Chapter 23. TEE in the Emergency Department

Echocardiography has been a component of emergency medicine practice for over 20 years, and serves as an integral diagnostic tool in the evaluation of patients with cardiac and noncardiac disorders. Echocardiography represents the only diagnostic modality capable of providing real-time bedside information for acutely ill patients in the emergency department (ED). In addition to cardiac anesthesiologists and cardiologists, trainees in U.S.-based emergency medicine residency and fellowship programs now often acquire specialized training in echocardiography.

Transthoracic echocardiography (TTE) has proven to have several applications in routine clinical practice including the detection of cardiac effusion and tamponade, estimating cardiac ejection fraction, and as an important adjunct during specialized procedures such as pericardiocentesis.1,2 Under certain circumstances, however, TTE may be impractical (ie, during active chest compressions in the arresting patient), impossible (ie, in the morbidly obese patient), or inadequate (ie, definitive imaging of the ascending aorta). In such cases, transesophageal echocardiography (TEE) may be a suitable alternative, and in some cases can be regarded as the first-line echocardiographic investigation. While TEE is still primarily under the purview of the cardiologist or cardiac anesthesiologist, suitably trained emergency medicine physicians have also begun to incorporate TEE into their practice. Regardless, physicians using TEE as a tool for clinical management need to be adequately trained in performance and interpretation, which has prompted the American Society of Echocardiography to publish a policy statement on echocardiography in the ED.3

• Guidelines for the clinical application of echocardiography in the ED. The ACC/AHA/ASE guidelines for the clinical application of echocardiography (2003) have established recommendations for the use of echocardiography in the critically ill or injured patient (Table 23–1), including that echocardiography is appropriate to use in patients with suspected aortic injury, hemodynamically unstable patients, patients with serious blunt or penetrating chest trauma, and suspected pre-existing valvular or myocardial disease in the trauma patient. Table 23–2 details the specific clinical conditions and diagnoses for which echocardiography can be a useful diagnostic tool in the ED.
• Transthoracic versus transesophageal echocardiography in the ED. The ACC/AHA/ASE guidelines for the clinical application of echocardiography (2003) have also delineated the conditions and settings in which TEE (as opposed to TTE) provides the most definitive diagnosis in critically ill or injured patients, including the hemodynamically unstable patient with suboptimal TTE images or those on ventilators; major trauma or postoperative patients; suspected aortic dissection or other aortic injury; and other conditions in which TEE is superior (ie, endocarditis and cardiac source of emboli).4

TEE has several general advantages that make it a reasonable and useful imaging modality in the ED. TEE is relatively noninvasive and, once placed within the esophagus posterior to the heart, is capable of transmitting real-time images even during active cardiopulmonary resuscitation (CPR), thus permitting cardiac functional information to be viewed by all providers within range of the display monitor. It has been demonstrated to be safe, with low complication rates when used by experienced operators,5 and may result in a management change in up to 80% of cases.6 However, one must exercise added caution when using TEE in the ED given the acuity of the patient population, and only adequately trained providers should perform and interpret studies. For example, one study examining the complication rate in 142 patients undergoing TEE in the ED found a 12.6% complication rate including respiratory issues (n = 7), hypotension (n = 3), emesis (n = 4), agitation (n = 2), cardiac dysrhythmia (n = 1), and death (n = 1),7 a higher complication rate than reported in other clinical situations.8 Other disadvantages of the use of TEE in the ED relate to problems with incomplete (or absent) historical details in ED patients about the last oral intake, which can result in a need for gastric aspiration and decompression; the unexpected encountering of contraindications to TEE probe placement (such as esophageal webs, strictures, or varices); or unstable/unconscious patients incapable of providing consent.

While most studies of echocardiography in the ED have employed TTE as the first-line study, in many clinical situations including obesity, mechanical ventilation, lung disease, and poor echocardiographic windows, TTE limits clinical evaluation. For example, in a study by Varriale and Maldonado,9 they employed TTE as first-line echocardiographic imaging while studying patients with cardiac arrest, but required subsequent TEE to obtain acceptable images in 20% of their patients. In other studies, the failure rate of the transthoracic approach in the ICU setting has been reported to be 10% to 40%.10–12 The critically ill patient in the ED is often managed by intubation and mechanical ventilation, and up to one-half of such patients cannot be adequately imaged by TTE.13 TTE can also be impeded by objects on the chest wall such as lines, catheters, and electrocardiographic (ECG) leads, and even if performed by an experienced technician, TTE requires frequent interruption of CPR for repeated, intermittent acquisition of images during resuscitation. This is particularly relevant given the emphasis on minimizing the interruption of chest compressions in ACLS guidelines. Recently, an algorithm has been proposed for the use of TTE in resuscitation, to be executed simultaneously during CPR cycles to reduce interruptions.14

With TEE, on the other hand, once the probe is placed, it serves as a continuous monitor as it can remain in the esophagus even during active chest compressions15 and is not affected by objects on the chest wall. Continuous monitoring permits images to be viewed by the entire resuscitation team if a display-ready monitor is available, and can allow confirmation of adequate chest compressions by observing appropriate changes in chamber size as the heart alternately fills and empties.

There are several specific situations in the ED for which TEE is an important diagnostic tool for clinical care and may be considered the first-line echocardiographic investigation, including assessment of patients presenting with cardiac arrest to identify reversible causes; evaluation of patients with chest pain to rule out acute aortic dissection; evaluation of the patient with chest trauma; and evaluation of the patient with unexplained hypotension. Examples of such situations are reviewed in a case-based format below.

