2004 Vol.22 Issues 3 Ultrasound in Emergency Medicine

Emerg Med Clin N Am 22 (2004) xvxvi
Preface
Ultrasound in Emergency Medicine
Carlo L. Rosen, MD Richard E. Wolfe, MD Guest Editors
Ultrasound has become an integral part of the practice of emergency medicine and trauma care. It now is routinely used in most emergency medicine residencies and level I trauma centers. It has become part of the Advanced Trauma Life Support guidelines as a diagnostic tool in trauma patients and is a skill that is part of the required training of emergency medicine residents. The use of ultrasound in emergency medicine has its roots in Japan and Germany where it has been used for years as an adjunct to the physical examination for the diagnosis of intraperitoneal bleeding in trauma patients. Emergency ultrasound refers to an examination performed at the bedside by an emergency physician. This examination has limited goals that allow the physician to make rapid decisions, identify life-threatening diagnoses, and expedite the operative management of emergency patients. This type of examination is very dierent from that performed by a radiology technician, radiologist, or cardiologist. The scope of practice of emergency ultrasound includes its use for detecting intraperitoneal bleeding in trauma patients; for diagnosing pericardial eusion and tamponade; for detecting abdominal aortic aneurysm, cholelithiasis, cholecystitis, and hydronephrosis in patients with abdominal or ank pain; and for evaluating pregnant patients with vaginal bleeding or abdominal pain. Other current uses include the detection of deep venous thrombosis, pelvic masses, and as a procedural aid primarily for venous access. Newer uses of ultrasound include the detection of testicular masses and torsion. As the experience and expertise of emergency physicians with this technique
0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.013
xvi
C.L. Rosen, R.E. Wolfe / Emerg Med Clin N Am 22 (2004) xvxvi
increases, more uses for emergency ultrasound are being developed and dened. Examples of these include the use of ultrasound for musculoskeletal applications, for detecting ophthalmologic pathology, and for extending the use of this technique into the prehospital setting. This issue of the Emergency Medicine Clinics of North America uses an evidence-based approach to characterize the current state of the use of ultrasound in emergency medicine. It also includes a discussion of future uses of emergency ultrasound. Experts in the technique of ultrasound in emergency medicine have been assembled to contribute to this issue. These are the people who have worked hard to pioneer the use of emergency ultrasound. Others who have contributed to this issue are Maureen Murphy, who deserves special credit for her work on compiling the articles, and Karen Sorensen, Clinics Editor, who deserves credit for her help with the production of this issue. Carlo L. Rosen, MD Beth Israel Deaconess Medical Center Harvard Aliated Emergency Medicine Residency Harvard Medical School One Deaconess Road, West Clinical Center 2 Boston, MA 02215, USA E-mail address:crosen2@bidmc.harvard.edu Richard E. Wolfe, MD Department of Emergency Medicine Beth Israel Deaconess Medical Center Harvard Medical School One Deaconess Road, West Clinical Center 2 Boston, MA 02215, USA E-mail address:rwolfe@bidmc.harvard.edu
Emerg Med Clin N Am 22 (2004) 581599
Ultrasound in abdominal trauma

John S. Rose, MD
Department of Emergency Medicine, University of California Davis Medical Center, 2315 Stockton Blvd., PSSB 2100, Sacramento, CA 95817, USA
Ultrasound in the evaluation of abdominal trauma has evolved over the past 30 years. The use of ultrasound for abdominal trauma was described initially by Kristensen and colleagues [1] in 1971. In 1976, Ascher and colleagues [2] rst reported the accuracy of ultrasound in Radiology, with 80% sensitivity for the detection of splenic injury. In a study of 808 patients, Tiling and colleagues [3] in 1990 reported a sensitivity of 89%, a specicity of 100%, and an accuracy of 98%. This same group also was rst to comment on the eect of training and experience and reported that surgeons with extensive ultrasound experience could diagnose intra-abdominal uid with a sensitivity of 96% and an accuracy of 99%. Interest and experience with ultrasound for trauma grew steadily around the world among surgeons and emergency physicians during the early 1990s [47]. During this period, ultrasound technology was improving with regard to price, portability, and resolution, allowing its use during resuscitation. At the same time, in the United States, there was continuing reliance on diagnostic peritoneal lavage (DPL) and CT and much less interest in sonography for abdominal trauma. This all changed when emergency physicians and surgeons in the United States began to publish their experience with ultrasound [4,8,9]. The term Focused Assessment with Sonography for Trauma (FAST) was coined by Rozycki et al [10] in 1996 and has persisted as the accepted acronym for the trauma ultrasound evaluation. The basic four-view examination (perihepatic, perisplenic, pelvic, and pericardial views) has become the foundation of the FAST examination. The rapid, noninvasive, and practical nature of ultrasound for bedside evaluation of critically injured patients has changed the evaluation of blunt abdominal trauma.
E-mail address: jsrose@ucdavis.edu 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.007
582
J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599
Technique In the supine patient, the hepatorenal space is the most dependent area and the least obstructed for uid ow. Fluid in the abdomen can move freely up the right pericolic gutter into this space. The left pericolic gutter is higher and the phrenicocolic ligament blocks the ow; consequently, uid tends to ow to the right pericolic area. On the right, uid ows into Morisons pouch, the potential space in the hepatorenal recess (Figs. 1 and 2). On the left, uid ows preferentially into the subphrenic area and not into the splenorenal area (Fig. 3), which is important because the subphrenic area may be dicult to visualize due to bowel gas and splenic exure gas. Fluid in the pelvic region ows to the retrovesicular area in the male patient and to the pouch of Douglas in the female patient because these areas are the most dependent areas of the pelvis. Given these anatomic relationships, the FAST examination has evolved into three to ve intraperitoneal views and one cardiac view. The FAST scan can be completed in less than 5 minutes and may involve up to six views (Fig. 4), depending on examiner preference. These views include (1) a subxiphoid or parasternal view to detect pericardial uid; (2) a right upper quadrant view to assess the hepatorenal interface (Morisons pouch) and right chest; (3) the right paracolic gutter; (4) a left upper quadrant view to assess the splenorenal interface and left chest; (5) the left paracolic gutter; and (6) longitudinal and transverse pelvis views to look for free uid adjacent to the bladder [11]. Although the bladder is not a peritoneal organ, a full bladder greatly enhances the detection of free uid for the pelvis view (Fig. 5) by giving readily identiable landmarks and providing an acoustic window. The pericolic gutter views
Fig. 1. Normal view of hepatorenal interface (Morisons pouch).
583
Fig. 2. Free uid in hepatorenal interface.
are optional views that some providers use (Fig. 6). Their role in improving the accuracy of the FAST scan has not been studied. The focus of the FAST examination is the detection of free uid; however, during the procedure, specic organs occasionally may be visualized, providing potential injury localization. Given the ow of uid in the abdomen and free ow toward the right, the utility of a single right-sided view has been examined. Ma and
Fig. 3. Fluid in splenorenal interface. Note that uid in the subphrenic space is larger than in the splenorenal space due to the phrenciocolic ligament.
584
Fig. 4. FAST examination views.
colleagues [12] compared a single view of Morisons pouch with a standard ve-view FAST examination for the detection of hemoperitoneum. Sensitivity of the complete FAST examination versus the single view for hemoperitoneum was 87% and 51% respectively. In a blinded prospective
Fig. 5. Longitudinal pelvic view with intraperitoneal uid seen outside of the bladder.
585
Fig. 6. Fluid in pericolic gutter. Note oating bowel loops.
study, Ingeman and associates [13] demonstrated that for single views of the abdomen for hemoperitoneum, the hepatorenal view was superior to the splenorenal or pelvic views with regard to sensitivity. In a multicenter prospective trial, Rozycki et al [14] found that the hepatorenal interface was the most common location of free uid in blunt abdominal trauma associated with signicant parenchymal injuries. Although a single hepatorenal view has been advocated in certain situations [15], a multiple-view FAST examination is still recommended because the sensitivity is higher [11,12]. The role of the FAST examination for detecting pericardial uid also is very important. Work originally done by Plummer [16] in 1992 demonstrated the value of bedside ultrasound in the evaluation of penetrating cardiac injuries. In this study, focused bedside echocardiography was performed on 49 patients with penetrating cardiac injuries who were compared with a control group of patients who did not undergo ultrasonography. The ultrasound group had a signicantly faster rate of diagnosis and disposition and higher survival rates (100% versus 57%). A multicenter trial by Rozycki et al [17] in 1999 demonstrated similar results. In 261 patients, ultrasound was found to have a sensitivity of 100% and a specicity of 97% for penetrating cardiac injuries. Thus, bedside echocardiographic evaluation as a component of the FAST examination has proved its diagnostic role for penetrating cardiac injuries. The role of bedside ultrasound in blunt chest trauma patients for evaluating myocardial contusion, rupture, and valve injury is much less clear and involves an echocardiographic examination that is beyond the scope of the limited bedside ultrasound examination and this article.
586
Clinical implications of the focused assessment with sonography for trauma examination results Much of the initial research into the use of ultrasound in trauma focused on the test characteristics of ultrasound for detecting hemoperitoneum and on establishing its role in blunt abdominal trauma. Indeed, when describing the history of the FAST examination, sensitivity and specicity dene its usefulness compared with DPL and CT. From a purely statistical sense, sensitivity and specicity calculations rely on all injuries being conrmed or excluded by a gold standard test; however, this is impractical in much of clinical research. A signicant number of trauma ultrasound studies have used these clinical outcomes and, thus, sensitivity and specicity have been calculated using dierent outcomes and gold standards, depending on the study. For example, a false negative in one study may be a true negative in another if the patient required no operative intervention for the injury. This situation is exemplied by a study performed by McGahan et al [18] that evaluated the use of ultrasound in 500 blunt trauma patients. This group compared ultrasound ndings to CT or laparotomy ndings as the gold standard, not clinical observation. They reported a sensitivity of 63%, a specicity of 95%, and an accuracy of 85% for the detection of hemoperitoneum and organ injury. This calculation was based on very conservative positive and negative criteria, dened as the presence of any injury regardless of whether a therapeutic procedure was performed. If a study endpoint like clinical outcome or observation was used, however, then their reported sensitivity would have increased to greater than 90%. When assessing ultrasound in the evaluation of the blunt trauma patient, calculations such a sensitivity and specicity may be misleading. Consequently, the question should be, How does this ultrasound result aect the patient for whom I am caring right now? In detecting intra-abdominal injuries in trauma patients, the governing paradigm always has been to quickly recognize those patients who require laparotomy and to prevent further morbidity or mortality. Underlying this paradigm is the understanding that ongoing evaluation through serial examinations or other imaging will be needed to determine whether a patient has any intra-abdominal injury. Most experts would concur that ultrasound has performed best when limited to detecting free intraperitoneal uid [10,11,19]. Historically, the presence of any intraperitoneal uid indicated a signicant intra-abdominal injury and warranted an immediate exploratory laparotomy. Over the past decade, however, the practice of nonoperative management for intra-abdominal injuries in adults and children has increased [2022]. This changing practice has relied on the increasing use of high-performance CT to image the abdomen. The availability of rapid CT scanners has allowed many trauma centers to develop criteria for obtaining abdominal CT scans during the resuscitative phase [23,24]. Thus, patients who previously would have had an ultrasound and serial
587
examinations as the primary means of detecting abdominal injuries now are taken quickly to a high-speed CT scanner and undergo extensive body imaging. Given this practice, the question of sensitivity and specicity is a question not only of detecting an intra-abdominal injury with ultrasound but also of detecting intra-abdominal injuries that require laparotomy. The more important question is, What is the role of ultrasound in determining the need for laparotomy? The presence of free uid is only one surrogate indicator of serious intra-abdominal injury [25]. A small amount of uid in a stable patient with a small liver laceration is much dierent than a large amount of uid in an unstable patient with a high-grade splenic disruption. Both examinations are positive, but each patient falls into a very dierent management category. A 1999 consensus conference on the performance of ultrasound in trauma examined the question of the positive and negative studies and how they apply to clinical practice. This report suggested the following guidelines [11]: in hemodynamically unstable patients, a positive FAST examination generally should be followed by a laparotomy, and a negative FAST examination warrants examination for an extra-abdominal source of hemorrhage. In hemodynamically stable patients, a positive FAST examination should be followed by an abdominal CT scan to better dene the injury, and a negative FAST examination should be followed with serial exams for 6 hours and a follow-up ultrasound or abdominal CT scanning, depending on the clinical scenario. Although this practice paradigm has not been validated, it represents a consensus among experts. Variations in this practice, however, are commonplace. For example, given a signicant mechanism and a negative FAST examination, a stable patient may still undergo CT scanning [25]. An unconscious patient with severe head trauma requiring operative intervention who has a negative FAST examination still may undergo CT scanning to provide preoperative clearance. Suce it to say that stable patients with a negative FAST examination dene a more ambiguous population. Patients can have signicant intra-abdominal injuries and not have free uid or they may have delayed uid accumulation, especially if the examination is performed shortly after the injury. Does the FAST examination add much if the patient is going to undergo an abdominal CT anyway? Given the speed of the average FAST scan, it likely would not delay care. One could make an argument that the FAST examination allows for better resource use in multipatient scenarios or in situations where a single emergency physician is managing the trauma patients alone. In addition, smaller hospitals may not have CT as readily available; consequently, ultrasound would be important. Branney et al [26] developed a key clinical decision pathway using ultrasound in blunt trauma patients that reduced CT and DPL use and cost, without an increase in missed injuries. In addition, recent evidence in a randomized trial demonstrated that when ultrasound is absent, more abdominal CT scans were ordered, suggesting
588
that some abdominal CT scans are being performed as screening tests, especially in patients with a low clinical probability of intra-abdominal injury [27]. Likewise, a positive FAST examination in a stable patient also is somewhat ambiguous. The consensus document recommends CT imaging in this population, which is somewhat dierent from the 1995 American College of Surgeons Committee on Trauma recommendation [28] for a DPL, abdominal CT, or laparotomy to further evaluate for intra-abdominal injuries. The amount of uid and its location become more important when examining this population. In addition, there is some preliminary evidence that there may be criteria in this population that allow for the development of a clinical decision rule, including right upper quadrant uid, hypotension, femur fracture, abdominal tenderness, and age [60 years [29]. Fluid volume and scoring systems Fluid in the abdomen appears as an anechoic signal. Fluids such as unclotted blood, ascites, urine, and bowel uid may have a similar appearance. In 1970, Goldberg [30] demonstrated that with ultrasound, as little as 100 mL of intraperitoneal uid could be visualized in a right lateral decubitus position. Using the single view of Morisons pouch, Branney and coworkers [7] scanned supine patients undergoing DPL and found, on average, that a minimum of 619 mL was needed before free uid could be visualized in Morisons pouch by most examiners. When 1 L of crystalloid was infused, the sensitivity was 97% using the single view in their study. It is likely that most of the infused uid localized to the gravity-dependent pelvis. Current evidence suggests that in lower volumes, uid accumulates in the pelvis or near the site of injury. The acoustic window created by a full bladder enhances detection of uid in the pelvis [18]. It is not until there are larger intraperitoneal uid volumes (>500 mL) that uid is detectable in the perihepatic and perisplenic spaces [31]. A study by Abrams and colleagues [32] conrmed these ndings and, further, determined that 5( of Trendelenberg positioning resulted in the detection of free uid in Morisons pouch (668 mL in the supine position and 443 mL in Trendelenbergs position). The available data suggest that the average volume of uid detectable by the FAST scan ranges from approximately 250 mL to 620 mL [7,33]. A confounder associated with uid detection is the learning curve. As examiners gain more experience, their sensitivity improves. Gracias et al [33] demonstrated that examiners who had performed over 100 examinations were signicantly better at diagnosing smaller uid volumes. Despite these results, a discussion of the minimal detectable uid volume alone is not helpful to the clinician for decision making. As stated previously, a small amount of uid may not change clinical decision making if the patient is stable and able to undergo an abdominal CT scan. Determining which volume of intraperitoneal uid will require surgical
589
intervention is the next step. Applying a semiquantitative measure such as small, moderate, or large generally is not helpful clinically and is subject to signicant inter-rater variability. Development of a more standardized scoring system would allow for improved transfer of clinical data and overall care. Two scoring systems currently exist for the FAST examination. Huang and associates [34] in 1994 created a scoring system to estimate the amount of hemoperitoneum detected by ultrasound and applied this system to 442 patients in a prospective study. One point was assigned to each anatomic site in which free uid was detected during the FAST scan, with a score ranging from 0 to 8. Fluid of more than 2 mm in depth in the hepatorenal or the splenorenal space was given 2 points instead of 1. Floating loops of bowel were given 1 point. Ninety-six percent of patients with scores !3 required exploratory laparotomy; however, 38% of patients with scores \3 still required surgery. This system was 84% sensitive and 71% specic for quantifying hemoperitoneum greater or less than 1 L compared with intraoperative ndings. McKenney et al [35,36] developed and prospectively evaluated a scoring system that measured the depth of uid in the deepest pocket, and 1 point was added for uid in each of the other areas (four areas maximum.) In this study, 85% of patients with a score [3 required a therapeutic laparotomy, whereas 15% of patients with a score of 2 required surgery. These investigators concluded that their scoring system was better than systolic blood pressure and base decit in determining the need for therapeutic laparotomy. These scoring systems are relatively reproducible and easy to apply but have yet to be validated by other centers. In addition, they rely exclusively on uid volume scoring, without considering any clinical criteria. For example, taking the systems to an extreme, a patient could have an ultrasound score of 2 and be in hemorrhagic shock but not meet criteria for laparotomy. Although a simple scoring system would allow for reliable information transfer and could be an objective measure for serial examinations, a large trial is needed to validate such a scoring system. In addition, any useful scoring system should be easy to apply and be combined with reliable clinical variables.
