Chapter 44. Point-of-Care Echocardiography in the Emergency Department

Point-of-care echocardiography is ideally suited to the care of the critical patient in the emergency department or intensive care unit. It is highly accurate, noninvasive, portable, rapidly performed, easily repeatable, and simple to learn. It can provide critical information in real time that otherwise may not be available, particularly in the setting of life-threatening emergencies. Point-of-care echocardiography increases patient safety, improves diagnostic accuracy, reduces diagnostic uncertainty, improves efficiency, and saves lives.

The purpose of this chapter is to provide a general understanding of point-of-care echocardiography, as it is applied and accepted in the emergency department and ICU setting. We present an overview of how echocardiography can be utilized in the management of the critically ill patient. This chapter is not intended as a complete reference and assumes a basic understanding of ultrasound physics, image generation, ultrasound modes, terminology, and system operation.

Point-of-care ultrasound developed originally in Japan and Europe, and came into emergency medicine practice in United States in the 1990s. Point-of-care echocardiography is considered core content of the specialty of emergency medicine by the American Board of Emergency Medicine.1 The American Medical Association supports the use of ultrasound by appropriately trained clinicians in varied specialties, and supports specialty-specific guidelines for training, education, and oversight.2

Point-of-care echocardiography is not the same as a comprehensive echocardiogram or an ultrasound study done in a traditional imaging suite. It is performed, interpreted, and integrated into patient care in real time at the bedside. Our goal is to immediately impact patient care by quickly and accurately performing a brief or focused exam designed to answer simple yes/no questions. Our exam will focus on immediate life-threatening conditions and to assess response to resuscitative measures. Point-of-care echocardiography has evolved over the past two decades into a bedside diagnostic tool, a method to safety guide invasive procedures, and a way to noninvasively assess and monitor ongoing resuscitation. The most current American College of Emergency Physicians (ACEP) Emergency Ultrasound Guidelines provide a comprehensive overview of the scope of practice, and training and credentialing guidelines, and serve as an excellent reference for any department starting and developing a point-of-care ultrasound program.3

Echocardiography is an essential skill for emergency physicians and is especially well suited to the critically ill patient. Using simple qualitative echocardiography, emergency physicians can rapidly and definitively assess for cardiac activity in cardiac arrest, evaluate for pericardial effusion and tamponade, estimate left ventricular systolic function, noninvasively estimate preload and right ventricular filling pressure, identify acute right heart strain, direct resuscitation and medical decision making, and immediately differentiate treatable causes of PEA and shock. This chapter will focus on image acquisition, image interpretation, and integration of this information obtained into the care of the critically ill patient.

Echocardiography is a technically challenging study to perform for many reasons. The heart is surrounded by the bony ribs and sternum, as well as the air-filled lungs that expand and compress with a varying respiratory rate. Both of these impede image acquisition by reflecting and scattering the sound waves, respectively. In addition, differences in body habitus, particularly obesity and deformities of the chest wall, and chronic disease conditions, such as emphysema, may make performing the study more difficult. The left lateral decubitus position is preferred for obtaining high-quality images, but many critical patients cannot be optimally positioned. Despite these inherent challenges, echocardiography rapidly provides critical, high-yield information, making it an invaluable tool.

The heart sits inside the left chest cavity at an oblique angle with the long axis of the heart running along a plane from the right shoulder to the left hip. The great vessels and base are cephalad and the apex is caudad. The right side of the heart is anterior inferior and the left side of the heart is posterior superior. Knowing this basic anatomy will improve image acquisition and help to explain orientation of structures.

The phased array transducer is the transducer of choice for echocardiography (Figure 44-1). Its small footprint and wide field of view allow easy manipulation and imaging between intercostal spaces. Image resolution is worse than transducers with similar frequencies, but temporal resolution (frame rate) is more important in the dynamic imaging of echocardiography.

Standard cardiac orientation orients the image toward the patient's head or left side, with the image indicator appearing in the upper right of the ultrasound screen. Views will be obtained with the transducer indicator oriented to the patient's head or left side. This leftward orientation is the opposite of abdominal sonography, in which the image is oriented toward the patient's head or right side. Most ultrasound systems have a cardiac or echo preset that will automatically orient the image. The techniques and images in this chapter are presented in the traditional leftward or cardiac orientation.

Echocardiography is a dynamic medium, and still images do not convey the same amount of detail and information as live imaging. Most modern ultrasound systems have the capability to store digital video clips, which is the preferred image storage modality.

Many imaging windows and views have been described in echocardiography. We present a series of five views that allow for a rapid and comprehensive evaluation. These include the subxiphoid four-chamber view, subxiphoid longitudinal inferior vena cava (IVC) view, parasternal long-axis view, parasternal short-axis view, and apical four-chamber view. Additional views, such as the apical five-chamber view, will be discussed when evaluating left ventricular systolic function. The authors recommend obtaining as many views as is possible on each patient. Each view has strengths and weaknesses, and additional views may add additional critical information. One view may be sufficient in certain settings, such as cardiac arrest, but generally two to five views allow for a more accurate and comprehensive assessment.