A 52-year-old male with a history of coronary artery disease, hypertension, and Type 2 diabetes mellitus arrives via local emergency medical service transport. Approximately 20 minutes prior, the patient had reportedly become short of breath with associated chest pressure and nausea. Thereafter, he became unresponsive. Upon arrival at the home, EMS personnel found him prone on the floor, pulseless and apneic. Standard advanced cardiac life support (ACLS) protocol was initiated; organized electrical activity was present on the monitor, and hence the patient appeared to be in pulseless electrical activity (PEA) arrest. After transport to the ED, CPR was continued. Transthoracic echocardiography was attempted but the cardiac windows were inadequate due to the large body habitus of the patient and the resuscitation paraphernalia.

After several rounds of ACLS, discussion to end the resuscitation was initiated. Simultaneously, a TEE probe was inserted and a midesophageal four-chamber view revealed coordinated myocardial contraction but with severe left ventricular dysfunction and an estimated ejection fraction of 10% to 15% (Figure 23–1). Inotropic support with dobutamine and dopamine was initiated. Two days later, the patient recovered neurologically intact.

Despite resources allocated to such initiatives as 911 emergency services, broad distribution of automated electronic defibrillators (AEDs), national campaigns to promote bystander CPR, and standardized ACLS guidelines and training, the most recent data suggest that less than 5% of patients suffering an out-of-hospital cardiac arrest survive to discharge.16 The presence of comorbidities contributes to worse outcomes after presenting with cardiac arrest.17,18 In an effort to improve this dismal survival rate, several studies have detailed theoretical and actual benefits of ultrasound in the setting of cardiac arrest, and some have even lobbied for echocardiography to be incorporated into resuscitation protocols for PEA and asystole.19 Although currently there are no guidelines or recommendations for the use of echocardiography in cardiac arrest, it is thought that a protocol inclusive of echocardiography could lead to a decrease in time between onset of arrest and administration of appropriate therapies.

Role of TEE in the Arresting Patient in the ED

The definitive role of TEE during resuscitation of the arresting patient in the ED and its impact on mortality and morbidity remains to be established. However, a limited number of studies that may be generalizable to this patient population have been suggestive. For example, one small study of in-hospital cardiopulmonary resuscitation using echocardiography found that asystole was initially observed in 90% of patients during CPR; the return of ventricular contractions in four patients prompted positive inotropic therapy. Ventricular wall motion was detected in two patients with bradyarrhythmia (pseudo-electromechanical disassociation) and the causes of cardiac arrest were identified as massive pulmonary embolism and hypovolemia, respectively.9 In another study specifically evaluating TEE, van der Wouw15 performed TEE on 48 patients who arrested in either the in- or out-of-hospital setting. Patients underwent resuscitation and TEE anywhere in the hospital (including the ED but excluding the intensive care unit [ICU]). The underlying pathologic process was elucidated in 31 of 48 (64%) patients with 27 of 31 having the diagnosis confirmed by other studies (angiography or postmortem). Based on these results, TEE had a sensitivity of 93% and specificity of 50% with a positive predictive value (PPV) of 87%. Most importantly, in 31% of all cases (15 of 41), the treatment was changed after the TEE established a diagnosis.15 Memtsoudis et al20 reported a case series of TEE performed on 22 patients who had an unexpected arrest during noncardiac operative procedures. A suspected primary diagnosis was established with TEE on 19 of 22 patients, including acute myocardial infarction, thromboembolism, pericardial tamponade, and hypovolemia. The authors state that TEE aided further management in 18 patients.

Although these results suggest the utility of TEE as an adjunctive diagnostic tool for in- or out-of-hospital arrest, they are confounded by the lack of a “gold standard” for comparison and lack of a control comparison group. For example, intracardiac thrombi identified may not represent causal factors in the cardiac arrest, but may just be a side effect of low flow states after cardiac arrest. Further, the diagnostic utility of TEE in myocardial infarction (MI) is limited in differentiation of acute wall motion abnormalities versus old MI, and is contingent upon an organized rhythm, which may be absent during cardiac arrest. Comparative effectiveness analyses are necessary comparing TEE to other diagnostic modalities during cardiac arrest before definitive statements about routine use of echocardiography in cardiac arrest management can be made.

TEE in Patients Presenting with Pulseless Electrical Activity

Often, cardiac arrest is due to pulseless electrical activity (PEA), also known as electromechanical dissociation (EMD). Upon echocardiographic evaluation, many of these cases are actually found to have some degree of cardiac activity, hence representing “pseudo-EMD.” Establishment of this condition has important diagnostic and prognostic implications, as patients who do have residual cardiac function (ie, those with severe left ventricular dysfunction like our case example) have a better prognosis than those patients with true EMD.21 In one study of echocardiography in 169 cardiac arrest victims, cardiac standstill was visualized in 139 patients; of these, none with echocardiographically identified cardiac standstill survived to leave the ED regardless of the initial electrical rhythm.21 Echocardiography can also be used to confirm asystole and ventricular fibrillation in those patients in whom the rhythm is unclear from the cardiac monitor.

In addition to helping to differentiate the underlying rhythm, echocardiography in cardiac arrest due to PEA or asystole can play an integral and comprehensive role in diagnosing the underlying primary cause (and hence the appropriate treatment), including cardiac effusion with tamponade, pulmonary embolus,22,23 severe hypovolemia,24–27 myocardial infarction,28 myocardial rupture, cardiogenic shock, mitral valve failure or papillary muscle rupture,29,30 and even tension pneumothorax. Unlike ventricular fibrillation and pulseless ventricular tachycardia, identification of the underlying cause is a key focus of treatment for PEA, and hence echocardiography can play a significant role in patient management.