Training and credentialing The training and prociency in performing a FAST examination has been an area of increasing research. Ultrasound training has been required in Germany since the 1970s. More recently, ultrasound training has been required in emergency medicine residencies approved by the Accreditation Council for Graduate Medical Education. In addition, the FAST examination has been adopted as a modular component of the American College of
590
Surgeons ultrasound training curriculum [37]. Given the increasing educational requirements of the FAST examination, data are slowly emerging outlining required experience. The 1999 consensus document addresses this issue; its major recommendation was a 4-hour didactic component, a 4-hour practical component, and 200 supervised examinations [11]. The document referred to alternative recommendations, including a similar 1-day course followed by 50 examinations. The recent American College of Emergency Physicians Ultrasound Guideline recommendations published in 2001 recommended 25 to 40 supervised examinations [38]. As evidenced by this discrepancy, the true required number for prociency seems ill-dened. Assessing ultrasound competency and prociency is important. Initial training and ongoing review is critical to the safe and eective implementation of a FAST program. Prociency requires an adequate number of supervised examinations with a signicant number of positive ndings. After an initial 1-day course, which includes didactic and hands-on training, the reported sensitivity is 81% and specicity is 91%, with an overall accuracy of 98% [39]. Sensitivity is the parameter most aected by increased practice. The rst 30 to 50 examinations demonstrate a rapid improvement in overall sensitivity. The learning curve appears to level o after 50 examinations, and improvement is more gradual until 200 examinations are performed. Little measured improvement in sensitivity or specicity is seen after 200 examinations [33]. The overall number of examinations is not the only issue in FAST training and prociency. A critical part of training is the presence of an adequate number and variability of positive examinations. The positive FAST rate is reported to be 9% to 13% [27,4042]. Thus, with 50 sequential examinations, a provider may have less than 10 positive examinations. The American College of Emergency Physicians Ultrasound Guidelines specify that 50% of examinations should be positive [38]. In 2002, Gracias et al [43] demonstrated the importance of positive examinations in overall training and prociency with FAST scanning. They found that training with peritoneal dialysis patients increased sensitivity from 43% in their control group to 87% in the study group after a 2-hour practicum. Thus, the inclusion of peritoneal dialysis models is a recommended adjunct in FAST training. Standardizing the training experience also is important. Objective Structured Clinical Examinations (OSCEs) are used in several areas of surgical education. Sisley et al [44] used an OSCE developed for FAST training to assess knowledge and interpretation skills. The OSCE was eective in measuring improvement after a standardized FAST training course. The OSCE best measured factual knowledge improvement. Tools such as OSCEs allow for the comparison of dierent instruction styles and can be used to determine the ecacy of these styles in FAST education.
591
Salen et al [45] summarized many of the important issues surrounding FAST prociency and training. Their review demonstrates the lack of consensus in current FAST training. After appropriate training with didactic, practical, and experiential curriculum, they recommended 25 to 50 proctored examinations. This recommendation is very dierent from the FAST consensus document that recommended 200 examinations [11]. To add more confusion, one series [46] could nd no learning curve in a 24month study of surgical residents. It appears that most investigators nd that sensitivity and specicity begin to plateau after 25 to 50 examinations [33,39,47]. Whether it is appropriate to label this as procient or wait until maximal experience is achieved at 200 examinations is unclear. Central to this discussion of examination numbers is that experience with an adequate number of positive examinations in actual clinical situations is critical to true prociency. Nothing takes the place of scanning during an actual resuscitation. Anechoic signals are seen very dierently in a trauma patient who may be intubated and have a fractured femur. Seeing uid in positive FAST examinations may be subtle and is learned only through supervised practice. This clinical experience is central to technical prociency and reducing falsenegative examinations. Parenchymal and bowel injuries The reported sensitivities and specicities of ultrasound for detecting parenchymal intra-abdominal injuries are much lower than for hemoperitoneum [11,48,49]. Isolated solid-organ injuries without hemoperitoneum are much more dicult to detect. Brown et al [50] reported on 2693 blunt trauma patients and found that 26% of the patients with injuries had no hemoperitoneum. They also found that they could detect subtle ndings of injury in 46% of those patients, including parenchymal injuries and retroperitoneal uid. They used very experienced sonographers and did not separate therapeutic interventions from all injuries. Severe solid-organ injury may not produce sonographically detectable quantities of free intraperitoneal blood if the capsule remains intact [51]. These injuries, however, may be detected by ultrasound as aberrations in the normal parenchymal architecture of the spleen, liver, and kidney. Hematomas may be identied as cystic or mixed echogenic areas in a subcapsular or intraparenchymal distribution. Minor injuries may be isoechoic or present as a geographic hyperechoic pattern. Some injuries, such as subcapsular hematomas or bowel perforations, may not result in appreciable hemoperitoneum. These potentially lethal occult injuries may be missed with FAST examination alone. Detection of solid-organ injury requires greater skill in image interpretation that goes beyond simply searching for free uid [48]. Splenic injuries have a variable appearance. The parenchymal architecture may have a disorganized appearance, with cystic or hypoechoic regions
592
visualized. Subcapsular hematomas may appear as an echogenic or hypoechoic rim. The utility of the ultrasound for detecting free uid has been well established, but less is known about its accuracy for detecting solid-organ injuries. Richards and associates [49] recently evaluated ultrasound for detecting splenic injuries. Free uid and parenchymal abnormalities were assessed during FAST examinations, and the sonographic patterns of splenic injuries were reported. The overall sensitivity for all injuries was 69% but rose to 86% for grade III or higher splenic injuries. This series used radiologists for retrospective interpretation and the examinations were not read in real time during the resuscitation; consequently, the true practicality of evaluating for parenchymal injuries is uncertain from this data. It was concluded that inclusion of parenchymal abnormality without accompanying free uid improved the sensitivity of the FAST examination. The most common sonographic parenchymal nding was a diuse hyper- and hypoechoic pattern within the spleen. This examination, however, can be very dicult because the stomach lies close to the acoustic window of the spleen and can give a hyperechoic signal that appears to be within the spleen. This examination takes signicant practice, and its real role in the era of CT scanning is unclear. Hepatic injuries may be dicult to detect when the liver capsule remains intact. The liver has a greater volume than spleen; thus, small intraparenchymal lesions may be missed with the rapid FAST technique. In a recent study by Ohta and colleagues [52], a geographic hyperechoic liver pattern was detected with ultrasound in 33 patients and believed to represent a mild form of liver injury not requiring surgical repair. Bowel injuries from blunt trauma notoriously are dicult to diagnose, even with CT. Ultrasound is inaccurate for detecting bowel injuries, and the most common nding is free uid [53]. Clinical suspicion for this injury mandates further observation and laboratory tests. Despite research into the value of ultrasound for detecting parenchymal injuries, most of the studies have used radiologists in retrospective reviews. The added time and clinical value, given the advent of high-speed CT imaging, makes for a less compelling argument for using the FAST examination to detect parenchymal organ injuries.
The pediatric trauma patient Most advocates of FAST evaluation of pediatric trauma patients still recommend abdominal CT scans in hemodynamically stable children with positive ultrasound examinations [22,51,54]. As with adults, a CT scan is thought to be necessary to obtain detailed information regarding specic organ injuries that FAST examinations do not reliably provide. Although a negative FAST examination does not obviate the need for a subsequent abdominal CT scan, it likely provides enough extra information to decrease
593
the use of abdominal CT in children at low risk of intraperitoneal injury [55]. The FAST examination may be useful for the evaluation of pediatric trauma patients in several circumstances, although these circumstances have not been studied in a controlled fashion as yet. For hemodynamically unstable pediatric trauma patients, the FAST examination can help to identify rapidly the source of hypotension and assist decision making with regard to timing of diagnostic testing versus operative intervention [56]. For children with head and abdominal trauma, the FAST examination oers important clinical information to assist in making decisions regarding the timing of head CT and abdominal CT versus laparotomy. The FAST examination also may be of use in evaluating alert pediatric trauma patients without abdominal tenderness who otherwise would not routinely undergo radiographic imaging and who, on occasion, may have intra-abdominal injuries. The FAST examination also may allow for prioritization of imaging studies after initial evaluation and resuscitation are complete. Patients with intraperitoneal uid on ultrasound could be triaged to abdominal CT with greater expediency than those patients with no intraperitoneal uid detected on FAST examination. The additional information oered by a negative FAST examination also may be sucient for the clinician to decide against abdominal CT in pediatric patients with an already low likelihood of intra-abdominal injury. Thus, although the FAST examination routinely may not replace abdominal CT scanning of pediatric trauma patients, there are specic clinical scenarios in which ultrasound likely is useful and may enhance clinical eciency [55,56]. Controversies regarding the use of ultrasound technology in the setting of trauma were highlighted in a recent survey of general emergency physicians, pediatric emergency physicians, and trauma surgeons [57]. In the case of adult trauma patients, 91% of the respondents considered ultrasound to be somewhat to extremely useful. In the case of pediatric trauma patients, however, 73% of respondents considered ultrasound to be useful and only 57% of pediatric emergency physicians considered it useful. Only 14% of pediatric emergency departments routinely used the FAST examination for their trauma patients [57]. The investigators recommended further studies to evaluate the clinical utility of the FAST examination for pediatric trauma patients. These data suggest that the role of ultrasound in pediatric trauma patients generally is less clear than for adults. With most intra-abdominal injuries in children being managed through nonoperative management, the true role of ultrasound in children is yet to be fully claried.
Clinical algorithms The development of clinical algorithms for the use of ultrasound in trauma is a logical extension of its growing application. The FAST examination must be performed in the context of the resuscitation and
594
guided by a given clinical scenario. Luks [58] described it accurately, stating, ultrasound does not have to surpass other diagnostic modalities as long as it identies the life-threatening conditions. Emphasis has moved away from ultrasound replacing other diagnostic modalities and moved toward incorporating it into the resuscitation of trauma patients. Clinical algorithms attempt to integrate ultrasound into trauma care. Two algorithms that illustrate the dierent issues surrounding ultrasound use in abdominal trauma are shown in Fig. 7. These algorithms are derived from the FAST consensus results published in 1999. In Fig. 7A, the FAST examination is performed on every patient and then coupled with
FAST scan
Unstable hemodynamics
Stable hemodynamics
Negative
Positive
Negative
Positive CT/OR1
Other extra-abdominal source/DPL
OR YES
Serial exams/CT
Stable hemodynamics?
NO
CT and/or Serial exams (+/FAST)
FAST Positive?
YES
Laparotomy
NO/Equivocal
Other extra-abdominal source/DPL

Fig. 7. (A) FAST-oriented algorithm. (B) Hemodynamics-oriented FAST algorithm. (Adapted from Scalea TM, Rodriguez A, Chiu WC, Brenneman FD, Fallon WF Jr, Kato K, et al. Focused Assessment with Sonography for Trauma (FAST): results from an international consensus conference. J Trauma 1999;46(3):46672; with permission.)
595
hemodynamic stability. This type of algorithm is used by centers that believe that all patients should undergo a FAST scan. Fig. 7B illustrates a dierent algorithm based on the same consensus recommendations [11]. In this algorithm, hemodynamic status is the initial decision point. If stable, the patient can undergo serial physical examinations and FAST examinations or abdominal CT scanning alone, depending on the clinical scenario. If the patient is unstable, then a FAST examination is performed. If the FAST examination is positive, then the patient would go for laparotomy; if negative, then the clinician would look for an extraabdominal source of bleeding or perform a DPL if the examination was equivocal. In this algorithm, the FAST examination plays much less of a role in the stable patient. One could argue that in this algorithm, if CT is used, a FAST examination would not be needed in the hemodynamically stable patients. Although both algorithms follow the consensus recommendations, they do so with very dierent emphases: one emphasizes the initial FAST and the other emphasizes hemodynamic status. Another algorithm that is more logical and incorporates clinical and ultrasound data is the Key Clinical Pathway developed by Branney et al [26]. This pathway uses hemodynamic status rst. From this point, various clinical parameters are applied to the pathway to select patients who require CT or laparotomy. This pathway then includes more relevant clinical data such as peritoneal irritation, confounding injuries, gross hematuria, or hematocrit \35% to help answer the question of whether a patient requires immediate laparotomy. Two issues arise from this pathway. First, if a patient is unstable but has no free uid, then a DPL is recommended. Some clinicians would recommend looking for another source of bleeding and not performing a DPL. Second, if a patient has a positive FAST examination but is stable and has no peritoneal irritation, then a DPL rather than a CT scan is recommended. A patient who is severely head injured would fall into this category. This pathway gives the impression that the investigators believe the sensitivity of ultrasound is not good enough to be used in isolation. From these pathways, even more questions are generated: Why perform a FAST examination in a stable patient? Will it change decision making? As stated earlier, given the accessibility and speed of newer-generation CT scanners, severely injured patients can be imaged eectively during the resuscitation period. This practice is an area of some debate and focuses on the overall practice paradigm of a particular institution. One could argue that determining the presence of hemoperitoneum is valuable even when CT scanning is readily available. Resource use, timing, and treatment of secondary injuries will be inuenced by the FAST examination results. Algorithms that include ultrasound and CT have been developed that minimize missed injuries while limiting overuse of CT scanning [23]. Future studies are needed to examine whether there are any predictive factors present during resuscitation that help to determine whether a positive FAST
596
scan requires a therapeutic laparotomy. Whatever pathway is chosen, understanding the overall strengths and limitations of the FAST examination is critical to its safe and eective implementation into the trauma resuscitation. Future directions Future research in trauma ultrasound likely will focus on integration with current diagnostic and imaging modalities. Preliminary research that is examining the role of ultrasound for detecting traumatic pneumothorax [59], evaluating portable ultrasound units [60], and evaluating transducer frequency for parenchymal injuries is revealing exciting results [61]. More important, a large, multicenter trial to evaluate previous hypotheses and clarify issues such as scoring systems and practice pathways still is needed. Certainly, the role of trauma ultrasound in the setting of a small, rural emergency department with limited backup is very dierent than ultrasound in a busy level I trauma center. Emergency physicians must continue to be on the forefront of this work. The role of ultrasound in pediatric trauma still needs to be claried. Great strides, however, have been made in clarifying the advantages and disadvantages of the FAST examination. Ultrasound has been an important addition to the management of trauma patients and will continue to play an important role in trauma management into the twenty-rst century. References
[1] Kristensen JK, Buemann B, Keuhl E. Ultrasonic scanning in the diagnosis of splenic haematomas. Acta Chir Scand 1971;137:6537. [2] Ascher WM, Parvin S, Virgilio RW. Echographic evaluation of splenic injury after blunt trauma. Radiology 1976;118:4115. [3] Tiling T, Boulion B, Schmid A. Ultrasound in blunt abdomino-thoracic trauma. In: Border, Allgoewer M, Hanson ST, editors. Blunt multiple trauma: comprehensive pathophysiology and care. New York: Marcel Decker; 1990. p. 41533. [4] Tso P, Rodriguez A, Cooper C, Militello P, Mirvis S, Badellino M, et al. Sonography in blunt abdominal trauma: a preliminary progress report. J Trauma 1992;33(1):3943[discussion: 434]. [5] Bode PJ, Niezen RA, van Vugt AB, Schipper J. Abdominal ultrasound as a reliable indicator for conclusive laparotomy in blunt abdominal trauma. J Trauma 1993;34(1): 2731. [6] Liu M, Lee CH, PEng KF. Prospective comparison of diagnostic peritoneal lavage, computed tomographic scanning, and ultrasonography for the diagnosis of blunt abdominal trauma. J Trauma 1993;35(2):26770. [7] Branney SW, Wolfe RE, Moore EE, Albert NP, Heinig M, Mestek M, et al. Quantitative sensitivity of ultrasound in detecting free intraperitoneal uid. J Trauma 1995;39(2): 37580. [8] Chambers JA, Pilbrow WJ. Ultrasound in abdominal trauma: an alternative to peritoneal lavage. Arch Emerg Med 1988;5(1):2633.
597
[9] Jehle D, Guarino J, Karamanoukian H. Emergency department ultrasound in the evaluation of blunt abdominal trauma. Am J Emerg Med 1993;11(4):3426. [10] Rozycki GS, Ochsner MG, Schmidt JA, Frankel HL, Davis TP, Wang D, et al. A prospective study of surgeon-performed ultrasound as the primary adjuvant modality for injured patient assessment. J Trauma 1995;39(3):4928 [discussion: 498500]. [11] Scalea TM, Rodriguez A, Chiu WC, Brenneman FD, Fallon WF Jr, Kato K, et al. Focused Assessment with Sonography for Trauma (FAST): results from an international consensus conference. J Trauma 1999;46(3):46672. [12] Ma OJ, Kefer MP, Mateer JR, Thoma B. Evaluation of hemoperitoneum using a single- vs multiple-view ultrasonographic examination [comments]. Acad Emerg Med 1995;2(7):5816. [13] Ingeman JE, Plewa MC, Okasinski RE, King RW, Knotts FB. Emergency physician use of ultrasonography in blunt abdominal trauma. Acad Emerg Med 1996;3(10):9317. [14] Rozycki GS, Ochsner MG, Feliciano DV, Thomas B, Boulanger BR, Davis FE, et al. Early detection of hemoperitoneum by ultrasound examination of the right upper quadrant: a multicenter study. J Trauma 1998;45(5):87883. [15] Rose JS, Bair AE, Mandavia D, Kinser DJ. The UHP ultrasound protocol: a novel ultrasound approach to the empiric evaluation of the undierentiated hypotensive patient. Am J Emerg Med 2001;19(4):299302. [16] Plummer D, Brunett D. Emergency department echocardiography improves outcome in penetrating cardiac injury. Ann Emerg Med 1992;21(5):70912. [17] Rozycki GS, Feliciano DV, Ochsner MG, Knudson MM, Hoyt DB, Davis F, et al. The role of ultrasound in patients with possible penetrating cardiac wounds: a prospective multicenter study. J Trauma 1999;46(4):54351 [discussion: 5512]. [18] McGahan JP, Rose J, Coates TL, Wisner DH, Newberry P. Use of ultrasonography in the patient with acute abdominal trauma. J Ultrasound Med 1997;16(10):65362 [quiz: 6634]. [19] Yoshii H, Sato M, Yamamoto S, Motegi M, Okusawa S, Kitano M, et al. Usefulness and limitations of ultrasonography in the initial evaluation of blunt abdominal trauma. J Trauma 1998;45(1):4550 [discussion: 501]. [20] Cogbill TH, Moore EE, Jurkovich GJ, Morris JA, Mucha P Jr, Shackford SR, et al. Nonoperative management of blunt splenic trauma: a multicenter experience. J Trauma 1989;29(10):13127. [21] Bose SM, Mazumdar A, Gupta R, Giridhar M, Lal R, Praveen BV. Expectant management of haemoperitoneum. Injury 1999;30(4):26973. [22] Minarik L, Slim M, Rachlin S, Brudnicki A. Diagnostic imaging in the follow-up of nonoperative management of splenic trauma in children. Pediatr Surg Int 2002;18(56):42931. [23] Jacobs DG, Saran JL, Marx JA. Abdominal CT scanning for trauma: how low can we go? Injury 2000;31(5):33743. [24] Navarrete-Navarro P, Vazquez G, Bosch JM, Fernandez E, Rivera R, Carazo E. Computed tomography vs clinical and multidisciplinary procedures for early evaluation of severe abdomen and chest traumaa cost analysis approach. Intensive Care Med 1996; 22(3):20812. [25] Chiu WC, Cushing BM, Rodriguez A, Ho SM, Mirvis SE, Shanmuganathan K, et al. Abdominal injuries without hemoperitoneum: a potential limitation of focused abdominal sonography for trauma (FAST) [comments]. J Trauma 1997;42(4):61722 [discussion: 6235]. [26] Branney SW, Moore EE, Cantrill SV, Burch JM, Terry SJ. Ultrasound based key clinical pathway reduces the use of hospital resources for the evaluation of blunt abdominal trauma. J Trauma 1997;42(6):108690. [27] Rose JS, Levitt MA, Porter J, Hutson A, Greenholtz J, Nobay F, et al. Does the presence of ultrasound really aect computed tomographic scan use? A prospective randomized trial of ultrasound in trauma. J Trauma 2001;51(3):54550. [28] American College of Surgeons. Advanced trauma life support for physicians. Chicago: ACS; 1997.
