Chapter 12: Imaging of the Critically Ill Patient: Bedside Ultrasound

Ultrasound technology was introduced to the medical field in the early 1950s. It was not until 1970s that multiple specialties in medicine adopted ultrasound as a useful diagnostic tool. Slasky et al mentioned the importance of ultrasound in the ICU. He reported a 2-year retrospective study of ultrasound indications in the ICU and found that 64.4% of patients had abnormal findings and 52% of these patients' clinical course and therapy were modified by ultrasonography.¹ Other authors reported similar findings.^2,3,4

Bedside ultrasonography differs from the ultrasound studies done by consult services (ie, cardiology, radiology, etc). The intensivist uses real-time (image acquisition and interpretation are done at the same time) point-of-care ultrasonography (POC-US) to answer simple clinical questions that will change patient's management. One clear example of POC-US frequently used by emergency physicians and trauma surgeons is the focused assessment by sonography for trauma (FAST). This became standard of care and it is fully integrated into advanced trauma life support teaching. For intensivists, POC-US helps diagnose, evaluate response to treatment and guide procedures safely.

In this chapter we will cover the basics of image acquisition and clinical applications of POC-US.

Understanding the basic physics of how ultrasound images are generated will help interpreting images and identifying artifacts. The ultrasound frequency ranges within millions of Hertz (MHz). It is generated when electric current is applied to a piezoelectric crystal. These mechanical vibrations are transmitted through tissues; each tissue will have different impedance, generating echoes throughout. The same transducer, then receives and transduces into a grayscale image that is generated at 20 to 40 frames per second. This is perceived as continuous images on the screen. For deeper structures, low-frequency waves such as 1 to 5 MHz are preferred and for superficial structures, high-frequency waves such as 5 to 15 MHz are preferred. By convention, tissues like the liver and kidneys are isoechoic. Bone, stones, and other tissues that reflect more echoes are hyperechoic and fat, fluid, and blood have low impedance generating hypoechoic or anechoic images.

Basic understanding of acoustic artifacts can help with image interpretation; these are described in Table 12–1.

Training and Competency

Despite the potential benefit of POC-US, the use of ultrasound is not universal in critical care medicine. Benefits will only be appreciated if the operator is capable of acquiring and interpreting ultrasound images. Multiple international critical care societies had partnered to establish guidelines to provide competency in POC-US.^5,6 These guidelines are intended to create a framework toward an universal standard. For those intensivists interested in POC-US, the World Interactive Network Focused on Critical Ultrasound (WINFOCUS) and the American College of Chest Physician (ACCP) have well-developed training programs that consist of basic concepts of ultrasonography, hands-on image acquisition, and interpretation.

Competency for general critical care ultrasound (GCCU) and critical care echocardiography (CCE) are different, CCE competency is described in chapter 95 of this book.

It is a consensus that courses designed to teach GCCU and CCE should be at least 10 hours divided into lectures and didactic image-acquisition training. The numbers of examination to achieve competency is controversial.⁶

The basic requirements needed to achieve competency as per the ACCP allows the operator to foster self-directed learning (Table 12–2.)

To date, the only evidence-based recommendation for POC-US is for lung ultrasonography (LUS).⁷ A panel of experts voted for the level of evidence in each condition where lung ultrasonography is used. The level of evidence was classified as following: Level A: high-quality evidence and further research is very unlikely to change the estimated effect or accuracy of evidence. Level B: Moderate quality and further research is likely to have an important impact on estimated effect or accuracy and may change level of evidence. Level C: Low quality of evidence and further research is likely to change the estimated effect or accuracy of evidence (Table 12–3).

Lung Ultrasonography

Ultrasound Probe Selection and Orientation

The preferred probe to evaluate the thorax is a 3.5 to 5 MHz with a small footprint and the cursor pointing cephalad. If more superficial structures of the thorax are being investigated, a 7.5- to 10-MHz linear probe can be used.

The classical views of the chest are in the midclavicular line, midaxillary line, and posterior axillary or mid scapular line. However, scanning other parts of the chest may be indicated as long as good images are obtained for interpretation.