TEE-Aided Diagnosis of the Underlying Cause in PEA Arrest

Pericardial Effusion

Pericardial effusion is usually easily detected by several routine TEE views, including the transgastric short-axis view and midesophageal four-chamber view (Figure 23–2). One must take care, however, to distinguish pericardial from pleural effusions. In the standard TTE parasternal long-axis view, for example, pericardial effusions normally lie anterior to the aorta, whereas pleural effusions lie posterior to the aorta (Figure 23–3). While visualization of an effusion is relatively straightforward, the diagnosis of pericardial tamponade is more difficult given that traditional echocardiographic evaluations such as mitral inflow respiratory flow variation, absent inferior vena caval inspiratory collapse, and presence of right ventricular diastolic collapse may be very difficult if not impossible in the patient with cardiac arrest. Hence one is reliant upon the physical examination and history to establish a diagnosis of pericardial tamponade. Visualization of a “swinging heart” may also help in diagnosing pericardial tamponade. Pericardial tamponade can be caused by trauma, aortic dissection, infection, neoplasm, congestive heart failure, uremia, autoimmune diseases, and radiation therapy.

Hypovolemia

Hypovolemia can be diagnosed on TTE and TEE by the presence of a small, flattened left ventricle in the four-chamber view. Further interrogation of the left ventricular outflow tract with Doppler echocardiography can help establish whether dynamic outflow obstruction related to the hypovolemia is present. In one study, left ventricular end-diastolic volume was found to correlate well with the presence of blood loss.31

Pulmonary Embolism

Pulmonary embolism is typically manifest as right ventricular hypokinesis and/or enlargement, and has been shown to be the cause of almost 5% of cardiac arrests, with PEA being the initial rhythm in 63%.32 Further details of the echocardiographic evaluation of pulmonary embolism are provided in Case 4.

Primary Cardiac Pathology

Echocardiography can be a very useful adjunct diagnostic tool for primary cardiac pathologies leading to cardiac arrest, including cardiogenic shock, myocardial infarction28 (diagnosed by findings of wall motion abnormalities in the proper clinical context), and complications of myocardial infarction including myocardial rupture, ventricular septal defect (VSD), mitral valve failure, and/or papillary muscle rupture.29,30 In an international registry from 19 medical centers of 251 patients after myocardial infarction, the cause of cardiogenic shock was severe left ventricular failure in 85%, mechanical complications in 5%, right ventricular infarct in 2%, and other comorbid conditions in 5%.33 Cardiac free wall rupture is usually a fatal complication of acute myocardial infarction. Echocardiographic diagnosis requires a careful search for the site of rupture and should be suspected if a region of thin myocardium or a small amount of pericardial effusion is present, particularly if a loculated effusion or clot is detected. Detection of a free wall rupture in such patients can lead to surgical repair with a subsequent survival rate of greater than 50%.34 In some cases, a pseudoaneurysm that forms after a free wall rupture is contained within a limited portion of the pericardial space. This occurs most frequently in the inferolateral wall, and is characterized by a small neck communication between the left ventricle and the aneurysmal cavity, with to-and-fro blood flow through the rupture site seen on Doppler and color-flow imaging.34 Mitral regurgitation often occurs in the setting of myocardial infarction, can be severe leading to hemodynamic compromise, and can be due to left ventricular dilatation leading to mitral annular dilatation, papillary muscle dysfunction, or papillary muscle rupture (Figure 23–4). Differentiating these causes of mitral regurgitation is important, as papillary muscle rupture is a serious complication that mandates urgent surgical repair.34

TEE during Cardiopulmonary Resuscitation

In addition to diagnostic utility, TEE can play a role in directing resuscitation efforts and minimizing CPR interruptions by virtue of allowing more rapid assessment of “pulse” following interventions such as defibrillation or epinephrine administration. Survival with shockable rhythms is better with shorter times to defibrillation, and patients who present with a non-shockable rhythm but convert to a shockable one in the ED have better outcomes if the latter is recognized and the appropriate intervention occurs.35 In this regard, TEE may be the most expedient way to detect such rhythms, particularly during an evolving code where a switch to a different part of the ACLS algorithm may be warranted. Since TEE can differentiate among low ejection fraction, electromechanical dissociation, and asystole, TEE can aid in the decision to cease resuscitative efforts since cardiac standstill (asystole) predicts negative outcome in the ED.36

A 68-year-old male presents to a community hospital with severe central substernal chest pain radiating to his neck, jaw, and back. Upon arrival to the ED, he is diaphoretic and complaining of weakness with a blood pressure of 203/99. An ECG performed on arrival reveals anterior ST segment elevation. The nearest cardiac center with catheterization capability is at least 1 hour away, and therefore the decision is made to initiate the thrombolytic protocol established at the hospital.

A portable chest x-ray is completed. The film is difficult to interpret given the patient's body habitus, but a wide mediastinum cannot be excluded. Unequal blood pressures are noted in the arms, being approximately 20 mm Hg less in the left side. Given this physical exam finding and concern for aortic dissection, a TEE probe is placed. The image in Figure 23–5 is obtained, confirming an ascending aortic dissection extending into the left anterior descending artery. Thrombolytic therapy is deferred, and the patient's blood pressure is controlled to 160 mm Hg with esmolol. An emergency consultation with thoracic surgery is obtained, and the patient is taken to the operating room (OR) for primary repair.

Aortic pathology constitutes a subset of cardiovascular disease with high morbidity and mortality; this is compounded by presentation that frequently mimics more common disease processes. ED physicians must maintain a high level of suspicion for aortic disease, as failure to consider it as part of the differential diagnosis can result in delivery of a contraindicated treatment (such as administering thrombolytic therapy in the patient thought to have an ST elevation MI) or a delay in recognition, which could jeopardize outcomes.