Subxiphoid Four-Chamber View

To obtain the subxiphoid four-chamber view, place the transducer in subxiphoid region, point the transducer indicator toward patient's left side, and aim the ultrasound beam toward patient's left shoulder at a shallow angle (Figure 44-2). Identify the liver, cardiac silhouette, right ventricle, left ventricle, right atrium, left atrium, and pericardial space (Figure 44-3). Adjust the image by tilting, rocking, rotating, or sliding the transducer in order to align the beam angle so that all structures are in view.

This view is the easiest four-chamber view to obtain, and provides a great global overview. It allows for evaluation of pericardial effusion and evaluation of chamber size comparison, and is the best view during cardiopulmonary resuscitation because it does not interfere with resuscitative efforts including chest compressions, central lines, and pacer pads. This is generally the easiest view to obtain in patients with emphysema or other chest deformities.

Common mistakes when performing this view include too steep an angle of the ultrasound beam, as well as not enough depth to visualize the entire heart. A tip to allow a shallow angle of approach is to grasp the transducer from above, which allows a shallower angle without the operator's hand in the way. Stomach gas may obscure this view, which may be improved with either steady transducer pressure or having the patient take and hold a deep breath. Becoming familiar with the various troubleshooting techniques will allow for more efficient imaging.

Subxiphoid Longitudinal IVC View

To obtain the subxiphoid longitudinal IVC view, place the transducer in subxiphoid region, point the transducer indicator toward patient's head, and sweep into the patient's right upper quadrant to find IVC running through liver in a longitudinal plane (Figure 44-4). Identify the liver, IVC running through the liver, junction of the IVC and hepatic veins, junction of the IVC and right atrium, right atrium, right ventricle, and pericardial space (Figure 44-5). Adjust the image by tilting, rocking, rotating, or sliding the transducer, in order to align the beam angle so that all structures are in view. The IVC may also be imaged in cross-section, but the longitudinal view is the authors' preference.

This view allows noninvasive assessment of central venous pressure during normal respiration, using both IVC diameter and respiratory change. M-mode is useful to obtain both maximal and minimal diameters of the IVC during the respiratory cycle (Figure 44-6).

A common mistake when performing the subxiphoid longitudinal IVC view is failing to tilt the transducer to shallow the angle of the ultrasound beam. When measuring the IVC diameter, it is important to have the long axis of the IVC perpendicular to the ultrasound beam. In addition, the IVC and abdominal aorta may be confused. The IVC is more to the patient's right, runs through the liver, has thin walls, generally exhibits respiratory variation, and enters the right atrium. The aorta is more to the patient's left, runs posterior to the liver, has thick echogenic walls, and has the celiac and SMA vessels exiting anteriorly.

Parasternal Long-Axis View

To obtain the parasternal long-axis view, place the transducer perpendicular to chest wall in the left fourth to sixth parasternal space, and point the transducer indicator toward patient's right shoulder (Figure 44-7). Identify the right ventricle, left ventricle, left atrium, mitral valve, aortic valve, aortic root, and descending thoracic aorta posterior to the left atrium (Figure 44-8). Adjust the image by tilting, rocking, rotating, or sliding the transducer, in order to align the beam angle so that all the structures are in view. Rolling the patient into the left lateral decubitus position may allow for better image quality.

This is the best view for measurement of aortic root diameter, which should be less than 3.8 cm in diameter. In order to perform Doppler estimation of cardiac output, the left ventricular outflow tract (LVOT) diameter (LVOT D) should be measured in this view. This is also an excellent view for estimation of left ventricular systolic function using qualitative methods.

Parasternal Short-Axis View

To obtain the parasternal short-axis view, place the transducer perpendicular to the chest wall in the left fourth to sixth parasternal space, and point the transducer indicator toward the patient's left shoulder (Figure 44-9). This view can also be obtained by rotating the transducer 90° clockwise from the parasternal long-axis view. Identify the right ventricle, left ventricle, and papillary muscles indenting the left ventricle (Figure 44-10). The papillary muscles serve as a landmark to make sure the section in view is through the left ventricle, and not the left atrium or aortic root. To obtain this view, the operator may have to tilt the transducer slightly downward, toward the patient's left hip, along the long axis of the heart. Rolling the patient into the left lateral decubitus position may allow for better image quality.

This is an excellent view for estimation of left ventricular systolic function, and the best view for identification of regional left ventricular systolic dysfunction.

Apical Four-Chamber View

To obtain the apical four-chamber view, place the transducer at the point of maximum impulse (PMI), point the transducer indicator toward the patient's left axilla, and aim the ultrasound beam toward patient's right shoulder at a shallow angle (Figure 44-11). Alternatively, the apical window can be found by finding the apex of the heart in the parasternal long-axis view toward the left of the screen. The transducer is moved over the apex in real time, and then rotated toward the patient's left axilla and the beam angle flattened. This is the most challenging basic view to obtain, and rolling the patient into the left lateral decubitus position usually allows for better image quality. Identify the left ventricle, mitral valve, left atrium, right ventricle, tricuspid valve, and right atrium (Figure 44-12).