598
[29] Rose J, Richards JR, Bair AE, Battistella F, McGahan JP, Kuppermann N. The ultrasound is positive, now what? Derivation of a clinical decision rule for predicting therapeutic laparotomy among adult patients with a positive trauma ultrasound [abstract]. Aca Emerg Med 2002;9:4634. [30] Goldberg GG. Evaluation of ascites by ultrasound. Radiology 1970;96(15):21721. [31] Paajanen H, Lahti P, Nordback I. Sensitivity of transabdominal ultrasonography in detection of intraperitoneal uid in humans. Eur Radiol 1999;9(7):14235. [32] Abrams BJ, Sukumvanich P, Seibel R, Moscati R, Jehle D. Ultrasound for the detection of intraperitoneal uid: the role of Trendelenburg positioning. Am J Emerg Med 1999;17(2): 11720. [33] Gracias VH, Frankel HL, Gupta R, Malcynski J, Gandhi R, Collazzo L, et al. Dening the learning curve for the Focused Abdominal Sonogram for Trauma (FAST) examination: implications for credentialing. Am Surg 2001;67(4):3648. [34] Huang MS, Liu M, Wu JK, Shih HC, Ko TJ, Lee CH. Ultrasonography for the evaluation of hemoperitoneum during resuscitation: a simple scoring system. J Trauma 1994;36(2):1737. [35] McKenney KL, McKenney MG, Cohn SM, Compton R, Nunez DB, Dolich M, et al. Hemoperitoneum score helps determine need for therapeutic laparotomy. J Trauma 2001; 50(4):6504 [discussion: 6546]. [36] Ong AW, McKenney MG, McKenney KA, Brown M, Namias N, MaCloud J, et al. Predicting the need for laparotomy in pediatric trauma patients on the basis of the ultrasound score. J Trauma 2003;54(3):5038. [37] American College of Surgeons. Ultrasound education program bluebook. Chicago: ACS; 2000. [38] American College of Emergency Physicians. Use of ultrasound imaging by emergency physicians. Ann Emerg Med 2001;38:46970. [39] Thomas B, Falcone RE, Vasquez D, Santanello S, Townsend M, Hockenberry S, et al. Ultrasound evaluation of blunt abdominal trauma: program implementation, initial experience, and learning curve. J Trauma 1997;42(3):3848 [discussion: 38890]. [40] Ma OJ, MJ, Ogata M. Prospective analysis of a rapid trauma ultrasound examination performed by emergency physicians. J Trauma 1995;18:87985. [41] Rozycki GS, Ballard RB, Feliciano DV, Schmidt JA, Pennington SD. Surgeon-performed ultrasound for the assessment of truncal injuries: lessons learned from 1540 patients. Ann Surg 1998;228(4):55767. [42] Dolich MO, McKenney MG, Varela JE, Compton RP, McKenney KL, Cohn SM. 2,576 ultrasounds for blunt abdominal trauma. J Trauma 2001;50(1):10812. [43] Gracias VH, Frankel H, Gupta R, Reilly PM, Gracias F, Klein W, et al. The role of positive examinations in training for the focused assessment sonogram in trauma (FAST) examination. Am Surg 2002;68(11):100811. [44] Sisley AC, Johnson SB, Erickson W, Fortune JB. Use of an Objective Structured Clinical Examination (OSCE) for the assessment of physician performance in the ultrasound evaluation of trauma. J Trauma 1999;47(4):62731. [45] Salen PN, Melanson SW, Heller MB. The focused abdominal sonography for trauma (FAST) examination: considerations and recommendations for training physicians in the use of a new clinical tool. Acad Emerg Med 2000;7(2):1628. [46] Smith RS, Kern SJ, Fry WR, Helmer SD. Institutional learning curve of surgeonperformed trauma ultrasound. Arch Surg 1998;133(5):5305 [discussion: 5356]. [47] Shackford SR, Rogers FB, Osler TM, Trabulsy ME, Clauss DW, Vane DW. Focused abdominal sonogram for trauma: the learning curve of nonradiologist clinicians in detecting hemoperitoneum. J Trauma 1999;46(4):55362 [discussion: 5624]. [48] Richards JR, McGahan JP, Pali MJ, Bohnen PA. Sonographic detection of blunt hepatic trauma: hemoperitoneum and parenchymal patterns of injury. J Trauma 1999;47(6): 10927.
599
[49] Richards JR, McGahan JP, Jones CD, Zhan S, Gerscovich EO. Ultrasound detection of blunt splenic injury. Injury 2001;32(2):95103. [50] Brown MA, Casola G, Sirlin CB, Hoyt DB. Importance of evaluating organ parenchyma during screening abdominal ultrasonography after blunt trauma. J Ultrasound Med 2001; 20(6):57783[quiz: 585]. [51] Krupnick AS, Teitelbaum DH, Geiger JD, Strouse PJ, Cox CS, Blane CE, et al. Use of abdominal ultrasonography to assess pediatric splenic trauma. Potential pitfalls in the diagnosis. Ann Surg 1997;225(4):40814. [52] Ohta S, Hagiwara A, Yukioka T. Hyperechoic appearance of hepatic parenchyma on ultrasound examination of patients with blunt hepatic injury. J Trauma 1998;44:1358. [53] Brown MA, Casola G, Sirlin CB, Patel NY, Hoyt DB. Blunt abdominal trauma: screening us in 2,693 patients. Radiology 2001;218(2):3528. [54] Katz S, Lazar L, Rathaus V, Erez I. Can ultrasonography replace computed tomography in the initial assessment of children with blunt abdominal trauma? J Pediatr Surg 1996; 31(5):64951. [55] Thourani VH, Pettitt BJ, Schmidt JA, Cooper WA, Rozycki GS. Validation of surgeonperformed emergency abdominal ultrasonography in pediatric trauma patients. J Pediatr Surg 1998;33(2):3228. [56] Holmes JF, Brant WE, Bond WF, Sokolove PE, Kuppermann N. Emergency department ultrasonography in the evaluation of hypotensive and normotensive children with blunt abdominal trauma. J Pediatr Surg 2001;36(7):96873. [57] Boulanger BR, Kearney PA, Brenneman FD, Tsuei B, Ochoa J. Utilization of FAST (Focused Assessment with Sonography for Trauma) in 1999: results of a survey of North American trauma centers. Am Surg 2000;66(11):104955. [58] Luks F, Lemiere A. Blunt abdominal trauma in children: the practical value of ultrasonography. J Trauma 1993;34(5):60710. [59] Chan SS. Emergency bedside ultrasound to detect pneumothorax. Acad Emerg Med 2003; 10(1):914. [60] Kirkpatrick AW, Simons RK, Brown R, Nicolaou S, Dulchavsky S. The hand-held FAST: experience with hand-held trauma sonography in a level-I urban trauma center. Injury 2002;33(4):3038. [61] Stengel D, Bauwens K, Sehouli J, Nantke J, Ekkernkamp A. Discriminatory power of 3.5 MHz convex and 7.5 MHz linear ultrasound probes for the imaging of traumatic splenic lesions: a feasibility study. J Trauma 2001;51(1):3743.
Bedside echocardiography in chest trauma

Diku P. Mandavia, MD, FACEP, FRCPCa,b,*, Anthony Joseph, MD, FACEMc,d
a
Los Angeles CountyUSC Medical Center, Room 1011, GH 1200 North State Street, Los Angeles, CA 90033, USA b Keck School of Medicine, University of Southern California, Los Angeles, CA, USA c Department of Emergency Medicine, Royal North Shore Hospital, St. Leonards, Australia 2065 Australia d Department of Medicine, University of Sydney, Sydney, Australia
Bedside ultrasonography has been used in emergency medicine for more than 15 years [13]. Emergency physicians successfully have integrated ultrasound (US) as the primary imaging modality for a number of emergent disorders. Of the well-established indications for bedside US, it is clear that the trauma patient benets signicantly from the use of this technique [410]. Trauma patients often arrive in the middle of the night when radiology backup is at a minimum, yet these patients often require the most extensive imaging evaluation of any emergency patient. This dichotomous relationship has required that trauma clinicians be experts in imaging injured patients, which includes bedside US techniques. Many life-threatening disorders can be encountered in patients with chest trauma and, therefore, a rapid and thorough evaluation is compulsory. Chest radiography is essential and can detect many injuries including hemothorax, pneumothorax, diaphragmatic rupture, and rib fractures. Other specic ndings can lead the clinician to suspect other disorders such as traumatic aortic injury (TAI). Helical CT has demonstrated excellent sensitivity and specicity in diagnosing this entity [11,12] and other serious chest injuries. Despite these advancements, however, some injuries can remain undetected. In particular, patients with penetrating chest injuries may harbor serious cardiac injury and a pericardial eusion that may be clinically dicult to determine. Becks triad is unreliable, and previous work has shown that chest radiography and electrocardiography are not helpful in the diagnosis of
* Corresponding author. E-mail address: mandavia@usc.edu (D.P. Mandavia). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.004
602
D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619
a pericardial eusion and acute tamponade [13,14]. In this scenario, bedside echocardiography can provide immediate, accurate information regarding the pericardium and the need for immediate surgery. It also can improve patient outcome. Plummer et al [15] demonstrated that early detection of a pericardial eusion by echocardiography improves outcome in patients with penetrating chest trauma. He conducted a 10-year retrospective review of penetrating cardiac injury and compared patients who received an emergency echocardiogram with those who did not. In this study, the echocardiography group had a shorter time to diagnosis (15.5 versus 42.4 minutes) and a better overall survival rate (100% versus 57.1%). In the past, bedside echocardiography has been performed solely by cardiologists and cardiac sonographers. Because a formal echocardiography service often is not provided 24 hours per day at most hospitals, state-of-theart trauma care has required that trauma clinicians play an increasing role in performing this technique [16]. Over the last decade, it has become increasingly apparent that trauma clinicians can be trained in this technique in a focused fashion that is benecial to patients. The main objective of a focused echocardiogram is the primary nding of a pericardial eusion, and clinicians can be trained eectively in this modality after a relatively short period of time [17]. In a prospective study of 515 emergency patients, Mandavia et al [18] showed that emergency physicians reliably could examine the pericardium for a pericardial eusion with an accuracy of 97.5%. In this study, a combination of approaches including the parasternal long axis, subcostal four-chamber view, and the apical four-chamber view were used to detect 103 pericardial eusions. The Focused Assessment with Sonography for Trauma (FAST) examination includes the subcostal fourchamber view of the heart as part of the standard protocol and has been validated extensively [19,20]. Recognizing this trend and the evolving standard in trauma care, the Committee on Trauma of the American College of Surgeons revised their Advanced Trauma Life Support course to integrate rapid US assessment [21]. They recommend bedside US in the diagnosis of pericardial eusions when equipment and trained personnel are available. Course participants currently are taught the basics of the FAST examination, although formal hands-on US training has not yet been integrated.
Penetrating cardiac injury Penetrating cardiac injuries are among the most lethal of all injuries and can present with complete hemodynamic stability, cardiovascular collapse with shock, or frank cardiac arrest. This extreme variation in presentation and propensity for sudden deterioration requires that a thorough evaluation be performed to rule out this injury [22]. Lacerations to the heart from small stab wounds to the chest tend to seal by clot and adjacent fat, causing
603
a natural predisposition to cardiac tamponade. In contrast, gunshot wounds to the heart cause extensive myocardial trauma and leave large rents in the pericardium where continued hemorrhage into a hemithorax often occurs [23]. Becks triad often is referenced in the clinical diagnosis of acute cardiac tamponade but, clinically, it may be found in only 10% of cases [24]. Distended neck veins are dicult to evaluate due to cervical collars or may not be present because of concomitant hemorrhagic shock. Hypotension is not specic and may be caused by acute blood loss, tension pneumothorax, or spinal cord injury. Finally, mued heart sounds are not reliably auscultated during a busy trauma resuscitation and also may be heard in patients with hemothorax or pneumothorax. Chest radiography adds little in the diagnosis of acute cardiac injury. Chronic medical eusions cause a slow distension of the pericardial sac and an enlargement of the cardiac silhouette, but this nding often is not seen in acute distension as it is in trauma. Intrapericardial air is diagnostic of pericardial violation but rarely is seen on plain radiography. Electrocardiography often does not show any reliable diagnostic ndings in patients with pericardial eusions. Diagnostic pericardiocentesis often can lead to false-negative results because of the propensity of blood to clot within the pericardium and, thus, also cannot be relied on to exclude an eusion [24]. Acute cardiac tamponade in trauma occurs due to the brous anatomy of the pericardium that makes the pericardium inelastic and noncompliant compared with other body tissues. Sudden blood loss makes the right ventricle smaller, more compliant, and more susceptible to acute collapse. Simultaneously, blood accumulation in the constrictive pericardium causes an acute rise in intrapericardial pressure, subjecting the heart to right ventricular collapse and subsequent acute tamponade physiology. With small cardiac wounds, a slow rate of hemorrhage is much better tolerated, and acute tamponade may be delayed [22]. Small penetrating cardiac wounds may have a relatively slow rate of hemorrhage that can delay the appearance of acute cardiac tamponade. Acute tamponade is thought to be protective initially and then deleterious subsequently in penetrating heart injury. The initial tamponade eect helps limit the degree of exsanguination into the hemithorax, but this is a delicate balance because enlarging amounts of pericardial blood lead to a lifethreatening rise in intrapericardial pressure. After the intrapericardial pressure exceeds the right ventricular diastolic pressure, acute tamponade ensues and rapidly leads to death unless the pericardium is decompressed [22]. In penetrating heart injury, early detection of acute cardiac tamponade, therefore, is imperative. Before the widespread implementation of echocardiography, subxiphoid pericardial window (SPW) was the initial method of evaluation in penetrating chest injury with wounds in proximity to the heart. SPW remains the gold standard for diagnosing heart injury but is an invasive procedure and requires general anesthesia. At most US trauma
604
centers, SPW currently is used when echocardiography is equivocal or in unstable, penetrating thoracoabdominal trauma patients who require emergent transfer to the operating room [22,25]. Bedside echocardiography allows a rapid and noninvasive method of examining the heart, pericardium, and pleural spaces and can be performed in the resuscitation suite after the primary survey has been completed. A number of studies have evaluated US in the setting of penetrating chest trauma. Freshman et al [26] conducted a study of 36 patients with penetrating precordial trauma. They detected four pericardial eusions using echocardiography and concluded that echocardiography was more ecient than ICU observation in making the diagnosis and less invasive than SPW or thoracotomy. Aaland et al [27] evaluated bedside echocardiography in 53 patients with penetrating chest injury and concluded that echocardiography was a sensitive method of evaluating these patients. Jimenez et al [28] conducted a prospective trial that compared SPW with echocardiography and likewise concluded that echocardiography could eliminate SPW in many patients. In both of these studies, echocardiograms were performed by cardiology fellows. Nagy et al [29] evaluated echocardiography performed by trauma surgeons for detecting occult penetrating heart injury in 121 clinically stable patients. Using their selective pericardial window protocol (SPW performed after a positive echocardiogram), all signicant cardiac injuries were identied and successfully repaired. Meyer et al [30] prospectively evaluated 105 hemodynamically stable patients with penetrating chest injury for occult heart injury. In this study, all patients underwent SPW and bedside echocardiography performed by surgeons. Echocardiography missed four signicant injuries in patients with hemothorax but had a 100% sensitivity and a 91% specicity in patients without hemothorax. The investigators concluded that echocardiography was an acceptable diagnostic option in patients without hemothorax. Rozycki et al [31] conducted a study of 247 penetrating truncal injury patients in whom there was no immediate need for operative intervention. There were 236 true negative patients and 10 true positive patients. Echocardiography had a sensitivity and a specicity of 100%, with a mean evaluation time of less than 1 minute. Mean time from US to operation was 12.1 minutes and all patients survived. The investigators concluded that surgeon-performed US was both rapid and accurate in the diagnosis of hemopericardium. Rozycki et al [32] also completed a prospective multicenter trial of US in patients with possible penetrating heart injury. In this large study, four centers recruited 261 patients with penetrating truncal wounds. Bedside echocardiography was performed by surgeons, cardiologists, or cardiac sonographers and had a sensitivity of 100% and an accuracy of 97.3%. Mean time from US to operation was 12.1 minutes. The investigators concluded that US should be the initial modality for the evaluation of patients with penetrating precordial wounds.