Acute aortic dissection occurs with an incidence of 2.9 per 100,000 per year.37 Any portion of the aorta can dissect. Aortic dissection typically begins with a tear from the lumen of the aorta through the intima into the medial layer. Subsequent propagation extends the intimal dissection away from the media. In addition to the classic aortic dissection instigated by a tear in the lumen of the aorta, spontaneous intramural hematoma is also a cause of aortic dissection. In this case, hemorrhage into the medial layer then dissects without rupture into the lumen. Intramural hematoma is more common in the descending aorta and arch, but can occur at any point along the aorta.

Classification of Aortic Dissections

Aortic dissections are classified by their location, with the most important factor being whether or not involvement of the ascending aorta is present. For example, the Stanford criteria differentiate aortic dissections by whether there is involvement of the ascending aorta, with such involvement classified as Stanford A (regardless of involvement of the descending aorta), and isolated descending aortic dissection classified as Stanford B. In the DeBakey classification, patients with both ascending and descending aortic involvement are classified as DeBakey I; isolated ascending aortic involvement DeBakey II; and isolated descending aortic involvement as DeBakey III. Additionally, isolated aortic arch dissections can occur. Patients with involvement of the ascending aorta have a higher risk of subsequent adverse outcomes, including pericardial effusion/ tamponade, rupture, dissection into the coronary arteries, and aortic insufficiency. It should be noted that most dissections will have multiple communication points between the true and false lumens, which may be important for surgical repair.38

More recent studies have suggested that intramural hematomas, intramural hemorrhage, and aortic ulcers may represent evolving aortic dissection or dissection subtypes. Therefore, a new classification scheme has been proposed (Table 23–3).39

TEE in Diagnosis of Aortic Dissection

An ascending aortic dissection is considered a surgical emergency; rapid, accurate diagnostic assessment is associated with improved survival, and hence TEE can play an important role in this assessment. TEE can reliably assess the aorta for atheromatous disease, penetrating ulcers, intramural hematomas, dissection, and rupture, as well as quantification of aortic insufficiency, if present. Although TTE provides a limited view of the ascending aorta, TEE can provide visualization of the entire length of the aorta from the aortic valve to the level of the gastroesophageal junction (although the aortic arch can be difficult to visualize). Other diagnostic modalities include aortic angiography (the traditional gold standard), helical computed tomography (CT), and magnetic resonance imaging (MRI). Each of these modalities plays an important role in the diagnosis of aortic dissection, but the risk-benefit ratio often supports TEE as the first-line imaging technique in suspected aortic injury in the ED.40

One study comparing diagnostic modalities for diagnosis of aortic dissection has shown similar sensitivities between single-plane TEE, CT, and MRI with sensitivities of 98% for all three, but with higher specificity for MRI (98%) than TEE (77%).41 Use of multiplane TEE has been shown to increase the specificity of TEE to greater than 90%.42 In another study comparing TEE with CT and aortography in the multicenter European Cooperative Study, the sensitivity and specificity of TEE was 99% and 98%, respectively; for CT, 83% and 100%; and for aortography, 88% and 94%.43 The positive and negative predictive values for TEE were 98% and 99%, respectively; for CT, 100% and 86%; and for aortography, 96% and 84%. Other studies have revealed similar findings for the high accuracy of TEE for diagnosis of aortic dissection, although some do note the lesser sensitivity for dissections in the aortic arch.44–46 The echocardiographer must also be aware of the not uncommon occurrence of artifactual echos in the assessment of aortic dissection (Figure 23–6).

In addition to having a similar sensitivity to other diagnostic modalities, TEE is also the most portable of the available options, which may be the sole determining factor for an unstable patient when transfer to the radiology suite of patient, staff, and equipment is not feasible. TEE has a relatively low cost, does not require contrast, and can provide ongoing, high-resolution image acquisition in real time. It can also provide important information for clinical decision making with regard to surgical candidacy and operative approach by virtue of being able to differentiate dissection from intramural hematoma and penetrating ulcers, localizing primary and secondary entry sites of dissection, differentiating true and false lumens, evaluating the aortic valve for regurgitation, establishing involvement of the coronary arteries, and for ruling out associated conditions such as pericardial effusion and tamponade. In addition, in the setting of trauma when the patient may be expected to have more complex injury patterns, TEE permits assessment of the heart and aortic branches.

TEE Views for Imaging the Aorta

TEE imaging of the aorta usually begins with imaging of the ascending aorta utilizing a 120° imaging plane with the probe in the mid- to upper-esophagus, allowing visualization of the proximal 5 to 10 cm of the ascending aorta. Rotation of the imaging plane to 30° to 60° allows visualization of short-axis views of the proximal ascending aorta and the aortic valve. Next, the probe is moved deeper into the esophagus towards the gastroesophageal junction; rotation of the probe to the left with use of a 0° scanning plane will result in a posterior facing plane and visualization of the descending aorta.34 The TEE probe is then slowly withdrawn, allowing visualization of the length of the descending aorta through a series of short-axis views, and at several points along the aorta the imaging plane should be rotated to 90°, allowing a longitudinal view of the aorta. As one transitions more proximally in the aorta, the probe position is often less well tolerated by awake patients, making visualization of the aortic arch difficult. The arch is visualized by withdrawing the probe to the level of the left subclavian artery and clockwise rotation with a 0° scanning plane to obtain an elongated view of the arch; rotation to a 90° scanning plane reveals a short-axis view of the arch. Clockwise and counter-clockwise rotation allows full evaluation of the arch with visualization of the takeoff of the great vessels.