This is the best view for assessment of valvular pathology, as well as comparing relative right and left ventricular sizes. The normal right ventricle to left ventricle ratio is <0.6 to 1 measured across valve leaflets. A good rule of thumb for qualitative inspection is that the image should be “1/3 right ventricle, 2/3 left ventricle.” This view is also excellent for Doppler interrogation of inflow and outflow velocities for advanced echocardiographic applications.

One prospective study showed that a focused 6-hour training course significantly improved emergency medicine residents' theoretical and practical knowledge of bedside echocardiography.4 This study enrolled 21 emergency medicine residents who underwent 5 hours of didactics and 1 hour of hands-on instruction on echocardiography. Subjects underwent preexposure and postexposure testing of their theoretical and practical knowledge. Practical scores significantly increased from 56% to 94%, and theoretical scores significantly improved from 54% to 76%. This study suggests that skills necessary for competent bedside echocardiography can quickly be learned and applied.

Current ACEP guidelines recommend performing at least 25 proctored exams prior to using bedside echocardiography independently in patient care decisions.3 The exception to this would be in a case where delaying treatment or further interventions in order to obtain some other reference standard would cause undue harm to the patient.

Echocardiography in Cardiac Arrest

Echocardiography is an excellent adjunct in the setting of cardiac arrest. It enables one to differentiate organized and agonal cardiac contraction from cardiac standstill, providing prognostic information. In addition, it allows rapid diagnosis and treatment of reversible causes of cardiac arrest, such as hypovolemia, tamponade, left ventricular dysfunction secondary to myocardial infarction, and acute right heart strain due to pulmonary embolus.

There is a strong association between cardiac standstill on echocardiogram and mortality. One prospective observational trial enrolled 169 patients in cardiac arrest.5 All 136 patients with cardiac standstill on initial bedside echocardiogram died in the emergency department. Another prospective observational study of 70 patients in cardiac arrest showed 100% mortality for patients with cardiac standstill.6 An additional prospective observational study of patients in cardiac arrest enrolled 20 patients, and found 100% mortality for patients in cardiac standstill.7 If replicated in larger studies, these results imply that substantial hospital resources and manpower could be saved by using echocardiography to identify cardiac standstill during resuscitation. At the time of this writing, a multicenter prospective study is underway to confirm these findings with sufficient power.

Cardiac standstill carries a grave prognosis and can be considered a marker for termination of resuscitative efforts in certain settings. Individual reflections from slow-moving red blood cells may be seen, or blood may be seen clotting in the heart, which is a late finding (Figure 44-13).

The presence of cardiac activity during any point of the resuscitative effort is strongly associated with survival to hospital admission. In a prospective observational trial on 102 patients in cardiac arrest, patients with cardiac activity at any point of the resuscitation survived at a much higher rate, 27% versus 3%.8 A similar study showed that 12 out of 18 patients (67%) with PEA and cardiac contractions survived to hospital admission.5 A third study showed that 8 out of 11 patients (73%) with PEA and cardiac contractions survived to hospital admission.6

Any cardiac motion corresponding with electrical impulses should be considered to be cardiac activity, and aggressive attempts at resuscitation should be continued. The presence of cardiac activity predicts return of spontaneous circulation.

Another benefit of bedside ultrasound during cardiac arrest is the ability to identify reversible causes, such as pericardial tamponade. In a prospective observational study of patients in cardiac arrest of 20 patients, the authors demonstrated pericardial effusions in 8 of 12 patients with cardiac motion, including 3 cases of tamponade.7 Finding a treatable cause of cardiac arrest, such as a large pericardial effusion, should prompt immediate definitive treatment. In this case, pericardiocentesis is indicated, preferably under ultrasound guidance. The findings of tamponade physiology are discussed later in the chapter.

As previously mentioned, the subxiphoid four-chamber view is an excellent view during cardiac arrest. It does not interfere with chest compressions or other resuscitative efforts, and provides an excellent global view of the heart. From this window, the heart can be quickly assessed for the presence of cardiac contraction, pericardial effusion and tamponade, left ventricular function, as well as right ventricular size, to detect right ventricular dilation in suspected pulmonary embolus or a flat right ventricle in suspected hypovolemia. Alternatively, the parasternal long axis can be used. Imaging should occur during pulse checks to minimize interruptions in CPR.

Echocardiography to Identify Pericardial Effusion and Tamponade Physiology

Emergency physicians can rapidly and accurately identify pericardial effusion and tamponade physiology using point-of-care echocardiography.