Clinical Implications and Description of Syndromes

In the hands of an experienced operator, LUS in the critically ill is superior to chest roentgenogram (CXR) and comparable to computerized tomography of the chest (CT-Chest) in several conditions.^8,9 It is important to understand the air-fluid ratio in the lung to understand and interpret LUS. The air-fluid ratio in pneumothorax (PTX) is pure air, no fluid; normal lung has air and very little fluid, the interstitial syndrome has air with more fluid; alveolar consolidation has little air but a lot of fluid and pleural effusion consists of pure fluid, no air. (Figure 12–1).

The only way to image the lung is in between the ribs (Figure 12–2). The lung has to be centered in the screen and the two rib shadows are seen on the side; this is referred to as the bat sign. The pleural line can be identified as a bright structure. Standard ultrasound probes cannot separate the visceral from the parietal pleura so only one line is seen. In a healthy lung, the pleural surfaces move against each other during the respiratory cycle causing a shimmering line, this is called lung sliding. Lung pulse is referred to the pleural line moving synchronously with each cardiac cycle. A normal aerated lung will show reverberations artifact from ultrasound reflection between the pleural line and the skin surface, these are called A-lines. In the presence of lung sliding, A-lines indicate normally aerated lung. In the absence of lung sliding, A-lines do not necessarily indicate normal lung and pneumothorax could be present.

The Interstitial Syndrome—B-lines represents thickened subpleural interlobular septa and/or ground-glass opacity,¹⁰ either of cardiogenic or noncardiogenic etiology. They are described as comet-tail artifact and must fulfill all of these requirements: (1) Starts from the pleural surface, (2) moves with lung sliding, (3) erases A-lines (4) reaches the bottom of the screen. If any one of the requirements is not met, it is not a B-line.

B-line artifact can be seen in interstitial syndromes (Table 12–4). A few B-lines in the bases of the lung are considered normal but more than three B-lines in a single field should be considered abnormal (Figure 12–3).

Diffuse bilateral B-line pattern with smooth pleura is consistent with cardiogenic interstitial edema.¹¹ Diffuse bilateral B-line pattern with an irregular pleura is usually seen in adult respiratory distress syndrome. Focal irregular pleura with B-lines can be interpreted as early or atypical pneumonia in an appropriate clinical setting.

The Alveolar Consolidation Syndrome—The alveolar consolidation syndrome refers to any condition that can fill the alveoli with fluid (Table 12–5). About 98.5% of alveolar consolidations reach the pleura, enabling ultrasound examination. In a supine patient, 90% of consolidations can be found in the most dependent area of the thorax.⁸ The tissue-like sign is an echoic pattern, with regular trabeculation reminiscent of an ill-defined liver (Figure 12–4). Ultrasound has a 90% and 98% sensitivity and specificity, respectively.⁸ In alveolar consolidation syndromes 2 types of air bronchograms can be seen, static and dynamic. Dynamic air bronchograms indicate the presence of air moving within the bronchus with each breathing cycle, ruling out absorptive atelectasis. A static air bronchogram indicates obstruction of the bronchus causing absorptive atelectasis.

Pleural Ultrasonography

PTX occurs in 6% of the ICU patients¹² and up to 30% of those can be misdiagnosed on conventional CXR owing to the supine position of the critically ill patient.¹³ The gold standard to diagnose PTX is CT-Chest; this involves transferring the patient with all the risks associated with it. Luckily ultrasound can help us identify PTX by scanning the anterior chest. Once the bat sign is identified, the absence of lung sliding plus a diffuse A-line pattern should prompt us to consider the diagnosis of PTX (Figure 12–5). Lung point indicates the point where the lung is touching the chest wall and moves with each respiratory cycle, this is 100% specific for PTX. The sensitivity of lung point for PTX detection is not 100% since this is usually absent in large PTX. It is important to mention that the presence of lung pulse or B-lines rule out PTX in 100% of the cases.