Aortic dissections are visualized as an intimal “flap” undulating within the aorta, often demonstrating a “true lumen” and a “false lumen” (Figure 23–7). Intramural hematomas will have the appearance of a smooth, homogeneous thickening of the wall corresponding to thrombus formation between the intima and adventitia (Figure 23–8), are a precursor for aortic dissection, and should be treated like an aortic dissection. In fact, 15% to 20% of patients with aortic dissection may present with intramural hematoma.47 The etiology of concurrent aortic insufficiency can also be characterized by TEE, which can occur as a result of dilation of the sinotubular junction and resultant aortic valve malcoaptation, dissection propagation into the sinus of Valsalva, and prolapse of the dissection flap through the aortic valve orifice.

Differentiation of True Aortic Dissection from Artifacts on TEE

Differentiation of true aortic dissection flaps from artifact can be accomplished using several imaging views and using color-flow imaging (a true flap will demonstrate margination of flow, whereas artifact will not disrupt the color Doppler signal). Further, a true flap will show mobility as opposed to artifacts appearing more rigid and fixed in location. Artifacts (see Figure 23–6) are often caused by side lobes of the aortic sinotubular junction, and hence their intensity diminishes in the lumen, whereas a true dissection will not lose its echo density along its course.38 In addition, venous structures can create an adjacent echo creating the appearance of a linear density suggestive of a dissection flap. Color-flow imaging of this will show that the larger lumen contains pulsatile flow (ie, the true aorta) and that the smaller lumen contains continuous flow typical of a venous pattern.38 Agitated saline contrast can also be injected to differentiate the venous structure (which is often the brachiocephalic vein). Vignon et al48 conducted a TEE study of patients with a high risk of aortic dissection (n = 188) or traumatic disruption of the aorta (n = 42; covered later in this chapter), with comparison to final confirmed clinical diagnoses. They found linear artifacts within the ascending and descending aorta in 59 of 230 patients (26%) and 17 of 230 patients (7%), respectively.48 TEE findings associated with linear artifacts in the ascending aorta (as opposed to true aortic dissection or traumatic disruption of the aorta) included (1) displacement or movement paralleling the aortic walls; (2) similar blood flow velocities on both sides of the “flap”; (3) angle with the aortic wall greater than 85° in short axis; and (4) thickness of “flap” greater than 2.5 mm. Further, none of the patients with linear artifacts within the ascending aorta (as opposed to true aortic pathology) had an associated pericardial effusion or evidence for an entry tear. In a subsequent prospective series of 121 patients, systematic use or these diagnostic criteria resulted in improved TEE specificity for the identification of intra-aortic flaps.48

A 29-year-old female was brought in by EMS due to presumed traumatic injuries secondary to a high-speed motor vehicle accident. She was combative and was intubated on-scene. Her initial vital signs showed a blood pressure 80/38, heart rate 122. She had visible facial lacerations, extensive bruising over the precordium, and a slightly distended abdomen with bruising corresponding to the area presumed to be under the location of the lap belt. After completion of the primary survey, the decision was made to take the patient to the OR for explorative laparotomy. Prior to transfer a single portable chest film revealed a widened mediastinum with a mild left pleural effusion. A TEE was able to be rapidly completed and examination of the aorta revealed partial disruption of the aorta (Figure 23–9). Thoracic surgery was immediately consulted, and the patient was taken to the OR for surgical repair.

Blunt or penetrating chest trauma can result in myocardial contusion or rupture, pericardial effusion and/or tamponade, hemopericardium, aortic and/or other vascular disruption, septal defects or fistula, and valvular regurgitation.49 Trauma patients can thus present in cardiogenic shock; in this situation, the most common causes are cardiac tamponade and ventricular akinesia.50,51

Both TTE and TEE can be useful diagnostic tools in these situations; however, TEE has been shown to be superior to TTE in the recognition and management of patients with cardiovascular injuries due to blunt chest trauma. For example, in one study of 134 consecutive patients suffering from severe blunt chest trauma, TTE resulted in adequate visualization in only 38% of patients (compared with 98% by TEE).52 In another study of intubated multiple injury patients not confined to blunt chest trauma, TEE was able to detect unsuspected myocardial contusion, pericardial effusion, and aortic injury.53 Therefore, although there are no randomized controlled trials, the observational evidence supports that in patients with suspected traumatic aortic injury TEE may be used as a first-line evaluation, especially if hemodynamic instability is present.54

Traumatic Aortic Injury in Patients With Chest Trauma

Traumatic aortic injury can range from small intimal tears to catastrophic rupture of the aorta, which usually causes rapid, lethal hemorrhage. Minor arterial wall lesions such as a mural hematoma or a limited intimal flap often have a benign course and regress spontaneously.55 Pseudoaneurysms can have a more insidious course, and may expand and/or rupture, leading to thromboembolic events, fistualization, or compression of nearby structures.55,56

Aortic injury should be particularly suspected when sudden deceleration is present in the history. In fact, up to 18% of deaths in the setting of high-velocity accidents are secondary to rupture of the aorta,57 and blunt aortic injury is the second most common cause of death in studies of blunt trauma deaths.4 Unfortunately, 60% to 85% of patients with traumatic aortic injuries die at the site of the accident or within hours of hospital admission.54,58,59 However, because of improvements in prehospital emergency care, an increasing number of patients with partial aortic rupture arrive at EDs for treatment. Prompt recognition and treatment of traumatic aortic injuries in such patients is vital for survival. Unless suspected early, this partial thickness rupture may not be noticed, especially since 70% to 90% of cases of thoracic trauma are associated with multiple injuries,60 and clinical attention may be focused (sometimes appropriately) on other even more emergent pathology. However, it is critical to assess possible aortic injuries in a timely fashion, since the goals of ACLS—ie, actively resuscitating someone to a target blood pressure of, perhaps, 90 mm Hg—may be in contrast to optimal management for aortic rupture, where one aims for a reduction in shear forces and hence treats with controlled hypotension.60