Several studies have shown emergency physicians can accurately identify pericardial effusion and tamponade physiology. One prospective observational enrolled 515 patients at high risk for pericardial effusion and found 103 positive studies.9 All studies were performed and interpreted by emergency physicians and subsequently reviewed by a cardiologist. Sensitivity and specificity of bedside echocardiography performed by emergency physicians for pericardial effusion were 96% and 98%, respectively. Another study showed that patients with unexplained new-onset dyspnea may benefit from emergency physician–performed echocardiography to rule out pericardial effusion.10 This prospective, observational trial enrolled 103 patients with new-onset dyspnea unexplained by pulmonary, infectious, hematologic, traumatic, psychiatric, cardiovascular, or neuromuscular disease after ED evaluation. Fourteen of 103 patients had effusions, with 4 classified as large.

In penetrating chest trauma, echocardiography has been shown to reduce time to diagnosis and decrease mortality. In one retrospective study, the authors reviewed the records of 49 patients with penetrating cardiac injury.11 Survival was 100% in the echo group, compared with 57% in the nonecho group. The average time to diagnosis and disposition for surgical intervention was significantly shorter at 15 minutes for the echo group versus 42 minutes for the nonecho group A prospective multicenter study enrolled 261 patients, with 29 true positives confirmed in the operating room.12 Bedside echocardiography had a sensitivity of 100% and a specificity of 97% for the diagnosis of hemopericardium, and mean time from ED arrival to operative intervention in positive cases was 12 minutes. Due to its excellent test characteristics and time savings, the authors recommended bedside echocardiography as the initial diagnostic modality of choice in penetrating chest trauma.

A pericardial effusion appears as an anechoic fluid collection in the pericardial space. Typically, effusions are circumferential, but may be loculated in postoperative patients and patients with inflammatory conditions. In these cases, there may be some echoes present within the effusion. The absence of pericardial effusion essentially rules out tamponade as a cause of hypotension.

Pericardial effusions are seen in many sizes. A small effusion may only be seen in the dependent portion of the pericardium, and generally measures less than 5 mm. Moderate pericardial effusions are generally circumferential, and measure between 5 and 10 mm. Large pericardial effusions are circumferential and measure greater than 10 mm. Swinging of the heart in the pericardial sac may be seen, and often manifests as electrical alternans on electrocardiography (Figures 44-14, 44-15, 44-16, 44-17, 44-18).

Pitfalls in diagnosing pericardial effusions do exist. The most common are mistaking a normal fat pad or pleural effusion for a pericardial effusion. These pitfalls can be avoided with careful scanning, observing the heart in multiple views, and identifying key landmarks.

A pericardial fat pad can often have a similar appearance to an effusion (Figure 44-19). Typically, a fat pad is only present anteriorly, has internal echoes, and moves with the heart. Using multiple views will allow the user to distinguish an anatomic fat pad from an abnormal pericardial effusion.

A left-sided pleural effusion is commonly mistaken for a pericardial effusion. Pericardial effusions are typically circumferential, where a pleural effusion is only seen posterior to the heart. Key landmarks for identifying pericardial from pleural fluid are the left atrium and descending thoracic aorta on the parasternal long-axis view. Pericardial fluid will appear between the left atrium and descending aorta, whereas pleural effusion will be seen posterior to the descending aorta (Figure 44-20).

When a pericardial effusion is identified, tamponade must be considered. Tamponade is a time-critical clinical diagnosis of hypoperfusion in the setting of a pericardial effusion. Increasing pressure in the nondistensible pericardium limits right ventricular filling and venous return, leading to circulatory collapse. Tamponade is difficult to diagnose solely on clinical findings. A recent case series showed that pericardial tamponade can often present without the classic findings of Beck's triad, and mimic more common disease processes.13

Qualitative echocardiography can show evidence of “tamponade physiology.” Findings consistent with tamponade physiology include early diastolic collapse of the right ventricle, late diastolic collapse of the right atrium, as well as dilation of the IVC with loss of normal respiratory variation (Figures 44-21, 44-22, 44-23, 44-24).

Doppler echocardiography can also be used to diagnose tamponade in the setting of pericardial effusion. Normal cardiac filling is influenced by respiratory phase, with filling reduced in inspiration due to negative intrathoracic pressure. This phenomenon leads to the presence of pulsus paradoxus. In tamponade, cardiac filling is further impaired, and this can be demonstrated using pulsed wave spectral Doppler interrogation of mitral inflow velocities. This is best seen in the apical four-chamber view. To obtain this information, place the pulsed wave spectral Doppler gate in the LV to measure mitral inflow velocity. Normal E-wave (early diastolic) velocity decreases with inspiration, but usually less than 10–15%. With tamponade, LV inflow is further restricted, leading to an exaggerated inspiratory decrease of greater than 25% in maximal E-wave velocity (Figure 44-25).14

One prospective observational study in 56 consecutive patients with pericardial effusion study compared qualitative findings of tamponade and Doppler findings of tamponade.15 Sixteen patients were found to have tamponade and underwent drainage. A 22% decrease in peak mitral inflow velocity during inspiration had sensitivities and specificities of 77% and 80%, respectively. Right ventricular collapse performed quite similarly with sensitivities and specificities of 75% and 85%, respectively.