605
Technical considerations The techniques of the FAST examination and transthoracic echocardiography are covered extensively elsewhere in this issue. In the emergencydepartment assessment of the patient with chest trauma, the primary focus is the evaluation of the pericardium and the detection of a pericardial eusion. It is important to understand that hemopericardium may not be completely anechoic as seen in most medical eusions. The presence of clot within the pericardium will cause the eusion to have internal echoes and a more hypoechoic appearance (Figs. 13) [33]. Echocardiography also can be used to perform a US-guided pericardiocentesis in a safe and expeditious manner [34,35]. Echocardiographic ndings of acute tamponade, such as diastolic collapse of the right ventricle, will precede clinically apparent cardiovascular collapse and may be detected by bedside US. The nding of an empty pericardium may represent a large pericardial tear that prevents the accumulation of a pericardial eusion. In these cases, hemorrhage occurs into the hemithorax, and thus, a normal pericardial examination in the setting of a hemothorax should be interpreted with caution [30]. Similarly, an initial small pericardial hemorrhage may drain into the chest cavity after insertion of an intercostal catheter and result in an initially normal pericardial US examination. The subcostal four-chamber view is a more commonly taught view in trauma, but the parasternal long-axis transthoracic approach oers some signicant advantages [36]. The parasternal long-axis view is a direct, clear approach through the precordium, with the heart examined through its long axis [37]. The pericardium is dependent on this view, allowing easier
Fig. 1. Parasternal long-axis view of a pericardial eusion from a stab wound to the heart. Upper arrow points to hypoechoic eusion and lower arrow points to the hyperechoic posterior pericardium.
606
Fig. 2. Subcostal four-chamber view of hemopericardium with clotted blood. Right arrow points to the clotted blood seen as a circumferential hypoechoic area surrounding the heart. Left arrow points to a rim of anechoic and nonclotted blood. LV, left ventricle.
detection of small eusions. In this plane, the posterior pericardium and the descending aorta are recognized easily, allowing a dierentiation between pericardial and pleural uid [38]. On the parasternal long-axis view, the pericardial reection will be noted to be anterior to the descending aorta,
Fig. 3. Subcostal four-chamber plane of a stab wound to the heart with acute hemopericardium. Note the clotted hypoechoic appearance of blood within the pericardium (both arrows). LV, left ventricle; RV, right ventricle.
607
whereas the pleural reection will be posterior to the aorta (Fig. 4). The parasternal view can be especially important given that many patients may have pericardial and pleural uid collections. The subcostal approach also requires a cooperative patient because signicant transducer pressure may be required in the subxiphoid area. Obese patients, who often have a small subxiphoid space, also pose a signicant limitation to the subcostal approach. Given these considerations, the authors believe the parasternal long-axis approach to be a superior view, although knowledge of both techniques is optimal for an unselected trauma population.
Blunt cardiac injury Myocardial rupture Blunt cardiac rupture rarely is seen in the emergency department because most of these injuries are rapidly fatal, with death ensuing at the scene [39]. Increasing sophistication of emergency medical systems and rapid transport of patients to hospitals has allowed a small number of these patients to arrive at the hospital alive. High-speed motor vehicle crashes, pedestrian accidents, or a fall from a height account for most of these injuries, and serious coexistent trauma is common in these cases [40]. The ability of a patient to survive such a lethal injury partially depends on the integrity of the pericardium [23]. Most survivors of blunt myocardial rupture have had rapid exsanguination prevented by an intact pericardium. These unstable patients may survive for variable periods with hemopericardium and cardiac tamponade.
Fig. 4. Parasternal long-axis view of a pleuropericardial eusion. Fluid is noted within the pericardium (small arrow and PE), and the descending aorta (Ao) and pleural uid is noted below this. Arrow A points to the more anterior reection of the pericardium adjacent to the descending aorta. Arrow B points to the posterior reection of the pleura adjacent to the descending aorta. LV, left ventricle; PE, pericardial eusion; RV, right ventricle.
608
Shock unresponsive to uid and blood administration, massive hemothorax, elevated central venous pressure, and cyanosis of the head, neck, arms, and upper chest in the setting of acute chest trauma should prompt a search for myocardial rupture [41]. Bruit de moulin is a precordial murmur that sounds like a splashing mill wheel and also may be present [25]. Chest radiography usually will not help make the diagnosis but may detect other coexistent trauma. Electrocardiography may show a bundle branch block, axis deviation, or nonspecic ndings. Echocardiography can be invaluable in making the diagnosis, as illustrated by published cases [4247]. Because of the infrequent nature of this injury, no large prospective trial has examined echocardiography specically in the setting of myocardial rupture. Patients with an intact pericardium may show evidence of hemopericardium on a limited echocardiographic examination, whereas formal echocardiography may reveal further ndings, including wall motion and valvular abnormalities. Myocardial contusion There is uncertainty as to the clinical signicance of a myocardial contusion. The diagnosis is based on the microscopic nding of intramural hematoma in the absence of damage to the other structures such as valves, papillary muscles, coronary vessels, and pericardium [48]. Establishing the diagnosis likewise has been problematic, with the only true gold standard being direct visualization at thoracotomy or examination at autopsy. Furthermore, signicant adverse cardiac events are rare in young patients, and a growing consensus of trauma experts favors the elimination of this diagnostic entity [49]. Echocardiographic ndings in patients with myocardial contusion include small pericardial eusions, wall motion abnormalities (especially of the anterior wall of the right ventricle), wall thickness changes, and ventricular dilatation. The role of formal echocardiography performed by cardiologists has been evaluated as a screening test in patients with blunt chest injury. Karalis et al [50] prospectively examined the utility of echocardiography in 105 blunt chest trauma patients and concluded that only patients with acute complications benet from echocardiography. Nagy et al [51] prospectively evaluated 315 patients with severe blunt chest trauma and completed extensive testing including electrocardiography, CPK-MB enzymes, and echocardiography. Overall, 22 patients were diagnosed with blunt cardiac injury, which was dened as evolving ST changes on electrocradiogram, dysrhythmias, elevated CPK-MB index, or hemodynamic instability. The investigators concluded that patients with blunt chest trauma who have a normal electrocradiogram, are hemodynamically stable, and have no dysrhythmias require no further testing. They also concluded that echocardiography is not useful as a screening investigation. Lindstaedt et al [52] performed a prospective study of 118 blunt
609
trauma patients. Their criteria for the diagnosis of myocardial contusion included echocardiographic evidence of akinetic wall motion or regional wall motion abnormality plus CK-MB enzyme elevation or electrocradiogram abnormality. They found no complications in 14 of 118 patients in whom they made the diagnosis, except for one case of left ventricular thrombus in the group followed for over 12 months. The investigators concluded that routine cardiac workup is not indicated and investigations should be limited to patients with complications. From the available literature, therefore, it appears that echocardiography (transthoracic or transesophageal) should be limited to patients with complications such as electrocradiogram changes, cardiac enzyme abnormalities, dysrhythmias, or hemodynamic instability. Unless the patient is unstable, echocardiography does not need to be performed in the ED. In addition, the use of bedside US by emergency physicians for detecting myocardial contusion has not been studied and currently is not recommended. Valvular injury Injury to cardiac valves after blunt or penetrating trauma is rare and usually coexists with other clinically signicant cardiac injuries [53]. Autopsy ndings have shown the aortic valve to be more frequently involved, followed by the mitral and tricuspid valves. Transthoracic and transesophageal echocardiography (TEE) are useful and able to detect right and left ventricular hypokinesis and acute aortic, mitral, and tricuspid incompetence. Invariably, these patients may have other thoracic injuries including pneumothoraces, which may limit the acoustic windows available for transthoracic echocardiography. The appropriate timing of surgical repair is dicult in these patients with multiple injuries, so TEE has been recommended as the investigation of choice because it allows accurate denition of the nature of the valve lesion and the underlying cardiac function [54].
Traumatic hemothorax Early diagnosis of massive hemothorax (more than 1500 mL of blood in the thoracic cavity) is important to ensure prompt and appropriate resuscitation by placement of a chest tube and restoration of blood volume. The thoracic cavity is a common source of bleeding, although only 5% to 15% of these injuries require urgent surgery. Clinical signs may be dicult to elicit in the resuscitation room and a chest radiograph usually involves some delay. The FAST examination incorporates views of both lung bases and will conrm a clinical suspicion of hemothorax, allowing early intervention before radiographic identication. The FAST examination includes views of the right and left upper quadrants. The right upper quadrant, when viewed from the eleventh rib
610
interspace in the midaxillary line, will include the liver parenchyma, the right diaphragm, and the pleural space superior to this. Any pleural eusion will be visible above the diaphragm as an anechoic space, and this may be the rst clue to a traumatic hemothorax (Figs. 5 and 6). A closer examination of the pleural space may be made at higher interspaces to conrm the superior extension of the pleural uid. During the examination of the left upper quadrant, by placing the probe longitudinally in the posterior axillary line in the tenth interspace, the left pleural space also may be inspected and will reveal an eusion or hemothorax if present. There is some evidence that experienced sonographers are able to diagnose accurately traumatic hemothoraces. Sisley et al [55] found that surgeon-sonographers could accurately detect traumatic eusions in blunt and penetrating trauma patients compared with chest radiography. In a study of 360 patients with 40 eusions, focused thoracic US detected 39 eusions, whereas chest radiography detected 37, with a slightly higher sensitivity for US (97.5%) compared with chest radiography (92.5%) and a comparable specicity (99.7% for each). More important, the investigators also noted that the time to perform the US was faster compared with the time to obtain a chest radiography (1.3 versus 14.2 minutes). US also was able to detect even small hemothoraces because the uid detected ranged from 60 mL to 1100 mL. Ma and Mateer [56] compared US to chest radiography in a study of 240 trauma patients. In their study, US had a sensitivity of 96.2% and a specicity of 100%. Rozycki et al [57] also conducted a study of 47 patients and 140 thoracic US examinations in critical care patients. They demonstrated a sensitivity of 83.6%, a specicity of 100%, and an accuracy of 94% in detecting pleural uid and concluded
Fig. 5. Hemothorax seen above the right diaphragm as an anechoic area (left arrow). The diaphragm is recognized as a hyperechoic line just superior to the liver (right arrow). L, liver.
611
Fig. 6. Hemothorax seen above the right diaphragm secondary to a stab wound. Left arrow points to anechoic nonclotted blood; right arrow points to echoic clotted blood.
that US can reliably detect pulmonary eusions in critically ill patients. Abboud and Kendall [58] recently published a prospective study of 155 patients using CT as the comparative standard. In their study, US had a low sensitivity of 12.5% and a specicity of 98.4%. Their interpretation of these contrasting results was that CT detects very small pulmonary eusions, many of which may not be clinically signicant in the acute resuscitation, and they concluded that further work is needed to clarify the size of hemothoraces detected by US.
Traumatic aortic injury The aorta is the most frequently injured of the major intrathoracic vessels in blunt and penetrating trauma. Penetrating trauma usually causes a fullthickness injury that is rapidly fatal, whereas a blunt mechanism such as sudden deceleration may result in partial or complete tears. The most common site for TAI after blunt trauma is the descending aorta just proximal to the ligamentum arteriosum, followed by the ascending aorta or arch (Fig. 7). Most patients who arrive in the ED have partial-thickness injuries in which the intima and media are torn but the adventitia remains intact. Of these patients, 50% will rupture in the rst 24 hours and 80% in the rst week if the diagnosis is not made [48]. Chest radiography may show a widened mediastinum and other specic ndings or it may be normal. Conventional angiography has been the gold standard investigation for diagnosis of this condition but has been challenged seriously by helical CT and TEE.
612
Fig. 7. (A) Long and short axis of descending aorta seen on TEE in a young adult with blunt chest trauma. Arrows point to intimal injury of the aorta. (B) Aortography of previous case conrms traumatic injury of the aorta (arrow).
TEE has been used in the operating room since the 1980s, with the introduction of two-dimensional transducers. The 5.0- to 7.5-MHz probes allow higher resolution, and the close proximity to the aortic and cardiac structures results in higher-quality images with better spatial resolution [59]. Doppler helped with the investigation of blood ow and assessment of valvular function. The biplane TEE in the 1990s provided a longitudinal plane that allowed imaging of vertically aligned structures such as the superior vena cava, right ventricular outow tract, and the ascending aorta. The multiplane transducer allowed even greater visualization of the cardiac structures. The advantages of TEE over aortography include the ability of TEE to be performed at the bedside with a short diagnostic time of 15 to 30
613
minutes and to be performed intraoperatively in patients undergoing lifesaving operations for other injuries. The use of TEE requires special training and is operator dependent. It also is relatively contraindicated in patients with cervical injury and absolutely contraindicated in those with esophageal injuries. TEE provides limited views of injury in the proximal ascending aorta and main branches and may miss injuries such as a partial rupture of the subclavian artery. TEE is ideal for use in the hemodynamically unstable patient who is unsuitable for transfer to the radiology suite for chest CT or arch angiography. Although it can be used at the bedside and in the potentially unstable patient, denitive airway management may be required before performing TEE. It is being increasingly used as the diagnostic test of choice for TAI in the unstable patient. It also can be used to diagnose other conditions such as acute valvular rupture and pericardial eusion or tamponade. The abnormalities found by TEE in conrmed cases of TAI are shown in Box 1 [60,61]. Smith et al [62] conrmed the accuracy of TEE in the diagnosis of TAI compared with surgery, aortography, and autopsy. In this study, cardiologists performed TEE successfully in 93 of 101 patients and detected TAI near the aortic isthmus in 11 of 93 patients. The ndings were conrmed in 10 of the 11 patients by surgery, aortography, and autopsy, resulting in a sensitivity of 100% and a specicity of 98%. There was one false-positive examination. The investigators concluded that TEE was sensitive and specic for the detection of TAI and can be used safely and accurately in critically injured patients. The ndings of this study were conrmed in a similar study by Vignon et al [63] in which TEE was performed on 31 consecutive trauma patients with a suspicion of TAI based on a major deceleration mechanism or chest radiography ndings of a mediastinum [8 cm. The investigators found two subsets of injury. Ten patients had subadventitial injury and 2 had tears conned to the intima only. There were 18 patients with a normal aorta conrmed on aortography and
Box 1. Transesophageal echocardiography ndings in traumatic aortic injury Dilated aortic isthmus Intimal ap with pseudoaneurysm Medial ap Intimal tear or mural thrombus Intramural hematoma Mediastinal hematoma (usually anteromedial)
614
1 patient with a false-negative study (a 2-mm medial tear found at autopsy). The patients with subadventitial tears were taken to the operating room and the patients with tears conned to the intima were treated conservatively. Smith et al [62] concluded that TEE should be the initial imaging modality for suspected TAI because of its portability, safety, accuracy, and benecial impact on patient care. Goarin et al [61] compared TEE with angiography and CT in a 9-year retrospective study. They evaluated 209 blunt trauma patients and demonstrated a 98% sensitivity and a 100% specicity of TEE for diagnosing TAI. They recognized that TEE also detects small injuries of unclear clinical signicance but concluded that TEE is an accurate method for the diagnosis of aortic injury. Buckmaster et al [64] conducted a prospective trial that compared TEE and aortography in 160 patients. TEE correctly identied all 14 injuries in this study with a 100% sensitivity. The investigators concluded that TEE is diagnostic and can obviate the need for aortography. They recommended aortography when TEE is equivocal, is contraindicated, is not tolerated, and when other vascular injuries that would be missed by TEE are suspected. TEE may be more sensitive in diagnosing small aortic intimal injuries compared with aortography or CT scan of the chest. A small study by Kepros et al [65] suggested that intimal tears \2 cm completely resolved within a few days. The study supported the safety of nonoperative management, which included beta-blockers, follow-up TEE, and invasive monitoring. The clinical relevance of these small intimal tears requires further study. There has been some uncertainty regarding the relative accuracy of helical CT and TEE in the detection of TAI. Vignon et al [60] prospectively compared the diagnostic accuracy of mutiplane TEE and helical CT in the detection of TAI and cardiac injury in patients who suered major chest trauma. Both modalities were reliable for the detection of subadventitial tears to the aortic isthmus or the ascending aorta. Helical CT detected one laceration of the innominate artery not seen on TEE, whereas TEE was more sensitive in diagnosing supercial (intima and media injuries) aortic injuries and blunt cardiac trauma such as acute valvular incompetence. Hence, these investigators recommended restricting the use of angiography to those with contraindications to CT or TEE, inconclusive results from other investigations, and suspected aortic branch injury. In summary, TEE should be used as a primary tool to diagnose TAI in patients who are too unstable for aortography or CT and who require emergent surgery for other life-threatening injuries. The main advantages are portability and rapidity; however, it appears to be user dependent like most other US applications. Presently, TEE is performed primarily by cardiologists and anesthesiologists and has not yet become part of the armamentarium of the emergency physician.
615
Traumatic pneumothorax There has been much recent interest in the extension of the FAST examination to include the thorax to detect the presence of an anterior pneumothorax that may not be clinically or radiologically obvious. Although these pneumothoraces are not immediately life threatening, it is essential to make the diagnosis, especially if the patient requires intubation for any reason (eg, head injury or imminent surgery) because positive-pressure ventilation may transform a simple pneumothorax into a tension pneumothorax. As US technology has improved rapidly over the last decade and the machines have become more compact, there is increased interest in the ability to diagnose a potentially life- threatening pneumothorax in situations such as air ight or space ight where radiology equipment is not practical. This technique was rst described in a veterinary journal when it was used to diagnose and treat a pneumothorax in a horse [66]. Since that time, there has been increased interest in the medical literature regarding the application of this procedure in humans. The technique involves the use of a 7.5-MHz linear probe applied over the anterior chest wall usually in the third or fourth interspaces in the midlateral clavicular line or the anterior axillary line. Normal lungtochest wall interface allows the detection of the lung sliding sign (Fig. 8). Lung sliding is the to-and-fro movement with respiration that is seen as a hyperechoic line between the chest wall and the aerated lung and represents the visceral and parietal pleura [67]. A normal examination also may show comet-tail (short-range reverberation) artifacts, which are high-amplitude echoes tapering and reducing in brightness with depth [68]. These artifacts arise from the visceral pleura,
Fig. 8. Sonographic appearance of a normal lung using a linear array 7.5-Mhz probe. The hyperechoic line represents the pleura (arrows). Absence of the lung-sliding sign (seen in real time) indicates a pneumothorax.