Furthermore, when secondary to blunt chest trauma, aortic injury frequently coexists with myocardial contusion, which can precipitate myocardial infarction or cardiac tamponade, culminating in overt cardiac failure. Early recognition may allow treatment measures to be instituted quicker. Up to 40% of patients with traumatic injury to the aorta die within 24 hours without surgical intervention.61

TEE in the Evaluation of Traumatic Aortic Injury in Patients With Blunt Chest Trauma

TEE plays a key role in the diagnosis of traumatic aortic injury in the patient with blunt chest trauma. In one study, 32 consecutive trauma patients were prospectively evaluated with TEE and the findings compared with aortography, surgery, or necropsy.62 The authors found two subsets of aortic injury with distinct echocardiographic signs: (1) subadventitial traumatic disruption of the aorta (n = 10), and (2) traumatic intimal tears (n = 3). Eighteen patients had a normal TEE confirmed by aortography. One 2-mm medial tear was missed by TEE (seen on necropsy). They found that the sensitivity of TEE for the diagnosis of subadventitial aortic disruption was 91% and the specificity was 100%.62 The authors conclude that TEE should be considered as first-line imaging for the evaluation of trauma patients with suspected injuries of the thoracic aorta. Another case series of 101 patients presenting to the ED with a diagnosis of possible traumatic rupture of the aorta who were evaluated simultaneously by TEE and aortography showed that TEE could be successfully performed in 93 patients (others not able to be completed because of lack of cooperation or maxillofacial trauma). No TEE-related complications were noted, and the results showed a sensitivity of 100% and specificity of 98% for the detection of injury to the aorta.63 A meta-analysis has shown a maximum joint sensitivity and specificity of TEE of 97%, corresponding to a false-positive and false-negative rate of 3%.54 When only considering studies that compared TEE with the gold standard of aortography, the maximum joint sensitivity and specificity of TEE was slightly lower (93% vs 95% with aortography). However, TEE performed better than aortography in patients with small aortic lesions not requiring surgical repair.54 Such lesions, though easily missed by aortography, require careful follow-up and can have an unpredictable clinical course.54 Some have advocated that TEE be routinely performed in patients with violent deceleration collisions even with a normal chest x-ray.62,64 However, the limitations of TEE should be noted, including limited visualization of branch and proximal arch disruption, which is better visualized with aortography.65 Of course, TEE should not be attempted in uncooperative, combative patients, or in those with unstable neck injuries.

Diagnosis of Traumatic Aortic Disruption

The most common sites of traumatic aortic injury are the aortic isthmus (at the junction of the relatively mobile aortic arch and the relatively fixed descending aorta) and the ascending aorta just proximal to the origin of the brachiocephalic vessels.54 Complete aortic transection is a fatal event, but patients with partial aortic disruption may survive and present to the ED. Of the 20% of such patients who survive to reach the ED, 40% die within the first 24 hours.4 The diagnosis is suggested by history of a deceleration accident and mediastinum greater than 8 cm on chest radiograph. Although CT or angiography has been the primary diagnostic modality, TEE is becoming the first approach in many centers because of the speed and superiority in evaluating aortic disease such as dissection.

The diagnosis of aortic disruption/transection requires the presence of a disrupted aortic wall with blood flow on both sides of the disruption.62 In the previous case series, TEE findings associated with aortic disruption included the presence of an abnormal intraluminal flap just distal to the aortic isthmus, corresponding to the disrupted wall and consisting of the entire intimal and medial aortic layers, which was thicker than those seen in patients with aortic dissection (4.2 ± 0.8 mm versus 2.2 ± 0.7 mm, respectively), and were usually extremely mobile and oscillating.62 This “thick flap” was usually accompanied by a regional deformity of the aortic isthmus contour due to formation of an acute localized pseudoaneurysm. In subtotal aortic disruptions (n = 8), this medial flap appeared as a linear structure in a transverse view that corresponded to a spiral tear at surgery. In complete aortic disruption (n = 1), the thick flap appeared as an open circle within the aortic lumen.62 In partial disruption, the aortic tear was visualized as a discontinuity of the intimal and medial aortic layers, with color flow entering a pseudoaneurysm consisting of the adventitial layer.

With aortic trauma where there has been some partial disruption through the media into the adventitia, often a periadventitial hematoma is encountered. Varying degrees of dissection and intimal tear may be visualized with the TEE probe, or occasionally a focal area of the aorta is visualized where the circular geometry of the aorta is lost and a limited ridge may be seen protruding into the lumen of the aorta.38 Formation of thrombus within the medial space or in the aortic lumen can also occur, and hence, if such a phenomenon is encountered in a young patient after chest trauma, aortic trauma rather than atheroma should be considered. Other consequences of aortic trauma that can be visualized with TEE include acute rupture of the sinus of Valsalva (typically into the right atrium) and formation of aorta vena caval fistulae, which are visualized as high-volume turbulent flow in the inferior vena cava.38

Specific signs of thoracic aortic branch vessel injury are rarely present. For example, pseudocoarctation or decreased blood pressure in the left arm occurs in only 5% of patients with rupture of the aortic isthmus.66 Clinical signs of injury to an arch artery are more common and include cervical or supraclavicular hematomas, bruits, and diminished peripheral pulses.67 Patients with rupture or dissection of a common carotid artery can present with coma or hemiparesis, and laceration of the brachial plexus (frequently associated with rupture of the subclavian artery), can present with denervation of the arm.67