Echocardiography to Estimate Left Ventricular Systolic Function

Emergency physicians can accurately estimate left ventricular systolic function using point-of-care echocardiography. Both qualitative and quantitative estimation methods will be reviewed. Simple qualitative methods are fast, easy to learn, and accurately correlate with quantitative methods.

Several studies have shown that emergency physicians can accurately estimate left ventricular systolic function. A prospective observational study enrolled a convenience sample of 51 patients with symptomatic hypotension.16 Patients underwent bedside echocardiography by emergency physicians and were classified to have normal, depressed, or severely depressed ejection fraction. A blinded cardiologist interpreted these images and served as the gold standard, while a second cardiologist reviewed the studies to determine interobserver reliability between cardiologists. Pearson's correlation coefficient between emergency physicians and the cardiologist was 0.86, compared with 0.84 between the cardiologists. Classification of ejection fraction between EP and cardiology showed a weighted κ of 0.61, representing substantial agreement. Another prospective observational study enrolled 115 patients in whom emergency physicians performed bedside echocardiography and classified ejection fraction as poor, moderate, and normal.17 The cardiology department performed and interpreted a comprehensive echocardiogram, which served as the gold standard. Results showed a Pearson's correlation coefficient of 0.71, with an overall agreement of 86%. The highest agreement was found in the normal category (92.4%), followed by the poor category (70.4%). While emergency physicians perform well overall when classifying ejection, they perform best when ejection fraction is clearly normal or poor, which represents the most clinically relevant determination. Another study showed that intensivists could classify left ventricular function appropriately in the critically ill patient.18 This prospective observational study enrolled 44 patients, who underwent echocardiography by intensivists, followed by the cardiology department, which was used as the gold standard. Intensivists classified left ventricular function as grossly normal or abnormal with excellent agreement and κ of 0.72. Intensivists also correctly placed ejection fraction into one of the three categories (normal, mild to moderately depressed, severely depressed) in 36 of 44 patients for a κ of 0.68.

One study showed that emergency physicians preferred the parasternal long-axis view over all other cardiac views studied, except the parasternal short-axis view.19 This prospective observational study enrolled 70 patients in a surgical intensive care unit on whom the following views were obtained: parasternal long axis, parasternal short axis, subxiphoid four chamber, subxiphoid short axis, and apical four chamber. Sonographers rated their preference for each window obtained on a 5-point Likert scale. Parasternal long-axis view was preferred over all other views (P < .05), except the parasternal short-axis view (P = .23). Since time needed for a study to be completed is an important factor in the evaluation of the critically ill, this should be considered when choosing the initial view for determining left ventricular function in this patient population.

Qualitative Estimation of Left Ventricular Systolic Function

Qualitative estimation of left ventricular systolic function can be performed by analysis of left ventricular end-diastolic diameter, change in left ventricular diameter during systole, change in thickness of left ventricular walls during systole, and rate and force of valve motion. Each of these criteria will be covered in further detail.

Normal left ventricular end-diastolic diameter is generally less than 5 cm. An LV end-diastolic diameter greater than 6 cm is consistent with dilated cardiomyopathy. Furthermore, it indicates high left ventricular diastolic pressure and diminished left ventricular squeeze (Figure 44-26).

Left ventricular diameter should change by about 40% from end diastole to end systole (Figure 44-27). This may be more accurately seen using M-mode (Figure 44-28). The left ventricular walls thicken during systole by about 40%. This can also be more accurately seen and measured using M-mode (Figures 44-28, 44-29, 44-30, 44-31).

The rate and force of valvular motion can also be used to estimate LV systolic function. The parasternal long-axis view, having both the mitral and aortic valves in view, is ideal for assessing rate and force of valvular opening. In early diastole, LV pressure is low, and the mitral valve should open wide and fast, with the anterior leaflet almost touching the intraventricular septum (Figure 44-32). The distance between the anterior MV leaflet and septum at their closest point is known as the E-point septal separation (EPSS). When LV pressures remain high with poor LV function, the mitral valve opens more slowly and not as wide (Figure 44-33). The mitral valve can be assessed with M-mode, and this more accurately depicts mitral valve opening (Figures 44-34 and 44-35).

Quantitative Estimation of Left Ventricular Systolic Function

Several methods exist to quantitatively measure left ventricular systolic function. These methods can be time consuming. Furthermore, qualitative estimation by an experienced sonographer has been shown to be as accurate as a measured quantitative estimation.