Another phenomenon described in left side PTX is the flickering image in the parasternal long axis view of the heart. The heart is seen in mid-diastole and disappears in midsystole. This is explained by the transient interposition of air between the chest wall and the heart, making it hard to visualize when it is empty and not in contact with the chest wall. This is called heart point (Figure 12–6).¹⁴

Pleural effusion is a very common finding in ICU patients.¹⁵ It collects in dependent areas unless loculated. By using the signs described below, ultrasound can achieve a sensitivity and specificity of 93%.⁹

A big pleural effusion is easily identified and must be distinguished from ascites. Using simple static and dynamic signs, we can tell them apart. The lung can be identified as a consolidated mass usually with air bronchograms, floating in fluid. This is called the jellyfish sign. The diaphragm should be identified above the liver and the spleen, if there is no jellyfish sign in a big effusion; the structure is likely to be the liver (or spleen), not the lung (Figure 12–7).

Ultrasound can also help identify the nature of the pleural effusion. A complex anechoic or hypoechoic space with septations most likely represents an exudative effusion (Figure 12–8).¹⁶ Sometimes this effusion can be seen as a “thick effusion” meaning a tissue-like echogenicity but with particles swirling around. This is called the plankton sign, which is seen in empyema and hemothorax (Figure 12–9).

Diaphragm Evaluation

This structure usually is easily visualized above the liver or spleen. There have been studies examining diaphragm thickness and strength. It is well known that thinning of the diaphragm starts within 18 hours of muscle inactivity.¹⁷ This was demonstrated by ultrasonography in 7 patients measuring the diaphragm using the 7.5- to 10-MHz linear probe placed in the midaxillary line to measure the apposition area (Figure 12–10).¹⁸ The utility of this should be further studied. Another promising use for ultrasonography of the diaphragm includes evaluation of its excursion during weaning from mechanical ventilation. Kim et al found that in patients with diaphragm dysfunction, defined when diaphragm excursion was less than 10 mm, weaning failure rate was higher when compared to patients with diaphragm excursion of more than 10 mm.¹⁹

Critical Care Echocardiography

Echocardiography, either transthoracic echocardiography or transesophageal is a very helpful tool in the critically ill patient. This will be discussed in detail in Chapter 94 of this book.

Abdominal Ultrasonography

Ultrasound Probe Selection and Orientation

The preferred ultrasound probe to evaluate the abdominal area will be a 3.5 to 5 MHz with a small footprint with the cursor pointing cephalad.

Clinical Implications of Abdominal Ultrasonography and Recognition of Important Structures

The intensivist should be capable of performing and interpreting simple abdominal ultrasonography. Detailed ultrasound examination of the abdominal organs is out of the scope of this book. It is our practice to perform abdominal POC-US paying attention to liver, spleen, kidney, bladder, gallbladder, aorta, and intestines.

Recognition of the Peritoneal Space

The parietal and visceral layers are in touch with each other and slide with each respiratory cycle; this is called gut sliding.

It is imperative to look for free fluid in an unstable patient. We have found FAST very useful in different situations outside the classic trauma population.

The peritoneal fluid can range from anechoic to hyperechoic, depending on the etiology. Transudates will have an anechoic nature, an hemoperitoneum can be visualized depending on the amount and time elapsed since bleeding started; it could be mildly hyperechoic with plankton sign or more hyperechoic if the blood had already organized. An exudative effusion causing infectious peritonitis could have septations. Ultrasound is more sensitive for detecting septations than CT scan.

Pneumoperitoneum (Figure 12–11) can be detected with ultrasound using the same concepts we have learned in pneumothorax. There are a couple of signs that will help us identify this condition: (a) Gut sliding: as described above, can rule out pneumoperitoneum. (b) Splanchnogram: it is when an abdominal organ can be seen with ultrasound, thus, ruling out pneumoperitoneum between the probe and the organ.

The presence of these 2 signs can rule out pneumoperitoneum accurately.

Kidney and Bladder

The probe should be oriented with the pointer cephalad and placed in the posterior axillary line under the costal border; once the liver or the spleen is identified, the kidney is easier to find.

Acute renal failure (ARF) is frequently encountered in the critically ill patient. The etiologies are classified into prerenal, renal, and postrenal. Ultrasound is useful as an initial screening tool.