Table 23–4 details echocardiographic criteria for differentiating between traumatic aortic disruption and acute aortic dissection. Based on their TEE study and previous categorization of nonpenetrating traumatic aortic injuries, Parmley et al61 suggest a new classification of distinct traumatic aortic injuries: (1) traumatic aortic intimal tears where the integrity of the aortic medial and adventitial layers is preserved—these lesions appear to regress spontaneously; (2) subtotal aortic disruption where more than two-thirds of the aortic wall circumference is involved; (3) complete aortic disruption with involvement of the entire aortic circumference (in both subtotal and complete disruptions, medial flaps are visualized); and (4) partial aortic disruption appearing as a limited discontinuity of both intimal and medial layers.61

Echocardiography in the Evaluation of Cardiovascular Injury in Patients With Penetrating Chest Trauma

In patients presenting with penetrating chest trauma (ie, gunshot wound, stabbing, etc), surgical exploration is often required to exclude cardiac injury, but carries a negative exploration rate of 80%. When compared with the usual surgical subxiphoid pericardiotomy, TTE has been found to be 96% accurate, 97% specific, and 90% sensitive in detecting pericardial fluid in juxtacardiac penetrating chest wounds,68 and hence TTE may prevent unnecessary surgical exploration. In another series of patients with penetrating chest trauma, TTE had an accuracy of 99.2% and positive and negative predictive values of 100% and 98%, respectively.69 However, others have reported that a normal TTE does not always exclude significant injury; for example, when hemothorax is associated with penetrating injury, TTE is not an adequate replacement for surgical exploration.70

A 47-year-old female with a history of diabetes, hypertension, and breast cancer has an acute onset of chest pain and shortness of breath at rest. Her symptoms progress and she presents to the ED for evaluation. She is initially tachycardic and diaphoretic with a blood pressure of 95/64. Before a 12-lead ECG can be performed, she loses consciousness and is found to have a repeat blood pressure of 63/48. Intravenous fluid boluses are given with no response. Pressors are initiated with some improvement but with sustained hypotension. A 12-lead ECG shows sinus tachycardia with a right bundle branch block. A bedside TEE is performed, which reveals marked right ventricular dysfunction and enlargement (Figure 23–10). A subsequent CT scan confirms a large saddle pulmonary embolus and the patient is successfully treated with thrombolytics.

Echocardiography can be a valuable tool in evaluation of the patient with unexplained hypotension (see Table 23–2). Both TTE and TEE can help quickly distinguish among a primary cardiac etiology resulting in decreased cardiac output versus a noncardiac etiology such as hemorrhage with hypovolemia. While TTE may provide sufficient imaging in the critically ill patient, visualization of cardiac structures is often impaired and TEE is therefore required. In one series of 2508 TEE studies in critically ill patients over 15 years, hemodynamic instability was the most frequent indication for TEE (39%), followed by suspected endocarditis (19%), assessment of ventricular function (9%), suspected aortic disease (8%), suspected pericardial tamponade (3%), and chest trauma (1%).71 Similarly, Burns et al72 demonstrated in acutely ill trauma patients that approximately two-thirds of patients improved because of changes in management resulting from the TEE findings. For example, several patients with what would be deemed as having acceptable filling pressures based on pulmonary arterial catheterization in fact had inadequate filling when visualized on TEE.72 In another study of 308 critically ill patients, 40% of patients were evaluated by TEE for hypotension; of these, the cause of the hypotension was identified in 67% of the cases, leading to a management change in 31% of patients.73

TEE Assessment of Cardiac Etiologies of Hypotension

Echocardiography can delineate the cardiac etiology resulting in hemodynamic instability, ie, cardiogenic shock, acute valvular insufficiency, or concurrent cardiac abnormalities. One of the mainstays of echocardiographic evaluation in patients with unexplained hypotension is left and right ventricular systolic function assessment, which can usually be accomplished by real-time visualization alone through a combination of short-axis and long-axis views of the left and right ventricles. In one study, 115 critically ill patients were evaluated by TEE, with the most common indication being hemodynamic instability (67% of patients).74 Of these, 26% were found to have left ventricular dysfunction with left ventricular ejection fraction of less than 30%, suggesting that this is not an uncommon cause of hypotension in ill patients. In several series, echocardiography has been found to be more reliable than Swan-Ganz catheter pressure in determining the cause of hypotension.75–77 Further, left ventricular dysfunction can occur in the setting of septic shock,78 and hence left ventricular function assessment in such patients can be important to guide subsequent therapy.

In addition to ventricular dysfunction, cardiac hypotension can be caused by pericardial tamponade and/or severe valvular disease. In a study of 61 critically ill patients with unexplained, sustained (>60 minutes) hypotension, a TEE-guided diagnosis of nonventricular limitation to cardiac output (ie, valvular or pericardial) was associated with improved survival to discharge (81%) versus a diagnosis of ventricular disease (41%) or hypovolemia/low systemic vascular resistance (44%).76 TTE was inadequate in 64% of these patients, and TEE contributed new, clinically significant diagnoses (ie, not seen by TTE) in 28% of patients, leading to surgical interventions in 20%. The authors suggest that TEE makes a clinically important contribution to the diagnosis and management of unexplained hypotension and is prognostic in critically ill patients.76 TEE can also diagnose myocardial infarction and related mechanical complications as detailed earlier in this chapter.

