaillancourt C. Test characteristics of ultrasonography for the detection of pneumothorax: a systematic review and meta-analysis. Chest. 2012;141(3):703–708.

21.Volpicelli G, Caramello V, Cardinale L, et al. Bedside ultrasound of the lung for the monitoring of acute decompensated heart failure. Am J Emerg Med. 2008;26(5):585–591.

22.Agricola E, Bove T, Oppizzi M, et al. “Ultrasound comet-tail images”: a marker of pulmonary edema: a comparative study with wedge pressure and extravascular lung water. Chest. 2005;127(5):1690–1695.

23.Noble V, Murray A, Capp R, et al. Ultrasound assessment for extravascular lung water in patients undergoing hemodialysis. Chest. 2009;135:1433–1439.

24.Lichtenstein DA, Lascols N, Meziere G, et al. Ultrasound diagnosis of alveolar consolidation in the critically ill. Intensive Care Med. 2004;30(2):276–281.

25.Parlamento S, Copetti R, Di Bartolomeo S. Evaluation of lung ultrasound for the diagnosis of pneumonia in the ED. Am J Emerg Med. 2009;27(4):379–384.

26.Lichtenstein D, Hulot JS, Rabiller A, et al. Feasibility and safety of ultrasound-aided thoracentesis in mechanically ventilated patients. Intensive Care Med. 1999;25:955–958.

27.Grimberg A, Shigueoka DC, Atallah AN, et al. Diagnostic accuracy of sonography for pleural effusion: systematic review. Sao Paulo Med J. 2010;128(2):90–95.

28.Brooks A, Davies B, Smethhurst M, et al. Emergency ultrasound in the acute assessment of haemothorax. Emerg Med J. 2004;21(1):44–46.

29.Webb WR, Higgins CB. Thoracic Imaging: Pulmonary and Cardiovascular Radiology. Philadelphia, PA: Lippincott Williams & Wilkins; 2004.

30.Rothlin MA, Nat R, Amgwerd M, et al. Ultrasound in blunt abdominal and thoracic trauma. J Trauma. 1993;34:488–495.

31.Balik M, Plasil P, Waldauf P et al. Usefulness of ultrasonography in predicting pleural fluid in mechanically ventilated patients. Intensive Care Med. 2006;32:318–321.

32.Marks WM, Filly RA, Callen PW. Real-time evaluation of pleural lesions: new observations regarding the probability of obtaining free fluid. Radiology. 1982;142:163–164.

33.Yang PC, Luh KT, Chang DB, et al. Value of sonography in determining the nature of pleural effusion: analysis of 320 cases. AJR Am J Roentgenol. 1992;159(1):29–33.

34.Diacon A, Brutsche M, Soler M. Accuracy of pleural puncture sites: a prospective comparison of clinical examination with ultrasound. Chest. 2003;123:436–441.

35.Copetti R, Soldati G, Copetti P. Chest sonography: a useful tool to differentiate acute cardiogenic pulmonary edema from acute respiratory distress syndrome. Cardiovasc Ultrasound. 2008;6:16.

36.Brenner D, Hall E. Computed tomography—an increasing source of radiation exposure. N Engl J Med. 2007;357:2277–2284.

37.Mayo PH, Beaulieu Y, Doelken P, et al. ACCP/La Societe de Reanimation de Langue Francaise statement on competence in critical care ultrasonography. Chest. 2009;135:1050–1060.