An estimated ejection fraction can be calculated using LV measurements obtained when measuring fractional shortening. Fractional shortening is determined by the following formula: (LV end-diastolic diameter − LV end-systolic diameter)/LV end-diastolic diameter. The normal range for fractional shortening is 30–45%. M-mode using either the parasternal long- or short-axis views can be used to accurately measure fractional shortening. Most ultrasound systems can calculate fractional shortening using built-in calculator packages from left ventricular measurements. Fractional shortening can be used to calculate a measured ejection fraction (Figure 44-36). Ejection fraction is calculated from fractional shortening from the following formula: EF = ([LV end-diastolic diameter]3 − [LV systolic diameter]3)/(LV end-diastolic diameter)3. This is best performed in the parasternal long- or short-axis views. This technique is easy to learn and perform, but has several drawbacks. Measurements must be completely perpendicular to the ventricle, accurate, and avoid overestimating or underestimating the measurements, as any measurement error is compounded as measurements are cubed in the calculation. In addition, this measurement assumes a symmetrically contracting ventricle that is uniform in shape. Any area of hypokinesis will lead to overestimation of ejection fraction.

Left ventricular systolic function can also be estimated using Simpson's method of discs. To perform this method, obtain an apical four-chamber view. End diastole should be identified and the image frozen. The LV volume can be traced using a caliper along the endocardial border. LV volume in diastole will be calculated by forming several small discs in the LV cavity (Figure 44-37). Once the LV end-diastolic volume is calculated, obtain a view of the LV at end systole. Again, trace the endocardial border to obtain LV end-systolic volume. Once the end-diastolic and end-systolic volumes are calculated, ejection fraction will be calculated by the following formula: EF = (end-diastolic LV volume − end-systolic LV volume)/end-diastolic LV volume. For increased accuracy, the procedure would be repeated in the apical two-chamber view, obtained by rotating the transducer 90° counterclockwise from the apical four-chamber view. This method, although accurate, has a few drawbacks. It can be time consuming, the endocardial border can be difficult to clearly visualize, and errors can be made in identifying which frames represent end diastole and end systole.

Doppler echocardiography can also be used to quantitatively estimate stroke volume and cardiac output using the velocity–time integral (VTI) of left ventricular outflow, along with the LVOT D and heart rate. This method compares favorably with traditional pulmonary artery catheter thermodilution methods. First, obtain a parasternal long-axis view and identify the aortic root and aortic valve leaflets. Measure the LVOT D when the aortic valve is open where the valve leaflets attach. The US system will use this measurement to calculate the LVOT cross-sectional area (Figure 44-38). Next, obtain an apical five-chamber view. This view is very similar to the apical four-chamber view, with the probe in the same position and orientation. The transducer is angled slightly anteriorly toward the chest wall to bring the LVOT into view. Using spectral pulsed wave Doppler, obtain a Doppler tracing of LVOT outflow (Figure 44-39). The LVOT outflow waveform will be the tracing under the baseline. It is very important to have the Doppler gate as perpendicular to the LVOT as possible to avoid measurement error. Using the calculator package, trace the LVOT outflow curve. The area under this curve will be the VTI, and most systems will calculate this value automatically (Figure 44-40). Stroke volume is calculated as LVOT VTI times LVOT cross-sectional area, and cardiac output can be calculated by multiplying stroke volume and heart rate. Some US systems allow continuous tracing and instantaneous calculation of VTI, stroke volume, and cardiac output in real time. Drawbacks to this technique include needing to obtain an apical view, as well as not having the proper perpendicular angle when obtaining the Doppler tracing. In addition, the aortic root and LVOT may be difficult to identify in some patients.

Echocardiography to Estimate Central Venous and Rv Filling Pressures

Emergency physicians can accurately estimate central venous and right ventricular filling pressure using point-of-care echocardiography. Key methods for estimation of volume status include assessment of IVC diameter and IVC collapsibility index. These measures can be helpful in differentiating hypovolemia, patients in septic shock who will respond to fluid resuscitation, fluid overload states such as CHF, tamponade physiology, and elevated RV pressures in suspected pulmonary embolus.

The IVC is a capacitance vessel, and its volume and pressure dynamics are related to CVP. The standard location to measure the IVC diameter is just distal to the IVC and hepatic vein junction, as the IVC is fixed at the diaphragm, which limits evaluation for respiratory variation. The normal diameter of the IVC in adults is between 1.5 and 2.5 cm. Patients with low volume will tend to have IVC diameters less than 1.5 cm, while patients with volume overload will tend to have IVC diameters greater than 2.5 cm. With inspiration, thoracic pressure becomes negative, leading to increased venous return and a smaller IVC diameter. This can be accentuated and more accurately measured using M-mode (Figures 44-41 and 44-42). The ratio of maximal IVC diameter to minimal IVC diameter during inspiration is the IVC collapsibility index, also known as caval index. Normal variation with inspiration is about a 50% reduction in IVC diameter. CVP is estimated using both the IVC diameter and percent change during respiration. This estimation correlates with CVP best at the extremes, which is the most clinically relevant scenario. Using this technique, emergency physicians can rapidly and accurately diagnose the low and high CVP noninvasively.