616
which is a highly reective surface. The presence of these artifacts indicates the absence of a pneumothorax. Conversely, when a pneumothorax is present, the air between the visceral and parietal pleura does not allow the sound waves to be transmitted to the visceral pleura and, therefore, comet tails are not produced and lung sliding does not occur. The apparently normal lung should be scanned rst to demonstrate lung sliding and any comet tails. Lung sliding may be absent in patients with pleural scarring and dicult to see in the presence of subcutaneous emphysema. Rowan et al [69] compared the accuracy of thoracic US with that of supine chest radiography in the detection of traumatic pneumothoraces, using chest CT as the reference standard. They prospectively studied 70 trauma victims who presented to the ED over an 8-month period. Patients in respiratory distress who required immediate tube thoracostomy were excluded. Twenty-seven patients underwent CT of the chest and 11 of them were found to have a pneumothorax. Thoracic US detected 11 of 11 pneumothoraces, whereas chest radiography detected only 4 of 11, yielding sensitivities of 100% and 36%, respectively. Thoracic US had one false positive study in a patient with substantial bullous emphysema on CT. Another study of 382 patients (mostly trauma) compared chest radiography to thoracic US for the detection of pneumothorax and demonstrated sensitivities of 100% and 95.5%, respectively [70]. There were 39 pneumothoraces seen on chest radiography, but only 37 on US, with two false negative studies due to the presence of subcutaneous emphysema. It would appear from the available evidence that US can be used to detect occult pneumothorax and may be incorporated into the FAST examination.
Summary Bedside US has an established role in the evaluation of chest trauma patients. Transthoracic echocardiography and TEE can be used to obtain critical information at the bedside for many emergent conditions, including the immediate detection of hemopericardium and acute aortic injury. More recent work has demonstrated that US also can be used to detect hemothoraces and pneumothoraces with accuracy. These diagnostic techniques can improve patient outcome and are within the scope of practice of emergency physicians and trauma surgeons. Physicians caring for trauma patients should be familiar with these techniques. References
[1] Heller M, Jehle D. Ultrasound in emergency medicine. Philadelphia: W.B. Saunders Company; 1995. [2] Heller MB, Verdile VP. Ultrasonography in emergency medicine. Emerg Med Clin N Am 1992;10(1):2746.
617
[3] Plummer D. Principles of emergency ultrasound and echocardiography. Ann Emerg Med 1989;18(12):12917. [4] Rozycki GS, Ballard RB, Feliciano DV, et al. Surgeon-performed ultrasound for the assessment of truncal injuries: lessons learned from 1540 patients. Ann Surg 1998;228(4): 55767. [5] Rozycki GS, Newman PG. Surgeon-performed ultrasound for the assessment of abdominal injuries. Adv Surg 1999;33:24359. [6] Rozycki GS, Ochsner MG, Schmidt JA, et al. A prospective study of surgeon-performed ultrasound as the primary adjuvant modality for injured patient assessment. J Trauma 1995;39(3):4928 [discussion: 498500]. [7] Ma OJ, Mateer JR, Ogata M, et al. Prospective analysis of a rapid trauma ultrasound examination performed by emergency physicians. J Trauma 1995;38(6):87985. [8] McKenney MG, Martin L, Lentz K, et al. 1,000 consecutive ultrasounds for blunt abdominal trauma. J Trauma 1996;40(4):60710 [discussion: 6112]. [9] Boulanger BR, McLellan BA, Brenneman FD, et al. Prospective evidence of the superiority of a sonography-based algorithm in the assessment of blunt abdominal injury. J Trauma 1999;47(4):6327. [10] Jehle D. Bedside ultrasonographic evaluation of hemoperitoneum: the time has come [editorial; comment]. Acad Emerg Med 1995;2(7):5756. [11] Scaglione M, Pinto A, Pinto F, et al. Role of contrast-enhanced helical CT in the evaluation of acute thoracic aortic injuries after blunt chest trauma. Eur Radiol 2001; 11(12):24448. [12] Parker MS, Matheson TL, Rao AV, et al. Making the transition: the role of helical CT in the evaluation of potentially acute thoracic aortic injuries. AJR Am J Roentgenol 2001; 176(5):126772. [13] Manyari DE, Milliken JA, Colwell BT, et al. Detection of pericardial eusion by chest roentgenography and electrocardiography versus echocardiography. Can Med Assoc J 1978;119(5):44550. [14] Eisenberg MJ, de Romeral LM, Heidenreich PA, et al. The diagnosis of pericardial eusion and cardiac tamponade by 12-lead ECG. A technology assessment. Chest 1996; 110(2):31824. [15] Plummer D, Brunette D, Asinger R, et al. Emergency department echocardiography improves outcome in penetrating cardiac injury. Ann Emerg Med 1992;21(6):70912. [16] Rozycki GS, Shackford SR. Ultrasound, what every trauma surgeon should know [comments]. J Trauma 1996;40(1):14. [17] Mandavia D, Aragona J, Chan L, et al. Ultrasound training for emergency physiciansa prospective study. Acad Emerg Med 2000;7:100814. [18] Mandavia DP, Honer RJ, Mahaney K, et al. Bedside echocardiography by emergency physicians. Ann Emerg Med 2001;38(4):37782. [19] Dolich MO, McKenney MG, Varela JE, et al. 2,576 ultrasounds for blunt abdominal trauma. J Trauma 2001;50(1):10812. [20] Bode PJ, Edwards MJ, Kruit MC, et al. Sonography in a clinical algorithm for early evaluation of 1671 patients with blunt abdominal trauma [comments]. AJR Am J Roentgenol 1999;172(4):90511. [21] Advanced Trauma Life Support Program for Doctors. Chicago: American College of Surgeons; 1997. [22] Asensio JA, Soto SN, Forno W, et al. Penetrating cardiac injuries: a complex challenge. Injury 2001;32(7):53343. [23] Eckstein M, Henderson S, Thorax Markovchick V. In: Marx J, Hockberger R, Walls R, editors. Rosens emergency medicine: concepts and clinical practice. St. Louis (MO): Mosby; 2002. p. 381414. [24] Demetriades D, van der Veen BW. Penetrating injuries of the heart: experience over two years in South Africa. J Trauma 1983;23(12):103441.
618
[25] Ivatury R. Injury to the heart. In: Feliciano D, Moore E, Mattox K, editors. Trauma. Englewood Clis (NJ): Appleton & Lange; 1996. p. 40921. [26] Freshman SP, Wisner DH, Weber CJ. 2-D echocardiography: emergent use in the evaluation of penetrating precordial trauma [comments]. J Trauma 1991;31(7):9025 [discussion: 9056]. [27] Aaland MO, Bryan FC III, Sherman R. Two-dimensional echocardiogram in hemodynamically stable victims of penetrating precordial trauma. Am Surg 1994;60(6): 4125. [28] Jimenez E, Martin M, Krukenkamp I, et al. Subxiphoid pericardiotomy versus echocardiography: a prospective evaluation of the diagnosis of occult penetrating cardiac injury. Surgery 1990;108(4):6769 [discussion: 67980]. [29] Nagy KK, Lohmann C, Kim DO, et al. Role of echocardiography in the diagnosis of occult penetrating cardiac injury [comments]. J Trauma 1995;38(6):85962. [30] Meyer DM, Jessen ME, Grayburn PA. Use of echocardiography to detect occult cardiac injury after penetrating thoracic trauma: a prospective study. J Trauma 1995;39(5):9027 [discussion: 9079]. [31] Rozycki GS, Feliciano DV, Schmidt JA, et al. The role of surgeon-performed ultrasound in patients with possible cardiac wounds. Ann Surg 1996;223(6):73744 [discussion: 7446]. [32] Rozycki GS, Feliciano DV, Ochsner MG, et al. The role of ultrasound in patients with possible penetrating cardiac wounds: a prospective multicenter study. J Trauma 1999; 46(4):54351 [discussion: 5512]. [33] DCruz IA, Constantine A. Problems and pitfalls in the echocardiographic assessment of pericardial eusion. Echocardiography 1993;10(2):15166. [34] Salem K, Mulji A, Lonn E. Echocardiographically guided pericardiocentesisthe gold standard for the management of pericardial eusion and cardiac tamponade. Can J Cardiol 1999;15(11):12515. [35] Vayre F, Lardoux H, Pezzano M, et al. Subxiphoid pericardiocentesis guided by contrast two-dimensional echocardiography in cardiac tamponade: experience of 110 consecutive patients. Eur J Echocardiogr 2000;1(1):6671. [36] Carrillo EH, Guinn BJ, Ali AT, et al. Transthoracic ultrasonography is an alternative to subxyphoid ultrasonography for the diagnosis of hemopericardium in penetrating precordial trauma. Am J Surg 2000;179(1):346. [37] Feigenbaum H. The echocardiographic exam. In: Feigenbaum H, editor. Echocardiography. 5th edition. Philadelphia: Lea & Febiger; 1994. p. 68133. [38] Lewandowski BJ, Jaer NM, Winsberg F. Relationship between the pericardial and pleural spaces in cross- sectional imaging. J Clin Ultrasound 1981;9(6):2714. [39] Powell MA, Lucente FC. Diagnosis and treatment of blunt cardiac rupture. W V Med J 1997;93(2):647. [40] Brathwaite CE, Rodriguez A, Turney SZ, et al. Blunt traumatic cardiac rupture. A 5-year experience. Ann Surg 1990;212(6):7014. [41] Rogers FB, Leavitt BJ. Upper torso cyanosis: a marker for blunt cardiac rupture. Am J Emerg Med 1997;15(3):2756. [42] Baxa MD. Cardiac rupture secondary to blunt trauma: a rapidly diagnosable entity with two-dimensional echocardiography. Ann Emerg Med 1991;20(8):9024. [43] Carrillo EH, Schirmer TP, Sideman MJ, et al. Blunt hemopericardium detected by surgeon-performed sonography. J Trauma 2000;48(5):9714. [44] Indrani S, Raji V, Kalyani N, et al. Sonographic diagnosis of blunt trauma causing delayed hemopericardium and cardiac tamponade. J Ultrasound Med 1991;10(5):2913. [45] Fulton JO, Nel L, de Groot KM, et al. Blunt cardiac rupture. S Afr J Surg 1998;36(4):1325. [46] Schiavone WA, Ghumrawi BK, Catalano DR, et al. The use of echocardiography in the emergency management of nonpenetrating traumatic cardiac rupture. Ann Emerg Med 1991;20(11):124850.
619
[47] Symbas NP, Bongiorno PF, Symbas PN. Blunt cardiac rupture: the utility of emergency department ultrasound. Ann Thorac Surg 1999;67(5):12746. [48] Jones E, Chambers K, Haro L. Thoracic trauma: principles and practice of emergency stabilization, evaluation and managemen. Part 3: injuries to the heart and aorta. Emerg Med Rep 2001;22331. [49] Miller FB, Shumate CR, Richardson JD. Myocardial contusion. When can the diagnosis be eliminated? Arch Surg 1989;124(7):8057 [discussion: 8078]. [50] Karalis DG, Victor MF, Davis GA, et al. The role of echocardiography in blunt chest trauma: a transthoracic and transesophageal echocardiographic study. J Trauma 1994;36(1):538. [51] Nagy KK, Krosner SM, Roberts RR, et al. Determining which patients require evaluation for blunt cardiac injury following blunt chest trauma. World J Surg 2001;25(1):10811. [52] Lindstaedt M, Germing A, Lawo T, et al. Acute and long-term clinical signicance of myocardial contusion following blunt thoracic trauma: results of a prospective study. J Trauma 2002;52(3):47985. [53] Zakynthinos EG, Vassilakopoulos T, Routsi C, et al. Early- and late-onset atrioventricular valve rupture after blunt chest trauma: the usefulness of transesophageal echocardiography. J Trauma 2002;52(5):9906. [54] Wall MJ Jr, Soltero ER. Trauma to cardiac valves. Curr Opin Cardiol 2002;17(2):18892. [55] Sisley AC, Rozycki GS, Ballard RB, et al. Rapid detection of traumatic eusion using surgeon-performed ultrasonography. J Trauma 1998;44(2):2916 [discussion: 2967]. [56] Ma OJ, Mateer JR. Trauma ultrasound examination versus chest radiography in the detection of hemothorax. Ann Emerg Med 1997;29(3):3125 [discussion: 3156]. [57] Rozycki GS, Pennington SD, Feliciano DV. Surgeon-performed ultrasound in the critical care setting: its use as an extension of the physical examination to detect pleural eusion. J Trauma 2001;50(4):63642. [58] Abboud P, Kendall J. Emergency department ultrasound for hemothorax after blunt traumatic injury. J Emerg Med 2003;25(2):1814. [59] Leitman M, Peleg E, Rosenblat S, et al. Transesophageal echocardiographyan overview. Isr Med Assoc J 2001;3(3):198206. [60] Vignon P, Boncoeur MP, Francois B, et al. Comparison of multiplane transesophageal echocardiography and contrast-enhanced helical CT in the diagnosis of blunt traumatic cardiovascular injuries. Anesthesiology 2001;94(4):61522 [discussion: 615A]. [61] Goarin JP, Cluzel P, Gosgnach M, et al. Evaluation of transesophageal echocardiography for diagnosis of traumatic aortic injury. Anesthesiology 2000;93(6):13737. [62] Smith MD, Cassidy JM, Souther S, et al. Transesophageal echocardiography in the diagnosis of traumatic rupture of the aorta [comments]. N Engl J Med 1995;332(6):35662. [63] Vignon P, Gueret P, Vedrinne JM, et al. Role of transesophageal echocardiography in the diagnosis and management of traumatic aortic disruption. Circulation 1995;92(10):295968. [64] Buckmaster MJ, Kearney PA, Johnson SB, et al. Further experience with transesophageal echocardiography in the evaluation of thoracic aortic injury. J Trauma 1994;37(6):98995. [65] Kepros J, Angood P, Jae CC, et al. Aortic intimal injuries from blunt trauma: resolution prole in nonoperative management. J Trauma 2002;52(3):4758. [66] Ratanen N. Diagnostic ultrasound: diseases of the thorax. Vet Clin N Am 1986;2:4966. [67] Kirkpatrick AW, Ng AK, Dulchavsky SA, et al. Sonographic diagnosis of a pneumothorax inapparent on plain radiography: conrmation by computed tomography. J Trauma 2001; 50(4):7502. [68] Gent R. Artifacts in diagnostic ultrasound. In: Gent R, editor. Applied physics and technology of diagnostic ultrasound. South Australia: Milner Publishing; 1997. p. 177220. [69] Rowan KR, Kirkpatrick AW, Liu D, et al. Traumatic pneumothorax detection with thoracic US: correlation with chest radiography and CTinitial experience. Radiology 2002;225(1):2104. [70] Dulchavsky SA, Hamilton DR, Diebel LN, et al. Thoracic ultrasound diagnosis of pneumothorax. J Trauma 1999;47(5):9701.
Cardiac ultrasound
Teriggi J. Ciccone, MDa,*, Shamai A. Grossman, MDb
Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Harvard Aliated Emergency Medicine Residency Program, One Deaconess Road, West Campus Clinical Center 2, Boston, MA 02115, USA b The Cardiac Emergency Center and Clinical Decision Unit, Beth Israel Deaconess Medical Center, Harvard Medical School, One Deaconess Road, West Campus Clinical Center 2, Boston, MA 02115, USA
a
Since its introduction into medical practice in Europe in the 1950s, ultrasound has become an important diagnostic modality for physicians. During the late 1980s and early 1990s, the use of ultrasound in the emergency-department (ED) setting expanded dramatically. Ultrasound imaging has several properties that make it especially attractive to emergency physicians (EPs). Ultrasound oers a portable, noninvasive, and relatively rapid means of evaluating internal organs. Echocardiography the specic use of ultrasound for imaging the heart and great vessels has been employed since the 1960s in the United States. Specic advances in technology, including the advent of two-dimensional (2-D) and Doppler ow modes have made echocardiography one of the primary tools in the evaluation of cardiovascular pathology. As with other applications for ultrasound in the ED, echocardiography has evolved signicantly and now is standard practice in the management of various acute cardiovascular disease processes. There are two major types of echocardiography: transthoracic and transesophageal. Transthoracic or surface echocardiography uses a probe placed on the patients chest. Transesophageal echocardiography is an invasive procedure that requires specialized training. The ultrasound probe is mounted on a exible unit that is passed down the patients esophagus under sedation. Transthoracic echocardiography (more easily applied and, thus, used by the EP) is the focus of this review. ED echocardiography is commonly performed by two groups of providers: EPs and cardiologists. The role of EPs and cardiologists diers, however, because cardiologists undergo a far more rigorous training
* Corresponding author. E-mail address: tciccone@bidmc.harvard.edu (T.J. Ciccone). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.012
622
T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640
program in echocardiography [1]. Yet, unlike EPs, cardiologist may not be available promptly enough to evaluate emergency patients, many of whom are unstable and in need of rapid diagnostic assessment. It is important, therefore, to delineate the roles of both groups of providers and their abilities to accurately assess dierent entities of cardiovascular pathology. Mateer et al [2] created a model curriculum for training EPs in ultrasound. This curriculum consists of 40 hours of instruction and 150 ultrasound examinations, 20 to 50 of which involve the heart. There is clear evidence that EPs can assess certain diseases with echocardiography accurately (most notably, pericardial eusion and cardiac standstill during arrest). Other uses of echocardiographyspecically, its use in the evaluation of focal wall motion abnormalities in acute coronary syndromes, valvular heart disease, and transesophageal echocardiographypresently are beyond the scope of training for EPs. This article reviews both of these broad categories for the use of echocardiography on ED patients (Box 1). Modes of echocardiography The echocardiographic mode describes the type of image that returning sound waves produce. M-mode produces a nonanatomic, wave-like image that is useful in interpreting movement of cardiac structures. Although used frequently by the cardiologist in the assessment of cardiac wall and valve motion, it has limited utility for the EP. 2-D mode echocardiography uses returning sound waves to produce an anatomic representation of deep tissues. Specic structures and their movements can be viewed directly. This mode of echocardiography has the greatest application to emergency echocardiography and is the mode with which EPs are most familiar. Doppler mode echocardiography assesses the velocity and direction of ow within the heart and great vessels. It is especially applicable for the
Box 1. Uses of echocardiography for specic pathology Disease states that reliably are detected by emergency physicians Percardial effusion Cardiac standstill/asystole Disease states that generally require cardiologist interpretation Focal wall motion abnormalities Pulmonary embolism Endocarditis Aortic dissection/aneurysm Valvular disease
623
evaluation of stenotic and regurgitant valvular lesion. Color ow Doppler imaging, a specic variation of this mode, superimposes color-coded ow images onto a 2-D image.
Echocardiographic windows The term window is used to describe the anatomic image produced, depending on where the ultrasound probe is placed on the patients body. Common echocardiographic windows include the subcostal (or subxiphoid), the parasternal, the apical, and the supraclavicular windows (Fig. 1). The axis for a particular window can be changed by rotating the probe 90(. The subcostal and parasternal windows are most critical for EPs.