TEE Assessment of Hypovolemia as a Cause of Hypotension

Echocardiography can represent an important tool for the diagnosis of hypovolemia, documenting a small left ventricular volume and hyperdynamic motion. With TEE, left ventricular volume can be measured by subjective assessment of LV size or by quantitative determination of left ventricular cross-sectional area at end-diastole using the transgastric short-axis view at the level of the midpapillary muscles.79 In some patients, often those with a history of hypertension, an acquired dynamic left ventricular outflow tract obstruction can occur when volume depletion occurs, with the same hemodynamic consequences as obstructive hypertrophic cardiomyopathy. Elevated left ventricular outflow tract gradients and systolic anterior motion of the mitral valve with secondary mitral regurgitation may be seen, resulting in progressive hypotension and development of a systolic murmur. Gradients exceeding 100 mm Hg have been noted, and hemodynamic assessment with pulmonary artery catheterization may be misleading, showing an elevated pulmonary capillary wedge pressure suggesting an increased left ventricular filling volume.38 This abnormality resolves with volume repletion and removal of pharmacologic agents aimed at increasing contractility and/or reducing vascular resistance.

TEE Assessment of Pulmonary Embolism as a Cause of Hypotension

Pulmonary embolus (PE) has been shown to be the cause of almost 5% of cardiac arrests, with PEA being the initial rhythm in 63%.32 In fact, one small study of 25 patients presenting with PEA in whom TEE was routinely performed showed that 14 had evidence of right ventricular enlargement; of these nine were subsequently found to have pulmonary embolism, with the remainder having cardiac contusion, right ventricular infarction, right ventricular hypertrophy, or cor pulmonale.80 Diagnosis of PE in the setting of cardiac arrest is an important one as prompt treatment with thrombolytics has been shown to result in a significantly higher return of spontaneous circulation than in those patients not receiving thrombolytics (81% vs 43%, respectively).32

PE is diagnosed on TEE by findings suggestive of acute right ventricular pressure overload such as right ventricular enlargement and/or hypokinesis in the context of an appropriate clinical scenario. In the absence of myocardial infarction, significant left-side valve or ventricular disease, or known pulmonary disease, this finding indicates a high probability of pulmonary embolus.4 This can often be accompanied by right atrial enlargement, tricuspid regurgitation, dilation of the inferior vena cava and flattening of the interventricular septum with a “D-shaped” left ventricular geometry in the short-axis views due to pressure overload on the right ventricle, and paradoxical septal motion. Of note, right ventricular dysfunction can also be due to an inferior myocardial infarction, myocardial contusion (ie, in the setting of trauma), sepsis, and acute sickle-cell crisis.10 Often termed “McConnell sign,” a relatively specific sign of right ventricular dysfunction due to pulmonary embolism (as opposed to other causes of right ventricular dysfunction) is characterized by a distinct regional pattern of right ventricular (RV) dysfunction with normal RV apical motion but akinesia of the mid-free RV wall.81 This regional RV dysfunction was found to have a sensitivity of 77%, specificity of 94%, positive predictive value of 71%, and a negative predictive value of 96% for the diagnosis of PE.

It should be noted that the echocardiographic findings of pulmonary embolus are typically seen after acute obstruction of more than 30% of the pulmonary arterial bed.82 However, a recent small study has suggested that a decreased acceleration time of pulmonary artery outflow is associated with even small pulmonary emboli obstructing less than 25% of the pulmonary vasculature (acceleration time in pulmonary embolism 85 ± 22 milliseconds vs 117 ± 35 milliseconds in controls without pulmonary embolism).83 This needs to be further validated before being used clinically, but suggests additional echocardiographic criteria to discriminate smaller pulmonary emboli. In addition, some studies have suggested that the right ventricular to left ventricular end-diastolic diameter ratio (RVEDD/LVEDD) may be an important prognostic factor in patients with pulmonary embolism. In one study of 950 patients hospitalized for acute pulmonary embolism, it was found that the sensitivity and specificity of an RVEDD/LVEDD greater than 0.9 for predicting hospital mortality were 72% and 58%, respectively, with the ratio also being an independent predictor of hospitality mortality (odds ratio 2.66).84

Although rarely seen on TTE, TEE may be able to directly visualize pulmonary emboli,85 usually in the main or right pulmonary arteries (Figure 23–11), with limited sensitivity for more distal or left pulmonary arterial clots. In one study of intraoperative TEE in 46 patients with known PE immediately prior to pulmonary embolectomy, using definitive location of thromboemboli determined from the surgical record as the “gold standard,” they found that echocardiographic evidence of right ventricular dysfunction was present in 96% and leftward interatrial septal bowing was present in 98% of examinations. However, the sensitivity of TEE for direct visualization at any specific location was only 26%, with the least sensitivity for the left pulmonary artery.22 Another study of TEE in patients with known severe PE showed that central pulmonary thromboemboli were directly visualized in 58.3% of patients evaluated, leading the authors to conclude that TEE seems to be a useful method for the diagnosis of severe PE, with the potential ability to clarify the diagnosis within a few minutes without further invasive diagnostic procedures.85 Regardless, the combination of indirect echocardiographic criteria in combination with potential direct visualization of pulmonary thromboemboli make TEE a valuable tool in the diagnostic workup.

An additional use of echocardiography in the ED and ICU is evaluation of unexplained hypoxia. In such situations, echocardiography can help distinguish causes such as intracardiac shunt, pulmonary embolus, pulmonary hypertension, and right ventricular failure. Often TTE in such critically ill patients is insufficient, and TTE is often insensitive for evaluation of intracardiac shunts, necessitating TEE evaluation. If such an evaluation excludes a primary cardiac abnormality, the ED physician can assume that the cause of the hypoxia is extracardiac.

Echocardiography remains a mainstay in the diagnosis and management of acutely ill patients presenting to the ED. Given limitations to TTE imaging in this group of patients, TEE can serve as a useful and accurate bedside modality, allowing rapid assessment of a wide range of cardiovascular pathologies observed in patients presenting to the ED.

We thank Dr. Robert Preston for his editorial assistance and drafting of the questions associated with this chapter.