A recent prospective observational study enrolled 102 patients undergoing right heart catheterization and echocardiography.20 Initially receiver operating characteristics were analyzed to determine optimal cutoffs, which were then prospectively studied. An IVC diameter of 2 cm correctly predicted RAP above or below 10 with a sensitivity and specificity of 73% and 85%, respectively, and a collapsibility of 40% performed similar with sensitivities and specificities of 73% and 84%, respectively. A recent study showed that greater than 50% collapse of the IVC with inspiration was sensitive and specific for a central venous pressure measurement less than 8 mm Hg.21 This prospective observational study enrolled 73 patients who were undergoing central venous catheterization. Ultrasound measurements of the IVC during inspiration and expiration were performed and caval index was calculated. They found greater than 50% collapse was associated with central venous pressure less than 8 mm Hg with a sensitivity of 91%, a specificity of 94%, positive predictive value of 87%, and negative predictive value of 96%.

Several studies have shown that a flat IVC and elevated collapsibility index is an accurate and sensitive marker of hypovolemia. One study showed that IVC diameter correlated with hypovolemia in trauma patients.22 This prospective observational study enrolled 35 victims of trauma, 10 patients found to be in shock, defined as an SBP <90 mm Hg on arrival or within 12 hours of arrival, and a control group of 25 hemodynamically stable patients. The mean diameter of the IVC was much smaller at 7.7 mm in the shock group, compared with 13.4 mm in the control group. Subjects were also divided into two groups based on IVC diameter. Those with an IVC diameter of 9 mm or less had significantly larger blood transfusion volume, 11.3 U versus 0.3 U. Another study by the same authors showed that in patients being resuscitated with hemorrhagic shock, IVC diameter could predict recurrence of shock.23 This prospective observational study enrolled 30 patients with hemorrhagic shock, who after enrollment were fluid resuscitated until SBP was above 90 mm Hg. All patients then had IVC diameter recorded using bedside sonography. Patients were subsequently divided into two groups: those who remained stable after initial resuscitation (13 patients) and those who experienced recurrence of hypotension (17 patients). These two groups had no significant differences in vital signs after fluid resuscitation. However, those who experienced recurrence of shock had significantly smaller IVC diameters, 6.5 ± 0.5 mm versus 10.7 ± 0.7 mm (P < .05). Although larger studies are needed to confirm these results, this implies that IVC diameter may be a useful adjunct in predicting the clinical course of trauma patients. Another study showed that IVC diameter was consistently decreased even by a small amount of blood loss.24 This prospective observational study enrolled 31 healthy volunteers at a blood donation center. IVC diameter was measured before and after donation of 450 mL of blood. Mean IVC diameter before donation was 17.4 mm (95% CI 15.2–19.7 mm) and decreased to 11.9 mm after donation (95% CI 10.3–13.6 mm). A flat, collapsable IVC should prompt aggressive fluid resuscitation (Figures 44-43 and 44-44).

Additional studies have shown that a dilated IVC and low collapsibility index is an accurate marker of volume overload. One prospective observational study enrolled 75 patients hospitalized for acute decompensated CHF.25 The authors found that predischarge IVC diameter, collapsibility index, and BNP were predictive of the need for readmission. Another study showed that the IVC caval index can identify patients with right heart failure.26 This cohort study enrolled 95 patients without right heart failure and 32 patients with documented right heart failure. The caval index was measured in both groups and results of receiver operator characteristics were analyzed. A cutoff caval index value of 0.22 yielded a sensitivity of 78% and specificity of 98% for the presence of right heart failure. Another study showed the IVC caval index to be useful in the emergency department to diagnose congestive heart failure.27 This prospective observational study enrolled 46 patients presenting to the emergency department with dyspnea. IVC caval index was determined prior to initiation of therapy, and patients with a final diagnosis of CHF were compared with those who had an alternative final diagnosis. Respiratory variation in patients with CHF was smaller than in those without CHF—9.6% versus 46%. Receiver operating characteristic curve analysis with a cutoff of 15% yielded a sensitivity of 92% with 84% specificity (Figures 44-45 and 44-46).

Two studies evaluated the ability of IVC diameter and caval index to predict fluid responsiveness in patients in septic shock. One prospective observational study enrolled 23 patients with sepsis and respiratory failure on mechanical ventilation.28 IVC diameter at end expiration and end inspiration was measured and the distensibility index was calculated ([Dmax − Dmin]/Dmin). Cardiac index was calculated using Doppler flow before and after patients received a fluid challenge. IVC distensibility above 18% identified fluid responders, those with an increase in cardiac index of at least 15%, with a sensitivity and specificity of 90%. Similar results were found in another study.29 This prospective observational trial enrolled 39 mechanically ventilated patients in septic shock. IVC distensibility was calculated as the difference between expiratory and inspiratory diameters divided by their mean. Cardiac index was measured before and after a fluid challenge. IVC distensibility above 12% distinguished between responders (those with an increase in cardiac index of at least 15% following fluid challenge) and nonresponders with positive and negative predictive values of 93% and 92%, respectively.

Evaluation of Acute Right Heart Strain

Emergency physicians can identify acute right heart strain using point-of-care echocardiography. This can be very clinically helpful in the setting of massive or submassive pulmonary embolism to direct the proper course of treatment.