Fig. 1. Echocardiographic windows. 1, supraclavicular window; 2, left parasternal window; 3, apical window; 4, subcostal window; 5, right parasternal window; L, left; R, right. (Courtesy of C.A. Ciccone, Bedford, MA.)
624
The subcostal window (Figs. 2 and 3) is relatively easy to obtain and can provide the EP with useful data [3], particularly for pericardial eusions and cardiac standstill. The ultrasound probe is place in the patients epigastrum just below the xiphoid process of the sternum. The probe is oriented toward to patients left shoulder, with the notch on the ultrasound probe pointed toward the left side of the patient. The image produced with this window puts the patients right ventricle at the top of the image on the screen. The right atrium and left ventricle are lined up below the right ventricle. Below these, the left atrium is visible. For the left parasternal window (Figs. 46), the ultrasound probe is place just to the left of the patients sternum at a level between the fourth and sixth intercostal spaces. Ideal images are obtained with the patient in a left lateral decubitus position. For the long axis, the probe is oriented parallel to the heart, with the probe notch pointing toward the patients right shoulder. In this window and axis, the patients right ventricular outow tract is immediately supercial on the screen. Below this, the left ventricle, its outow tract, and the aortic valve are visible. The mitral valve and left atrium are the deepest structures viewed. Rotating the probe clockwise 90( on the patients chest provides the short-axis view (see Fig. 6), where the right and left ventricles appear next to each other. The four-chamber apical view (Fig. 7) is obtained by placing the probe at the point of maximal impulse on the chest. The probe is pointed toward the patients right shoulder. From this window, the ventricles are visible at the top of the screen, with the atria lined up below.
Fig. 2. Subcostal window. LV, left ventricle; RA, right atrium; RV, right ventricle. (Courtesy of W.J. Manning, MD, Boston, MA.)
625
Fig. 3. Schematic representation of structures viewed from the subcostal window. LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle. (Courtesy of C.A. Ciccone, Bedford, MA.)
Echocardiography in the emergency department Since the late 1980s, the use of echocardiography in the ED has gained wide acceptance [4]. One of the earliest works examining the use of echocardiography in the ED was a descriptive study by Mayron et al [5] in 1988. This study analyzed echocardiograms done by EPs over a 1-year
Fig. 4. Left parasternal window, long axis. Ao, aorta; LA, left atrium; LV, left ventricle. (Courtesy of W.J. Manning, MD, Boston, MA.)
626
Fig. 5. Schematic representation of left parasternal long axis window.; Ao, aorta; LA, left atrium; LV, left ventricle; MV, mitral valve; RV, right ventricle. (Courtesy of C.A. Ciccone, Bedford, MA.)
period. One hundred fty-six echocardiograms were performed on critical patients (43% trauma patients, 37% with nonperfusing cardiac rhythms, and 20% with unexplained hypotension). Seven examinations were considered positive. By early detection of diseases such as pericardial eusion and presence of ventricular motion despite a lack of central pulses, patient care was enhanced in all of the positive studies [5].
Fig. 6. Left parasternal short axis. LV, left ventricle; RV, right ventricle. (Courtesy of W.J. Manning, MD, Boston, MA.)
627
Fig. 7. Apical four-chamber window. LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle. (Courtesy of W.J. Manning, MD, Boston, MA.)
Levitt and Jan [6] examined the eects of 2-D echocardiography on medical decision making among EPs. One hundred emergency patients were enrolled prospectively in the study. Physicians were asked about their level of condence in diagnosis, treatment, and disposition before and after echocardiography was performed. Results showed that EPs were three times more condent in diagnosis, seven times more condent in treatment, and three times more condent in disposition after being provided with echocardiographic data. Additional clinical studies that examined the uses of echocardiography by EPs have demonstrated that for pericardial eusion and cardiac standstill, EPs can use echocardiography as a diagnostic tool reliably [711].
Pericardial eusion Pericardial eusions appear as dark regions of uid between the pericardium and myocardium (Figs. 8 and 9). Echocardiographic signs of pericardial tamponade, including the swinging heart and right ventricular collapse also can be found. Eusions of traumatic and medical etiologies can be detected by echocardiography. Echocardiography also can be therapeutically used to guide drainage of pericardial eusions by pericardiocentesis.
628
Fig. 8. Pericardial eusion. The eusion appears as a dark echo lucent region between the pericardium and the right ventricular wall. LV, left ventricle; PE, pericardial eusion; RV, right ventricle. (Courtesy of W.J. Manning, MD, Boston, MA.)
Fig. 9. Schematic representation of pericardial eusion. Ao, aorta; LA, left atrium; LV, left ventricle; PE, pericardial eusion. Arrow shows collapse of right ventricular wall due to increased intrapericardial pressure. (Courtesy of C.A. Ciccone, Bedford, MA.)
629
Echocardiographic examination for pericardial eusion has signicant applications in patients presenting with penetrating chest trauma. Jimenez et al [12] examined 73 consecutive patients with penetrating chest wounds and stable vital signs. The patients underwent beside echocardiography followed by surgical pericardial exploration. Results showed that echocardiography was 96% accurate, 90% sensitive, and 97% specic for detecting eusions compared with open pericardiotomy. Nagy et al [13] determined that 2-D echocardiography reduced the number of negative pericardial windows from 115 to 15 in a group of 121 stable patients with penetrating chest wounds. Plummer et al [7] examined a series of 49 cases of penetrating cardiac injuries. Twenty-eight of the 49 patients underwent immediate echocardiography. The measured outcomes included overall survival and time to diagnosis and surgical intervention. Results of this study revealed a 100% survival rate in the echocardiography group compared with a 57% survival rate in the nonechocardiography group. Echocardiography also decreased the average time to diagnosis and surgical disposition from 42 minutes in the nonechocardiography group to 15 minutes in patients undergoing echocardiography. Mandavia et al [8] studied the ability of EPs to detect pericardial eusions using echocardiography in 2001. Five hundred fteen patients at high risk for pericardial eusion due to medical and traumatic etiologies were enrolled, 103 of which had pericardial eusions. Echocardiograms performed by EPs were compared with a gold standard review by a cardiologist. Results of this study revealed that EPs detected pericardial eusions with an overall sensitivity of 96%, specicity of 98%, and accuracy of 97%. Ultrasound-guided pericardiocentesis has become the standard treatment for pericardial eusions of medical etiologies. Tsang et al [14] studied 1127 echocardiographically guided pericardiocenteses performed at the Mayo Clinic between 1979 and 2000. They found that an overall success rate of 97% was obtained89% on the rst attempt, with a complication rate of 4.7% (1.2% with major complications). This complication rate was signicantly lower than pericardiocenteses performed by other methods. Echocardiographic diagnosis of pericardial tamponade is a somewhat more controversial issue. The classic clinical ndings for tamponade, including mued heart sounds, jugular venous distension, and hypotension may be absent in cases in which echocardiographic signs of tamponade are evident. The most frequently used echocardiographic nding of tamponade is collapse of the right heart chambers during mid-to-late diastole. Levine et al [15] examined 50 medical patients with pericardial eusions and echocardiographic right heart chamber collapse (92% right atrial, 57% right ventricular collapse). All of these patients underwent right heart catheterization and measurement of intrapericardial pressures by pericardiocentesis. Pericardial pressures equaled right atrial pressures in 84% of patients, and overall pericardial pressure was elevated in all patients. These ndings far
630
exceeded the percentage of clinical signs of tamponade in these patients. Only 14% of these patients had systolic blood pressures less than 100 mm Hg, 36% had pulsus paradoxus, 24% had distant heart sounds, and 74% had jugular venous distension. In fact, the diagnosis of tamponade was not considered before echocardiography in 50% of these patients. Some research suggests that right heart chamber collapse is a poor predictor of pericardial tamponade. Merce et al [16] examined echocardiographic signs of tamponade against a gold standard of clinically diagnosed tamponade. One hundred ten patients with moderate-to-large pericardial eusions were studied. Their results found that although 90% of patients with clinical tamponade had collapse of at least one right heart chamber, 34% of patients without clinical tamponade had similar collapse of one right heart chamber. These nding indicate that right heart chamber collapse lacks specicity for pericardial tamponade. In an attempt to resolve this controversy regarding the use of right heart chamber collapse in the diagnosis of pericardial tamponade, many investigators contend that tamponade is not an all or nothing phenomenon [15,17]. Rather, there is a spectrum of pericardial disease, beginning with modest increases in pericardial pressures and ending with frank hemodynamic compromise. Whether or not patients with echocardiographic ndings of elevated intrapericardial pressures ultimately will go on to develop clinical tamponade is not clear; however, given the severity of pericardial tamponade, EPs should be aware of these ndings and consider them when appropriate. Cardiac arrest Lack of a palpable pulse may represent true asystole or another pathologic condition. This determination holds critical importance, especially when a potentially reversible etiology for pulselessness is found. Pericardial eusion, pulseless electrical activity, ventricular brillation, and ventricular tachycardia are potentially reversible etiologies that result in the absence of central pulses. Echocardiography has been used to demonstrate that many patients with pulseless electrical activity have some degree of myocardial and valvular motion [18]. Echocardiography also can dierentiate asystole from ne ventricular brillation [9]. The echocardiographers goal in pulseless arrest, therefore, is to determine whether cardiac activity is present at all or whether there is cardiac standstill. Presence or lack of cardiac activity in these situations has important prognostic implications in decisions to terminate resuscitative measures [10,11]. Blaivas and Fox [10] studied outcomes among 169 patients presenting to their ED with ongoing cardiopulmonary resuscitation. Within this group, 136 patients had cardiac standstill on initial echocardiographic assessment. Results of this study revealed that none of the patients without cardiac activity by echocardiogram survived to leave the ED, regardless of electric
631
cardiac activity. In patients presenting to the emergency room undergoing cardiopulmonary resuscitation, therefore, lack of cardiac activity on echocardiogram may provide useful information regarding withdrawal of further resuscitation. Salen et al [11] prospectively examined sonographic and capnographic data in 102 cardiac arrest patients. Among 41 patients with identiable cardiac activity on echocardiogram, 11 (27%) survived to hospital admission. Among the 61 patients with no cardiac activity, however, only 2 (3%) survived to hospital admission. Within the group of 11 patients with echocardiographic evidence of cardiac activity who survived, 8 patients were in pulseless electrical activity, 2 were in ventricular tachycardia, and 1 was in ventricular brillation. Undierentiated shock The hypotensive ED patient represents a particularly challenging diagnostic dilemma. In some cases, the etiology of shock is obvious; in other cases, the etiology is less clear. In cases of potential cardiogenic shock, echocardiography is a useful diagnostic adjunct. Kaul et al [19] used 2-D echocardiography in the assessment of 42 critically ill patients with systolic blood pressures less than 100 mm Hg or with pulmonary edema to determine cardiac versus noncardiac causes of shock. Echocardiography was performed within 2 hours of placement of a pulmonary artery catheter. Presence of normal versus abnormal left ventricular function on echocardiogram was compared with pulmonary artery catheter data. The investigators found that there was complete agreement between echocardiographic and pulmonary artery catheter data regarding the etiology of shock in 86% of the study patients. Moore et al [20] prospectively analyzed whether EPs with specialized echocardiographic training could assess left ventricular function. Four EPs completed 6 hours of didactics in echocardiography, 10 hours of observing and performing echocardiographic examinations under cardiologist supervision, and a pilot tape review of nonhypotensive patients to assess technique. After completion of this training, these physicians performed focused echocardiographic assessments of 51 systemically hypotensive patients to assess left ventricular function. Blinded cardiologists reviewed all studies. Results of this study demonstrated a Pearson correlation coecient of 0.86 between EPs and cardiologists. This value closely approximated the Pearson coecient of 0.84 between cardiologists. The investigators concluded that EPs with specialized training could accurately determine left ventricular function. These results must be interpreted carefully. The training received by these physicians far exceeds the training provided to most EPs [1,2]. Echocardiography also has been used in limited settings to determine cardiac output quantitatively [21]. Invasive monitoring devices such as
632
pulmonary artery and central venous catheters often are employed for these purposes. The use of rapid, noninvasive techniques for measuring hemodynamic parameters have important implications, especially for EPs. The left ventricular outow (LVOT) method correlates well with pulmonary artery catheter data, with a correlation coecient of 0.94 [21]. This method uses the measured aortic diameter during systole, the average ow velocity, and the heart rate (HR) to calculate cardiac output (CO): CO = HR (LVOT area) ow velocity. Future studies are required to better delineate the uses and limitations of echocardiography for these purposes. Acute coronary syndromes Decreased or absent perfusion to a specic region of myocardium results in focal wall motion abnormalities. Hauser et al [22] showed in 1985 that new or increased focal wall motion abnormalities occur almost immediately after coronary artery occlusion in patients undergoing percutaneous transluminal coronary angioplasty. Regions of left ventricular hypokinesis and akinesis occurred in the regions anatomically predicted, based on the coronary artery occluded. Since that time, further work demonstrating the use of echocardiography in the diagnosis of acute coronary syndromes has been done. Nearly all of these studies, however, involved echocardiography performed by cardiologists or by certied echocardiographic technologists. Currently, the use of echocardiography for the specic task of determining focal wall motion abnormalities remains above the scope of practice for EPs [23]. Peels et al [24] performed 2-D echocardiography on 43 ED patients with active chest pain to determine whether focal wall motion abnormalities were a meaning measure of myocardial ischemia and infarction. Echocardiographic ndings were compared with serum creatinine phosphokinase and creatinine phosphokinaseMB isoezyme levels and with results from coronary angiography studies performed within 3 weeks of presentation. Patients with previous history of coronary artery disease, myocardial infarction, or coronary artery bypass grafting were excluded. Twenty-six patients were found to have regional wall motion abnormalities during chest pain in the ED. Twelve of these patients subsequently were found to have had a myocardial infarction by serial cardiac enzymes. Among the remaining 14 patients with regional wall motion abnormalities, 11 were found to have coronary artery disease by coronary angiography. In the 17 patients without focal wall motion abnormalities on 2-D echocardiogram, 3 had coronary artery disease by angiography, and none were ruled in for myocardial infarction. Overall, 2-D echocardiography was found to be 88% sensitive and 78% specic for myocardial ischemia, with a positive predictive value of 85% and a negative predictive value of 82%. For myocardial infarction, 2-D echocardiography had a sensitivity of 92%, with a negative predictive value of 94%.
633
Sabia et al [25] studied the use of 2-D echocardiography on 180 ED patients with chest pain. Regional wall motion abnormalities were compared with admission rates, results of serial cardiac enzyme studies, and complication rates from coronary artery disease. Myocardial infarction occurred in 30 patients with positive cardiac enzyme studies. In this group, 27 patients were found to have regional wall motion abnormalities, 2 had no wall motion abnormalities, and 1 study was technically inadequate. Only 9 of these 30 patients had diagnostic electrocardiograms. The sensitivity for echocardiography among patients with acute myocardial infarction, therefore, was far superior to the sensitivity of electrocardiography. Among patients who had not had an acute myocardial infarction by serial cardiac enzymes, 60 had regional wall motion abnormalities. Myocardial infarction occurred in only 2 of the 180 patients without any regional wall motion abnormalities noted. Among the 13 patients with in-hospital complications from myocardial infarction (cardiogenic shock, arrhythmia, postinfarction angina), all were found to have regional wall motion abnormalities. Overall, 2-D echocardiography had a sensitivity of 93% and a negative predictive value of 98% for detecting acute myocardial infarction. Low specicity (60%) and positive predictive value (31%) indicate that the presence of focal wall motion abnormalities cannot identify myocardial infarction reliably. The results of Sabia et al [25] were supported by Krontos et al [26]. Two hundred sixty patients presenting to the ED with possible myocardial ischemia underwent 2-D echocardiography within 4 hours of presentation. Forty-ve patients developed cardiac events, dened as myocardial infarction, percutaneous transluminal angioplasty, or coronary artery bypass grafting. Forty-one of these patients had abnormal echocardiograms, dened by the presence of focal wall motion abnormalities or an overall ejection fraction of less than 40%. Of the 166 patients who had normal echocardiograms, only 4 had cardiac events. Sensitivity of an abnormal echocardiogram for cardiac events was 91%, with a negative predictive value of 98%. Buchsbaum et al [27] examined the utility of echocardiography in patients considered low-risk for coronary artery disease. This prospective cohort study enrolled 145 patients with no previous history of cardiac disease presenting to the ED with chest pain. After a normal electrocardiogram and one negative creatinine phosphokinase test, the patients underwent stress echocardiography. The patients were followed for 3 months to determine the occurrence of any cardiac events. One hundred thirty-eight patients had stress echocardiograms with no evidence of focal wall motion abnormalities. All of these patients were free of cardiac events over the next 3 months. It may be concluded, therefore, that when 2-D echocardiography is performed on patients with possible ischemic chest pain and when no focal wall motion abnormalities are found, myocardial infarct or ischemia is unlikely. The presence of focal wall motion abnormalities, however, does
634
not rule in coronary artery disease, and further diagnostic testing should be undertaken (Fig. 10).
Pulmonary embolism Pulmonary embolism (PE) remains a challenging diagnosis for EPs. Varied historical presentations and unreliable physical ndings are common in patients with PE. Current diagnostic tests used in the evaluation of PE, including serum D-dimer, CT pulmonary angiography, and ventilationperfusion scanning, have limitations [28]. Pulmonary angiography, considered the gold standard for the diagnosis of PE, is an invasive procedure that has the potential for serious complications. The use of echocardiography has been studied by a number of investigators in the diagnosis, prognosis, and treatment of PE. Echocardiographic signs of increased pulmonary arterial pressure seen in PE include right ventricular dysfunction or dilatation, paradoxical septal wall motion, tricuspid regurgitation, and the presence of clot in the right ventricle. Interpretation of echocardiographic signs of PE currently is considered above the scope of training for EPs. As with the recognition of focal wall motion abnormalities, this skill requires a signicant investment of time in instruction. Echocardiography is a poor diagnostic test for PE [29,30]. In a prospective observational study that examined the accuracy of 2-D echocardi-
ED patient with chest pain History, physical exam, electrocardiogram (EKG) EKG nondiagnostic EKG diagnostic for acute coronary syndrome
Perform echocardiogram Focal wall motion abnormalities during pain? No Low risk patient? Yes Consider Discharge No Admission and further workup Yes Intervention or further workup
Fig. 10. Echocardiographic management algorithm for patients with possible ischemic chest pain.