The evaluation of the renal parenchyma by ultrasound is complex but we can briefly mention some concepts. ARF can be differentiated from chronic renal failure (CRF) because in the latter small kidneys with thin parenchyma and irregular borders are seen. Corticomedullary dedifferentiation can be seen in multiple conditions. In acute medical renal disease the renal parenchyma becomes more echogenic. The renal pelvises are well visualized under ultrasound. Chronic dilatation causes rounded renal calyces and the acute dilatation causes flatter renal calyces (Figure 12–12).

The bladder is very well visualized in the suprapubic area with the probe held transversely. A full bladder can be identified as an anechoic structure, if the patient has an urinary catheter, this can be easily visualized (Figure 12–13). Furthermore, ultrasound of the bladder can be used to check for urinary retention without using a catheter.

Ultrasound of the Abdominal Aorta

This is a very important structure to evaluate in an unstable patient and in someone who has been complaining of abdominal or flank pain. Abdominal aortic aneurysm (AAA) rupture has a high mortality rate (around 50%) and ultrasound has been shown to be a reliable method of evaluating AAA.

The thoracic aorta is better evaluated with a transesophageal echocardiogram or contrast computerized tomography. The abdominal aorta is easily visualized by transabdominal ultrasound.

The aorta is evaluated with the ultrasound probe placed in the midline of the abdomen and the cursor pointing toward the patient's right when a transverse view is desired and pointing toward the head when a longitudinal view is desired. The aorta should be studied from the epigastrium until the bifurcation to the iliac arteries (Figure 12–14). A diameter greater than 3 cm is considered aneurysmal. Ultrasound is not very sensitive to evaluate for aortic dissections but if a false lumen is seen, it is highly specific for such disease.

Gastrointestinal Tract

The esophagus is usually very difficult to visualize under ultrasound. It becomes visible and feasible for ultrasound assessment once it enters the abdomen to join the stomach. Ultrasound evaluation of the abdominal part of the gastrointestinal (GI) tract is very useful in the critically ill patient.

One observational study showed the utility of identifying gastric contents with ultrasound and performing suctioning prior to emergent endotracheal intubation.²⁰ It is thought that this can potentially decrease the chances of aspiration during intubation. The use of ultrasound had also been described for confirmation of nasogastric tube and Sengstaken-Blakemore tube positioning.^21,22,23

When evaluating the GI tract, it is important to observe the wall thickness that usually ranges around 2 to 4 mm. It is also very important to observe for gut sliding and peristalsis; the latter are permanent dynamic crawling contractions seen from the antrum of the stomach until the terminal ileus. The presence of peristalsis is rare in a true surgical abdomen.

Liver, Gallbladder, and Spleen

The liver is rarely a target in a critically ill patient despite its size. It acts as an acoustic window to access the heart, Morrison pouch and inferior vena cava analysis. However in some occasions, gross abnormalities can arise such as the following: (a) Liver abscess: yields a round heterogenous hypoechoic image within the regular hepatic echostructure easily identifiable. A highly hyperechoic image can represent microbial gas. (b) Portal gas: as a result of mesenteric infarction. Hyperechoic images disseminated within the liver parenchyma mainly in the periphery and resembling alveolar consolidation with air bronchograms. (c) Liver cirrhosis: yields a coarse and nodular pattern with atrophy (Figure 12–15). (d) Hepatic tumors: Metastatic tumors are usually multiple iso- or hypoechoic masses. However an echoic heterogeneous mass within a cirrhotic parenchyma suggests hepatoma. A simple biliary cyst is usually described as a simple anechoic structure with thin walls (Figure 12–16).

The gallbladder can be viewed during ultrasound of the right upper quadrant. The following findings are suggestive of acute cholecystitis: (1) Enlarged gallbladder over 90 mm in long axis and over 50 mm in short axis. (2) Thickening of the wall greater than 3 mm. (3) Sludge or stones seen within gallbladder. (4) Perivesicular fluid collection. (5) Murphy sign. The sensitivity drops if all the criteria are used but specificity increases.

Cardiogenic or noncardiogenic gallbladder wall edema can confound the diagnosis of acute cholecystitis.

Spleen analysis in the critically ill is seldom revealing. It acts as an acoustic window for different organs and also to localize the diaphragm while planning for thoracentesis. In a trauma patient, splenic laceration can be detected and in rare instances a splenic abscess is seen.