A recent study showed that emergency echocardiogram could support the diagnosis of pulmonary embolism.30 The authors conducted a prospective observational trial enrolling 124 patients with a suspected diagnosis of pulmonary embolism who then underwent emergent echocardiography. A study was considered positive if it displayed any two of the following signs: right ventricular dilatation, abnormal septal motion, right ventricular hypokinesis, elevated pulmonary artery or right ventricular pressures, moderate to severe tricuspid regurgitation, or visualization of a clot within the right ventricle or pulmonary artery. CT, MRI, and VQ scan identified 27 cases of pulmonary embolism. Echocardiography performed with a sensitivity and specificity of 41% and 91%, respectively. This study shows that a positive study in a high-risk patient should prompt treatment, but a negative study should not be used to rule out pulmonary embolism.

There are several helpful simple findings to identify acute right ventricular strain on bedside echocardiography. These include a dilated RV, septal shift of the intraventricular septum during diastole, RV hypokinesis, and intracardiac thrombus. Each of these findings will be discussed in further detail.

The normal RV to LV ratio is less than 0.6 to 1. This is best seen and measured in the apical four-chamber view, and the standard location for measurements is at the tricuspid and mitral valve leaflets (Figure 44-47). Additional views may also show RV enlargement. The subxiphoid four-chamber view may also be used, although the transducer angle may overestimate or underestimate RV diameter. In the parasternal long-axis view, the RV diameter should be less than 2.5–3 cm (Figure 44-48). A good rule of thumb for simple qualitative visual inspection is that the image should be “1/3 RV, 2/3 LV.” As pressure builds, the RV may become larger than the LV.

When RV filling pressure exceeds LV filling pressure, the septum will paradoxically bow into the LV cavity during diastole. This finding is known as septal shift, and can be best visualized in the subxiphoid four-chamber or apical four-chamber views. Assess the septum for shift during diastole by watching for paradoxical motion while the tricuspid and mitral valves are open (Figure 44-49). In the parasternal short-axis view, septal shift causes a flattening and bowing of the intraventricular septum. This causes a finding known as the “D sign,” as the LV cavity takes on a “D”-shaped appearance, instead of the “O”-shaped appearance normally seen (Figure 44-50).

Another finding of acute RV strain is RV hypokinesis, especially of the midventricular walls. The RV typically pumps against lower pressure, and acute pressure overload causes pump failure. This finding is best seen in the subxiphoid four-chamber or apical four-chamber views. This finding is also known as McConnell's sign. In addition, the RV may lose its typical triangular or wedge-shaped appearance, and take on more of an ovoid shape.

Occasionally, mobile intracardiac thrombus is directly visualized on echocardiography. This represents a pulmonary embolus in transit, and should prompt aggressive treatment (Figures 44-51 and 44-52).

Chronic cor pulmonale can be differentiated from acute cor pulmonale by assessment of RV wall thickness and RV contractility. Over time, the RV can adapt to elevated pressures by hypertrophy of the ventricular walls. Normal RV wall thickness is less than 0.5 cm, and measurements greater than this suggest RV hypertrophy. In addition, the chronically overloaded RV regains the ability to contract forcefully, and RV hypokinesis is not seen in chronic cor pulmonale.

Differentiation of PEA and Shock States

Using the techniques illustrated in this chapter, emergency physicians can use echocardiography to differentiate shock states. This was best illustrated in a prospective study that showed using bedside ultrasound improved diagnostic accuracy in medical patients with undifferentiated hypotension.31 The authors conducted a randomized controlled trial, enrolling 184 nontrauma patients with hypotension and at least one clinical sign of shock. Patients were randomized to receive an immediate ultrasound exam versus an ultrasound exam delayed by 15–30 minutes. The ultrasound exam consisted of a five-view echocardiogram (parasternal long- and short-axis, apical four-chamber, subxiphoid four-chamber, and IVC views), as well as a view of the hepatorenal recess to evaluate for free intraperitoneal fluid and a transverse view of the abdominal aorta to evaluate for the presence of an AAA. This exam took less than 5 minutes to complete. At 15 minutes, in the immediate ultrasound group physicians entertained a smaller number of viable diagnoses (median 4 vs. 9, P < .0001), and ranked the correct final diagnosis as the most like one more frequently (80% vs. 50%, difference 30%). This study is a great example of how to use echocardiography in the critically ill patient and shows how using echocardiography can rapidly cone the differential diagnosis and allow greater certainty when faced with an undifferentiated hypotensive patient.

With relatively brief training, emergency physicians can learn to perform and interpret echocardiography. Using this technology, emergency physicians can rapidly and definitively assess for cardiac activity in cardiac arrest, immediately differentiate treatable causes of PEA and shock, evaluate for pericardial effusion and tamponade, estimate left ventricular systolic function, identify acute right heart strain, noninvasively estimate preload and right ventricular filling pressure, and direct resuscitation and medical decision making with improved diagnostic accuracy.