635
ography in the diagnosis of ED patients with possible PE, echocardiography missed 16 of 39 patients with PEs diagnosed by other modalities [29]. Overall sensitivity for echocardiography in detecting PE was only 41%. Ribeiro et al [31] examined the extent of right ventricular dysfunction determined by echocardiogram as a prognostic indicator of mortality in patients with PE. Echocardiography was performed on 126 patients with PE. These patients were stratied into two groups: those with little or no right ventricular dysfunction and those with moderate-to-severe right ventricular dysfunction. Patients in the latter group had a signicantly higher mortality, with a relative risk of 6.0. In addition, all in-hospital deaths from PE occurred in the group with moderate-to-severe right ventricular dysfunction. The investigators concluded that echocardiographic ndings of right ventricular dysfunction in patients with known PE were associated with mortality rate. Specic criteria for treatment of PE with thrombolytic therapy remain controversial. Although thrombolytic therapy for hemodynamically unstable patients with PE often is employed, patients with normal systemic arterial blood pressures and signs of right ventricular dysfunction on echocardiogram also may benet from thrombolysis [32,33]. Konstantinides et al [32] studied the use of thrombolysis in hemodynamically stable patients with submassive PEs. Two hundred fty-six patients with PE and echocardiographic signs of right ventricular dysfunction or pulmonary arterial hypertension were randomized to receive heparin with recombinant tissue plasminogen activator (rtPA) or heparin with placebo. Although no dierence in mortality was found between these two groups, the group receiving rtPA plus heparin showed less clinical deterioration (ie, requirement for vasopressor therapy, repeated thrombolysis, endotracheal intubation, cardiopulmonary resuscitation, surgical embolectomy, or thrombus fragmentation by catheter). Goldhaber et al [33] conducted serial echocardiographic determinations of right ventricular function in patients receiving rtPA with heparin versus heparin alone for treatment of PE. One hundred one patients with PE diagnosed by high probability ventilation-perfusion scanning were randomized into these two groups. Patients underwent baseline, 3-hour, and 24hour echocardiograms. Patients receiving rtPA plus heparin had a 39% improvement in right ventricular function between baseline and 24-hour echocardiograms, whereas patients receiving heparin alone showed only a 17% improvement in right ventricular function. The results of these studies indicate that echocardiography is a poor diagnostic tool for PE. Echocardiography is useful in determining right ventricular dysfunction caused by PE and, therefore, may have utility in the decision to treat with thrombolytic agents. All patients with PE and moderate-to-severe right ventricular dysfunction should be considered for thrombolytic therapy (Fig. 11) [34].
636
PE diagnosed by angiography, CT angiography, or ventilationperfusion scanning Stable patient? Assess right ventricular function with echocardiogram Right ventricular dysfunction present? No Yes Contraindication to thrombolysis? Unstable patient?
Consider thrombolysis
Yes
No
Conservative management
Fig. 11. Echocardiographic management algorithm for pulmonary embolism.
Transesophageal echocardiography in the emergency department Certain disease states may require the EP to obtain consultation from cardiologists to perform transesophageal echocardiography. These states include infective endocarditis and aortic dissection/aneurysm.
Infective endocarditis Infective endocarditis is a disease state in which an infectious lesion, known as a vegetation, forms on a cardiac valve. This is a true multisystem disease because systemic embolic phenomenon can occur throughout the body. Risk factors for endocarditis include pre-existing valvular lesions from rheumatic heart disease or congenital defects, prosthetic heart valves, and intravenous drug abuse. Infective endocarditis remains a dicult diagnosis for the EP because classic clinical ndings of fever, heart murmur, and peripheral embolic phenomenon may not be present early in the disease process [35]. Echocardiography has assumed a vital role in the diagnosis of this disorder, with its inclusion into the Dukes criteria [36]. The echocardiogram also can provide important data on complications of infective endocarditis such as perivalvular abscesses and valvular insuciency. Transthoracic and transesophageal echocardiography have applications in the diagnosis of infective endocarditis. Transthoracic echocardiography can identify vegetations as small as 5 mm, with a sensitivity of 30% to 100% [37,38]. Transesophageal echocardiography is considered the superior test and can identify vegetations as small as 1 mm, with sensitivities close to 100% [39,40]. EPs suspecting the diagnosis of infective endocarditis should
637
obtain prompt cardiology consultation for transthoracic echocardiography. If transthoracic echocardiography results are equivocal, performing transesophageal echocardiography should be considered. Aortic emergencies Structural pathology to the thoracic aorta can result in high morbidity and mortality. Although rare, these diagnoses often are missed during ED evaluation [41]. Echocardiography can oer a more rapid evaluation of aortic dissection and aneurysm than more traditional angiographic or CT angiographic studies. An echocardiogram can reveal an intimal ap within the true aortic lumen. Information on complications of aortic dissection such as valvular insuciency, pericardial eusion/tamponade, and focal wall motion abnormalities from dissection into the coronary arteries also can be found [37]. Although the sensitivity of transthoracic echocardiography is about 75% for aortic dissection, sensitivity close to 100% can be obtained when transesophageal echocardiography is used [37,4143]. If transesophageal echocardiography consultation can be obtained in a timely manner, then the EP may consider the use of this diagnostic modality in suspected aortic dissection.
Summary The use of echocardiography in the ED is well established and continues to gain widespread use in the evaluation of critically ill patients. In certain circumstances such as chest trauma, pericardial eusion, and cardiac arrest, EPs can perform and interpret echocardiographic examinations reliably. In other circumstances such as the diagnosis of acute coronary syndromes, PE, and endocarditis, the EP should be aware of the uses and limitations of echocardiography and obtain appropriate consultation when necessary. Academic- and community-based EPs should seek to incorporate further the use of echocardiography in their respective clinical practices, with special attention given to training and quality assurance. As EPs continue to improve their skills in cardiac ultrasound, their ability to diagnose a wider spectrum of cardiac diseases undoubtedly will grow proportionally.
Acknowledgments The authors extend special thanks to Warren J. Manning, MD, FACP, FACC (Beth Israel Deaconess Medical Center, Boston, MA) for supplying echocardiographic images and Cristina A. Ciccone (Bedford, MA) for original artwork.
638
References
[1] Stewart WJ, Douglas PS, Sagar K, Seward JB, Armstrong WF, Zoghbi W, et al. Echocardiography in emergency medicine: a policy statement by the American Society of Echocardiography and the American College of Cardiology. J Am Soc Echocardiog 1999; 12(1):824. [2] Mateer J, Plummer D, Heller M, Olson D, Jehle D, Overton D, et al. Model curriculum for physicians training in emergency ultrasonography. Ann Emerg Med 1994;23(1):95102. [3] Plummer D. Principles of emergency ultrasound and echocardiography. Ann Emerg Med 1989;18(12):12917. [4] Hauser AM. The emerging role of echocardiography in the emergency department. Ann Emerg Med 1989;18(12):1298302. [5] Mayron R, Gaudio FE, Plummer D, Asinger R, Elsperger J. Echocardiography performed by emergency physicians: impact on diagnosis and therapy. Ann Emerg Med 1988;17(2): 1504. [6] Levitt MA, Jan BA. The eect of real time 2-D echocardiography on medical decisionmaking in the emergency department. J Emerg Med 2002;22(3):22933. [7] Plummer D, Brunette D, Asinger R, Ruiz E. Emergency department echocardiography improves outcomes in penetrating cardiac injury. Ann Emerg Med 1992;21(6):70912. [8] Mandavia D, Honer R, Mahaney K, Henderson S. Bedside echocardiography by emergency physicians. Ann Emerg Med 2001;38(4):37782. [9] Amaya SC, Langsam A. Ultrasound detection of ventricular brillation disguised as asystole. Ann Emerg Med 1999;33(3):3446. [10] Blaivas M, Fox JC. Outcomes in cardiac arrest patients found to have cardiac standstill on the bedside emergency department echocardiogram. Acad Emerg Med 2001;8(6):61621. [11] Salen P, OConnor R, Sierzenski P, Passarello B, Pancu D, Melanson S, et al. Can cardiac sonography and capnography be used independently and in combination to predict resuscitation outcomes? Acad Emerg Med 2001;8(6):6105. [12] Jimenez E, Martin M, Krukenkamp I, Barrett J. Subxiphoid pericardiotomy versus echocardiography: a prospective evaluation of the diagnosis of occult penetrating cardiac injury. Surgery 1990;108(4):67580. [13] Nagy KK, Lohmann C, Kim DO, Barrett J. Role of echocardiography in the diagnosis of occult penetrating cardiac injury. J Trauma 1995;38(6):85962. [14] Tsang T, Enriquez-Sarano M, Freeman WK, Barnes M, Sinak L, Gersh B, et al. Consecutive 1127 therapeutic echocardiographically guided pericardiocenteses: clinical prole, practice patterns, and outcomes spanning 21 years. Mayo Clin Proc 2002;77(5):42936. [15] Levine MJ, Lorell BH, Divier DJ, Come P. Implications of echocardiographically assisted diagnosis of pericardial tamponade in contemporary medical patients. Detection before hemodynamic embarrassment. J Am Coll Cardiol 1991;17(1):5965. [16] Merce J, Sagrista Sauleda J, Permanyer MG, Evalgelista A, Soler-Soler J. Correlation between clinical and Doppler echocardiographic ndings in patients with moderate and large pericardial eusion: implications for the diagnosis and treatment of cardiac tamponade. Am Heart J 1999;138(4):75964. [17] Soler-Soler J, Sagrista Sauleda J, Permanyer-Miralda G. Management of pericardial eusion. Heart 2001;86(2):23540. [18] Bocka JJ, Overton DT, Hauser A. Electromechanical dissociation in human beings: an echocardiographic evaluation. Ann Emerg Med 1988;17(5):4502. [19] Kaul S, Stratienko AA, Pollock SG, Marieb MA, Keller MW, Sabia PJ. Value of twodimensional echocardiography for determining the basis of hemodynamic compromise in critically ill patients: a prospective study. J Am Soc Echocardiog 1994;7(6):598606. [20] Moore CL, Rose GA, Tayal VS, Sullivan DM, Arrowood JA, Kline JA. Determination of left ventricular function by emergency physician echocardiography of hypotensive patients. Acad Emerg Med 2002;9(3):18693.
639
[21] McLean AS, Needham A, Stewart D, Parkin R. Estimation of cardiac output by noninvasive echocardiographic techniques in the critically ill subject. Anaesth Intens Care 1997;25(3):2514. [22] Hauser G, Gangadharan V, Ramos R, Gordon S, Timmis G. Sequence of mechanical, electrocardiographic and clinical eects of repeated coronary occlusion in human beings. Echocardiographic observation during coronary angioplasty. J Am Coll Cardiol 1985;5(2): 1937. [23] Selker HP, Zalenski RJ, Antman EM, Aufderheide TP, Bernard SA, Bonow RO, et al. An evaluation of technologies for identifying acute cardiac ischemia in the emergency department: executive summary of a National Heart Attack Alert Program Working Group report. Ann Emerg Med 1997;29(1):6974. [24] Peels CH, Visser CA, Funke Kupper AJ, Visser FC, Roos JP. Usefulness of twodimensional echocardiography for immediate detection of myocardial ischemia in the emergency room. Am J Cardiol 1990;65(11):68791. [25] Sabia P, Afrookteh A, Touchstone DA, Keller MW, Esquivel L, Sanjiv K. Value of regional wall motion abnormality in the emergency room diagnosis of acute myocardial infarction. Circulation 1991;84(3):I8592. [26] Krontos MC, Arrowood JA, Paulsen WHJ, Nixon JV. Early echocardiography can predict cardiac events in emergency department patients with chest pain. Ann Emerg Med 1998; 31(5):5507. [27] Buchsbaum MS, Marshall ES, Levine BJ, Bennett MA, DiSabatino A, OConnor RE, et al. Emergency department evaluation of chest pain using exercise stress echocardiography. Acad Emerg Med 2001;8(2):1969. [28] Kline JA, Johns KL, Colucciello SA, Israel EG. New diagnostic tests for pulmonary embolism. Ann Emerg Med 2000;35(2):16880. [29] Jackson RE, Rudoni RR, Hauser AM, Pascual RG, Hussey ME. Prospective evaluation of two-dimensional transthoracic echocardiography in emergency department patients with suspected pulmonary embolism. Acad Emerg Med 2000;7(9):9948. [30] Goldhaber SZ. Echocardiography in the management of pulmonary embolism. Ann Intern Med 2002;136(9):691700. [31] Ribeiro A, Lindmarker P, Juhlin-Dannfelt A, Johnsson H, Jorfeldt L. Echocardiography Doppler in pulmonary embolism: right ventricular dysfunction as a predictor of mortality rate. Am Heart J 1997;134(3):47987. [32] Konstantinides S, Geibel A, Heusel G, Heinrich F, Kasper W. Heparin plus alteplase compared with heparin alone in patients with submassive pulmonary embolism. N Engl J Med 2002;347(15):114350. [33] Goldhaber SZ, Haire WD, Feldstein ML, Miller M, Toltzis R, Smith JL, et al. Alteplase versus heparin in acute pulmonary embolism: randomised trial assessing right-ventricular function and pulmonary perfusion. Lancet 1993;341(8844):50711. [34] Goldhaber SZ. Thrombolysis for pulmonary embolism. N Engl J Med 2002;347(15): 11312. [35] Dunmire SM. Infective endocarditis and acquired valvular heart disease. In: Rosen P, Barkin R, editors. Emergency medicine: concepts and clinical practice. 4th edition. St. Louis (MO): Mosby; 1998. p. 174554. [36] Durack DT, Lukes AS, Bright DK. New criteria for diagnosis of infective endocarditis: utilization of specic echocardiographic ndings. Am J Med 1994;96:2009. [37] Wellford AL, Snoey ER. Emergency medicine applications of echocardiography. Emerg Med Clin N Am 1995;13(4):83154. [38] Sachdev M, Peterson GE, Jollis JG. Imaging techniques for diagnosis of infective endocarditis. Infect Dis Clin N Am 2002;16:31937. [39] Erbel R, Rohmann S, Drexler M, Mohr-Kahaly S, Gerharz CD, Oversen S, et al. Improved diagnostic value of echocardiography in patients with infective endocarditis by transesophageal approach: a prospective study. Eur Heart J 1988;9(1):4353.
640
[40] Shively BK, Gurule FT, Roldan CA, Leggett JH, Schiller NB. Diagnostic value of transesophageal compared with transthoracic echocardiography in infective endocarditis. J Am Coll Cardiol 1991;18(2):3917. [41] Ross C, Deen J, Lydon D. Aortic dissection and aortic aneurysms: what the ED physician must consider when facing these conditions. Emerg Legal Briengs 1993;33:26. [42] Barbant S, Eisenberg M, Schiller N. The diagnostic value of imaging techniques for aortic dissection. Am Heart J 1992;124(2):5413. [43] Nienaber CA, Spielmann RP, von Kodolitsch Y, Siglow V, Piepho A, Jaup T, et al. Diagnosis of thoracic aortic dissection: magnetic resonance imaging versus transesophageal echocardiography. Circulation 1992;85(2):43446.
Renal ultrasound
Vicki E. Noble, MDa,b, David F.M. Brown, MDa,b,*
Department of Emergency Medicine, Massachusetts General Hospital, Bulnch 105, Boston, MA 02114, USA b Division of Emergency Medicine, Harvard Medical School, Boston, MA, USA
a
There are a number of diagnostic imaging modalities available to assist in the evaluation of patients who present to the emergency department (ED) with acute urologic disease. Traditionally, intravenous pyelography (IVP) was the test of choice but, in recent years, it has been supplanted by helical CT. Ultrasound (US), angiography, and MRI are the other available diagnostic imaging modalities. US is a particularly attractive option for many emergency physicians (EPs) because it is noninvasive, can be performed at the bedside, and does not require exposure to contrast material or radiation. Moreover, many EPs have acquired the skills to perform limited, focused examinations themselves. This article reviews the role of formal renal US in the evaluation of acute urologic emergencies, focusing on clinical indications, radiologic ndings, and relative merits compared with other diagnostic modalities. This article also reviews the indications for bedside renal US performed and interpreted by EPs and summarizes the growing body of evidence in support of this practice. In addition, the technique of performing the US examination is reviewed.