Deep Venous Thrombosis Assessment

Ultrasound Probe Selection and Orientation

The high-frequency 7.5 to 10 MHz linear probe is preferred in our institution with the cursor pointing toward the right of the patient.

Introduction of Venous Ultrasonography of the Lower Extremities

Deep vein thrombosis (DVT) and pulmonary embolism (PE) are a spectrum of the same disease. In a rapidly decompensating patient, either in the hospital ward or in the ICU, DVT and PE should always be in the differential diagnosis. Blood test such as d-dimer is nonspecific in the hospitalized patient. Ultrasound of the lower extremities can successfully diagnose DVT. When done by a trained intensivist, the sensitivity and specificity are very similar to examinations done by the radiology technician and interpreted by a radiologist.²⁴ A DVT can be found in the upper and lower extremities. We will focus on lower extremity anatomy here.

The patient should be positioned in the decubitus position with the leg slightly flexed and externally rotated (Figure 12–17). The leg should be examined by performing gentle and firm compression, enough to collapse (Figure 12–18) the vein but not the artery. A vein with DVT will not collapse. (Figures 12–19 and 12–20).

The operator should be familiar with the lower extremity venous anatomy (see Figure 12–20). The ultrasound probe should be placed at the level of the inguinal ligament to localize the femoral vein, which is medial to the femoral artery. Once this is identified, a compression movement is performed to check for collapsibility, a collapsible vein rules out DVT. A noncollapsible vein, rules in DVT. Color Doppler can be used but will increase study time without adding much information. Unfortunately, the external iliac veins are not visualized fully but can be partially visualized if the probe is tilted and pointing the ultrasound beam “under” the inguinal ligament. Once the common femoral vein (CFV) is studied, the drainage of the greater saphenous vein can be seen medially. A couple of centimeters below, the CFV bifurcates in deep femoral vein and superficial femoral vein (SFV). It is important to recall that the SFV is still part of the deep venous system. Compression should be done every 2 to 3 cm while identifying the perforating branches until the vein enters the adductor canal if visibility allows. The popliteal vein should also be studied with the leg bent and the sole of the foot on the bed (see Figure 12–17). The popliteal vein should be located superficially to the artery. Again, firm pressure is applied to evaluate compressibility.

How can we integrate these findings in our clinical assessment? A hypoxic patient with positive DVT and bilateral A-line pattern in the lung, PE should be suspected. Furthermore, a quick echocardiogram should be done and if there are signs of right ventricle overload, submassive or massive PE is highly suspected.

Optic Nerve Ultrasonography

Probe Selection and Orientation

It is usually performed with the high-frequency 7.5 to 10 MHz linear probe and it is placed over the eyelid. The optic nerve sheath (ONS) will be identified and then measured (Figure 12–21).

Competency Requirements

It is estimated that 10 studies could be sufficient in experienced US operators and 25 studies for inexperienced US operators to achieve competency.²⁵

Clinical Implications

The gold standard for detection of an elevated intracranial pressure (ICP) is invasive measurement. These methods can lead to complications such as infection or bleeding. Other methods are described but we found US measurement of the optic nerve sheath diameter (ONSD) an excellent screening test for elevated ICP.

When ONSD is compared to the gold standard, a pooled analysis using cutoff value range from 5.2 to 5.9 mm showed a sensitivity and specificity of 74% to 95% and 74% to 100%, respectively.²⁶ In another meta-analysis²⁷ the pooled sensitivity and specificity was 90% and 85%, respectively. One study using 5 mm as a cutoff²⁸ yielded 100% and 95% sensitivity and specificity, respectively.

Several procedures can be done in the intensive care unit (Table 12–6). These procedures will be described in details in subsequent chapters. When possible, the dynamic technique, which involves direct visualization of the needle tip, is highly recommended for nearly all procedures. Continuous vision of the needle tip will ensure low complication rates.

POC-US is a helpful noninvasive tool that benefits the critically ill patient. It is easy to perform and learn but requires practice to master. POC-US should be incorporated in the skills of every intensivist.