Formal renal ultrasonography Formal renal US is a detailed examination of the renal and urologic systems performed by a radiologist or radiology technician. US has several advantages over other diagnostic imaging modalities: it provides excellent detailed renal anatomic information (Fig. 1), it can be performed at the bedside without moving the patient to an unmonitored setting, it can provide information about other organ systems when the EP suspects alternative diagnoses, and it is noninvasive. In addition, it does not expose the patient to radiation or require the administration of iodinated intravenous (IV) contrast, which is
* Corresponding author. E-mail address: dbrown2@partners.org (D.F.M. Brown). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.014
642
V.E. Noble, D.F.M. Brown / Emerg Med Clin N Am 22 (2004) 641659
Fig. 1. Normal right kidney on longitudinal view. Note the echogenic white Gerotas fascia, the grainy gray renal cortex, and the central echogenic pelvicalyceal structures. (From Brown DFM, Rosen CL, Wolfe RE. Renal ultrasonography. Emerg Med Clin N Am 1997;15:878; with permission.)
important because IV contrast administration has two major drawbacks: it is nephrotoxic [1] and may be contraindicated in patients with renal insufciency [2]. IV contrast also can cause allergic reactionsmost signicantly, anaphylaxis, which can be life threatening. US has several limitations: it is more operator dependent than other modalities and, despite its anatomic detail, can provide only limited assessment of renal function, unlike CT or IVP. If the US machine is capable of Doppler technology and both ureter jets are observed, however, global renal function can be assumed [3]. The ED indications for formal renal ultrasonography are listed in Box 1 and are discussed in detail in the following sections. Renal colic Renal colic is one of the most common diseases of the urinary tract. It is estimated that 2% to 5% of the population will form a urinary stone at some point in their lives [4]. Renal colic is the term used to describe the
Box 1. Emergency department indications for formal renal ultrasound Renal colic Acute renal vein thrombosis Renal failure Renal mass Acute renal infection Renal trauma Urinary retention
643
abrupt, severe, sharp ank and low back pain caused by the acute obstruction and distention of the ureter and renal pelvis. Associated symptoms include nausea, vomiting, diaphoresis, and radiation of pain to the groin. The most common cause of renal colic is ureterolithiasis, although obstruction of the ureter also may be due to ureteral spasm, thrombus formation within the ureter, or the presence of sloughed papillae within the ureter in the setting of acute papillary necrosis. The diagnosis of nephrolithiasis frequently is made on clinical grounds in patients with microscopic hematuria and the appropriate history. In many patients, however, a conrmatory imaging study is needed. IVP, long the radiologic mainstay in the diagnosis of renal colic, largely has been supplanted by helical CT in the last several years [514]. CT has a shorter examination time, provides better visualization of the calculus, and allows its measurement anywhere along the urinary tract, including the bladder. In addition, CT gives detailed information about other abdominal structures, which can be useful if the cause of the symptoms is not within the urinary tract. Generally, CT is performed without the administration of IV iodinated contrast. If the stone is not visualized, then IV contrast may be used to look for other causes of pain and hematuria. When contrast is added, CT also can provide an accurate assessment of renal function and renal vasculature. CT, however, carries the same risk of radiation exposure as IVP and, when contrast is used, carries the same allergic and nephrotoxic risks outlined in the previous section [49]. The overall advantages of CT usually make formal renal US the second choice in diagnostic imaging of suspected renal colic. Renal US may identify the actual calculi, which are seen as small echogenic structures with posterior shadowing (Fig. 2), particularly when the calculi are intrarenal. Intraureteral stones are very difcult to demonstrate, especially when they are small or in obese patients [15,48]. More commonly, US will reveal some degree of unilateral hydronephrosis, a sign of ureteral obstruction. Hydronephrosis occurs when the central calyceal system is dilated with urine, and appears sonographically as dark black and homogeneously anechoic (Fig. 3) [16]. It can be characterized as mild, moderate, or severe (Table 1), but these characterizations have poor correlation with the degree and acuity of obstruction [17]. The test characteristics for ultrasonographic detection of urinary tract calculi depend on which diagnostic criteria are used (direct visualization of the stone, unilateral hydronephrosis, or both). Using both criteria, Sinclair and associates [18] demonstrated a sensitivity of 85% and a specicity of 100% for US (compared with 90% and 94%, respectively for IVP). The presence of an identiable calculus alone as the diagnostic criterion yielded a sensitivity of 64% and a specicity of 100%; the presence of obstructive hydronephrosis alone had a sensitivity of 85% and a specicity of 100% [18]. Svedstrom and colleagues [19] demonstrated that plain radiographs and US each had a sensitivity of 60%, but when used together, the sensitivity rose to 80% for the diagnosis of nephrolithiasis. Other
644
Fig. 2. Renal stone. Longitudinal view of kidney containing a hyperechoic calculus (open arrow) with posterior acoustic shadowing. (From Van Arsdalen KN, Banner MP, Pollack HM. Radiographic imaging and urologic decision making in the management of renal and ureteral calculi. Urol Clin N Am 1990;17:186; with permission.)
investigators have described similar results when combining results of US and plain radiographs, thus approaching the diagnostic accuracy of IVP [20,21]. More applicable to current clinical practice, however, are the studies that compare US with noncontrast helical CT. Fowler [8] showed that US depicted 24 of 101 calculi identied on CT, for a sensitivity of 24% and
Fig. 3. Hydronephrosis. Longitudinal view of the kidney reveals dilated renal collecting system and pelvis. Note the calculus in the renal pelvis (arrow). (From Chang TS, Lepanto L. Ultrasonography in the emergency setting. Emerg Med Clin N Am 1992;10:8; with permission.)
V.E. Noble, D.F.M. Brown / Emerg Med Clin N Am 22 (2004) 641659 Table 1 Grades of hydronephrosis Grade I Slight blunting of calyceal fornices Grade II (mild) Obvious blunting of calyceal fornices and enlargement of calices, but intruding shadows of papillae easily are seen Grade III (moderate) Rounding of calices with obliteration of papillae
645
Grade IV (severe) Extreme calyceal ballooning
Adapted from Grainger RG, Allison DJ, editors. Diagnostic radiology: a textbook of medical imaging. 4th edition. London: Churchill Livingstone; 2001; p. 1594.
a specicity of 90%. No difference in stone detection of the left versus the right kidney was found for US. Smith and colleagues [12] showed that CT was more sensitive and specic than KUB, US, IVP, or any combination of the these diagnostic modalities. The value of nonenhanced helical CT was demonstrated by its sensitivity of 97%, specicity of 96%, and overall accuracy of 97%. Finally, in a prospective comparison of US and nonenhanced helical CT, Sheafor and colleagues [10] showed that CT identied 22 of 23 known stones, for a sensitivity of 96%, whereas US identied 14 of 23 calculi, for a sensitivity of 61%. Both tests had a specicity of 100%. When the tests were compared for any clinically relevant abnormality (unilateral hydronephrosis or urolithiasis with obstructing calculus), the sensitivities of US and CT increased to 92% and 100%, respectively. US missed one case of appendicitis as an alternative diagnosis and CT missed one case of medullary calcinosis [10]. Despite this evidence for CT superiority, US remains the test of choice for pregnant women; when helical CT is not available; and for those patients with relative contraindications to IV contrast such as allergy, renal insufciency, and diabetes mellitus when dye must be used. Renal failure For ED patients with newly diagnosed or worsening renal failure, US remains the imaging modality of choice because contrast material is not required. The causes of renal failure classically are categorized as prerenal, postrenal, or intrinsic to the kidney. US immediately is helpful in immediately identifying postrenal causes of failure because obstruction (ureteral or urethral) can be ruled out by the absence of hydronephrosis and a quick estimation of bladder size [17,22,23]. Moreover, prostatic enlargement, a common cause of lower tract obstruction, also can be identied on US [16]. Prerenal causes of renal failure generally will not be associated with sonographic abnormalities, but intrinsic causes often will have notable ndings. The rst issue to address using US is that of renal size, which gives a rough indication of the chronicity of the renal failure. Enlarged kidneys suggest an acute case of renal failure, such as infection, renal vein thrombosis,
646
or transplant rejection. Small kidneys imply chronic disease. In addition, an assessment of the echogenicity of the renal parenchyma can be made. An increase in echogenicity suggests chronic renal failure (Fig. 4) [15]. Finally, identication of a solitary kidney is important in any evaluation of new renal failure and readily is readily accomplished by US. Acute renal infection Pyelonephritis, an infection of the upper urinary tract, is a very common ED diagnosis and usually does not require imaging studies for diagnosis or management. Patients present with ank pain, fever, nausea and vomiting, pyuria, and bacteriuria. A small subset of these patients will progress to focal abscess formation that is resistant to standard antibiotic therapy and requires diagnostic imaging. Renal US in uncomplicated pyelonephritis generally reveals a normal-appearing kidney, although in some patients, the involved kidney may become enlarged with hypoechoic foci at the corticomedullary junction [15,16,24]. As the disease process becomes more advanced, these foci become larger, representing areas of bacterial nephritis. Renal abscesses are identied as larger, well-dened hypoechoic masses that occasionally may appear cystic. Emphysematous pyelonephritis is a rare bacterial infection generally limited to people who have diabetes. It can be diagnosed by renal US when high-amplitude echoes are seen within the renal parenchyma or sinus, with shadowing that contains low-level echoes and reverberations [25,26]. Plain abdominal radiographs or CT scans can conrm the presence of gas within the renal parenchyma. Distention of the pelvis and calyces of the kidney with pus can occur when infection develops in an obstructed kidney and is suggested on US by the presence of echoes in a dilated collecting system.
Fig. 4. Renal failure. Longitudinal scan of right kidney demonstrates an increase in the echogenicity of the renal parenchyma suggesting acute renal failure. (Courtesy of D. Riley, MD, St. Lukes/Roosevelt Hospital, New York).
647
Finally, perinephric abscess can be detected by US, appearing as circumscribed uid collections around the involved kidney. These collections often have internal septations or mobile debris, appear hypoechoic, and have a degree of posterior acoustic enhancement [27,28]. Usually when these ndings are seen, further evaluation with CT scanning is indicated.
Urinary retention As mentioned previously, postrenal causes of renal failure are assessed rapidly with US by looking for obstruction and bladder distention. In addition, US is useful for patients presenting with complaints of urinary retention, urgency, or incontinence. In these clinical situations, bladder volume can be estimated by measuring the bladder at its maximal width, depth, and length and applying the following formula: width depth length 0.75 [29]. Any patient with a postvoid residual volume greater than 100 mL has urinary retention. The prostate can be assessed while looking at the bladder and is best seen on transverse views at the bladder neck. A normal prostate should be 5 cm in width. In addition, the bladder wall can be assessed and should be smooth and of uniform thickness. The wall thickness depends on the degree of bladder distention, but as a rule of thumb, a thickness greater than 5 mm is abnormal [30]. The technique for scanning the bladder is simple. The probe should be placed just superior to the patients symphysis pubis and angled toward the patients feet. The bladder should be scanned in the sagittal and transverse planes.
Acute renal vein thrombosis Patients with acute renal vein thrombosis present with ank pain and tenderness, hypertension, and proteinuria. It particularly is prevalent in patients with renal transplants but also should be suspected in patients with nephrotic syndrome, malignancy, infections, and trauma. Signicant improvements in MRI and MRV technology have greatly expanded their role in the evaluation of acute renal vein thrombosis [31]. Contrast enhanced CT scanning or angiography, however, remain the mainstay for diagnosis, except in certain clinical conditions or populations (pregnancy, pediatrics, transplant recipients) where renal US is preferred [31]. US is diagnostic when it includes Doppler studies that show the absence of blood ow in the renal vein [28,50].
Renal masses Renal masses are being identied with increasing frequency in the ED because of increased use of emergency sonography and other imaging modalities [3235]. Although rarely of emergent clinical signicance, these
648
masses, after being identied, always deserve referral and further diagnostic imaging (often a CT scan). There is no question that the morbidity and mortality of patients are improved by detection of malignancies before they are symptomatic [33]. In addition, polycystic kidney disease may be identied in the ED because it can present with hematuria, ank pain, hypertension, and renal failure. The multiple cysts of varying sizes in both kidneys that enlarge and distort regular renal architecture are best identied by US [27,28]. If these ndings are observed, referral for further workup is mandatory.
Renal trauma The kidney frequently is injured in victims of trauma. Although increasingly, these injuries are being managed nonoperatively, many still require surgery to control hemorrhage and to prevent delayed abscess formation and hypertension. Concern for missing a surgical lesion results in the high use of radiologic studies. Contrast enhanced CT scanning is the diagnostic imaging test of choice because it provides information about renal function and vascular status in addition to detailed anatomic data [36]; however, US plays a role in the evaluation of renal injuries in trauma. The traditional role for renal US in trauma is as part of the Focused Assessment with Sonography for Trauma (FAST) examination, whereby it is used as a screening tool to look for intraperitoneal uid. The specics of the FAST examination are discussed in other articles within this issue, but it should be noted that it is not designed to pick up renal injuries but only identies intraperitoneal uid, suggesting ongoing hemorrhage or other abdominal catastrophes necessitating laparotomy. For specic renal injuries, US has been found to be highly reliable in distinguishing renal contusion from more serious injuries in Austria, although these results have not been reproduced in trauma centers in the United States [37,38]. In addition, renal US can show bleeding into the retroperitoneal space as a hypoechoic area around the kidney, although the sensitivity for US in diagnosing retroperitoneal hemorrhage is low. Focal areas of parenchymal hemorrhage and edema may be seen as hypoechoic areas within the kidney. A linear, reproducible absence of echoes suggests renal fractures. If the collecting system is injured, then urine may leak out of the kidney yet be contained between the renal capsule and Gerotas fascia, creating a urinoma (Fig. 5). A urinoma should be considered when an anechoic ring is seen around a portion of the kidney. Although US is a sensitive method for demonstrating a urinoma, the differential diagnosis includes lymphocele, hematoma, abscess, cyst, and ascites. Patients with this nding require further testing [27,28]. New technologic advances such as power color Doppler may provide an alternative diagnostic strategy to CT for the diagnosis of renal perfusion injuries. This technique deserves further study because it has not been thoroughly evaluated in the trauma setting [16]. One
649
Fig. 5. Blunt renal trauma. (A) Longitudinal section of right kidney shows hydronephrosis with echogenic material in the collecting system (asterisk). Note the clubbed calyxes. (B) Transverse view demonstrates a mixed-density uid collection medial to the kidney (arrows). Diagnosis was ureteropelvic junction obstruction with blood in the dilated collecting system and perinephric urinoma/hematoma. (From Loberant N. Emergency imaging of the urinary tract. Emerg Med Clin N Am 1992;10:75; with permission.)
nal important role for US in renal trauma is in the management of patients with identied parenchymal injuries such as hematomas and lacerations. These lesions often are well visualized and can be evaluated periodically to monitor their resolution so that patients do not need to undergo further testing with radiation exposure risks [16,38].
Renal ultrasound performed by emergency physicians The use and acceptance of limited emergency abdominal US performed and interpreted by EPs has increased during the past decade. US training now is part of the core content for residency training in emergency medicine. In addition, standards for performing US and requirements for training have been published by the American College of Emergency Physicians and are the current standard for emergency medicine [39]. Emergency US generally is performed at the bedside in the ED using a portable machine and is designed to answer simple clinical questions. It is not intended to be a comprehensive formal US. Emergency renal US performed by EPs has the same previously described advantages and limitations as formal US compared with other diagnostic imaging modalities. In addition, emergency renal US offers several advantages over conventional formal renal US: it is immediately available regardless of the time of day and patients do not have
650
to leave the ED to go to the radiology suite or other minimally monitored settings. The examining physician performs the US and so the evaluation can be focused on the patients signs and symptoms. It is repeated easily one or more times as clinical parameters change or to monitor response to therapy [47]. For example, it is well known that dehydrated patients will have false-negative US ndings when looking for hydronephrosis. After IV hydration, these same patients often develop signicant ndings on repeat examinations [16]. Finally, the physician performing the examination can look at other abdominal structures, especially the abdominal aorta, to help in the consideration of alternative diagnoses. The literature that supports ED ultrasonography by EPs continues to grow. Mandavia [40] provided a prospective analysis of the impact of a US training program on a residency training program. After an introductory course (16 hours) and minimal hands-on training, it was reported that emergency medicine residents had a sensitivity of 92% and a specicity of 96% on 1138 focused US examinations. These focused US examinations were a combined group of all six ACEP-indicated US categories (renal, pelvic, trauma, right upper quadrant, aortic, and cardiac). Lanoix [36], in describing the introduction of a US curriculum into a training program, concluded that with a minimal amount of training [residents] display acceptable US technical skill and interpretive acumen. Specically, renal US as performed by residents was reported to be 94% sensitive and 96% specic for ruling out hydronephrosis. The resident-performed US examinations were over-read by radiologists or certied ultrasonographers to determine their accuracy. Among earlier studies, Rosen and colleagues [41] described their experience with EP-performed renal US. Renal US examinations were performed by EP operators with limited training (5-hour introductory course) and were found to have a sensitivity of 71%, a specicity of 75%, a positive predictive value of 83%, and a negative predictive value of 60% for detecting hydronephrosis compared with IVP. Henderson and colleagues [42] also compared EPperformed renal US with a plain radiograph of the abdomen (KUB) and IVP. They found that minimally trained EPs had a sensitivity of 97% but a specicity of 59% in detecting hydronephrosis. Their positive predictive value was 80.7% and their negative predictive value was 92%. One of the reasons for the higher sensitivity in this study was that all patients received a 500-mL bolus of normal saline before US. All of these studies lend support to the idea that minimally trained EPs can use renal US to diagnose hydronephrosis accurately.
Technique The technique of emergency renal US performed by EPs is simple. It can and should be performed at the bedside, obviating the need to move the patient. The study begins with the patient supine. Generally a 3.5-MHz
651
transducer is used, although a 5-MHz probe can provide high-quality images in thin patients and in children. The right kidney is more easily accessible to US because it is located adjacent to the liver, which serves as an excellent acoustic window. The probe should be placed (with the marker toward the patients head) along the right lateral subcostal margin in one of the lower right intercostals spaces in the anterior axillary line, scanning through the liver to locate the right kidney. When bowel gas is present, it will reect the US waves and obscure the right kidney in this position. The probe then should be moved laterally to the mid or posterior axillary line. After the kidney has been visualized, the position of the probe should be adjusted slowly to obtain the optimum longitudinal view. Thereafter, the probe should not be moved along the skin but rather rocked slowly back and forth to sweep medially and laterally through the kidney in the longitudinal plane (long-axis view). The probe then should be rotated 90( to visualize the kidney in the transverse plane (short-axis view). The left kidney is more dicult to visualize because of overlying bowel gas or air in the stomach, its more superior location, and the absence of the liver to provide an acoustic window. US of the left kidney is performed with the probe placed in the posterior axillary line or in the left costovertebral angle with the patient in the right lateral decubitus position. After the left kidney is located, the scanning technique is the same as for the right. Views should be obtained in the longitudinal and transverse planes. For both right and left kidneys, it is important to visualize the kidney longitudinally and include the tip of the inferior pole because uid often collects there rst as it is the most dependent position. An important US technique is to have the cooperative patient inspire and hold his or her breath. This action will displace both kidneys inferiorly as much as 2.5 cm and may provide a more optimal view [43]. Normal kidneys measure 4 to 5 cm in width, 9 to 12 cm in length, and generally are within 2 cm of each other in length [44]. Each kidney has an echogenic capsule that represents Gerotas fascia and surrounding perinephric fat [15,43]. The kidney has a long-axis appearance that is football shaped (Fig. 6), and is round or C shaped when viewed along its short axis (Fig. 7). The renal sinus is composed of renal vessels, the pelvicalyceal system, and surrounding brous and adipose tissue. It appears on the longaxis as a bright, echodense central complex surrounded by less echodense parenchyma. On transverse view, the central collecting system structure is echodense, round in shape, and more medially located.
Focused emergency department ultrasound objectives The focus of the bedside renal US examination in ED patients is to determine the presence or absence of hydronephrosis as an indicator of urolithiasis (Fig. 8). The bright echodense central collecting system will be
652
Fig. 6. Longitudinal view of normal kidney with length and width measurements. (From Brown DFM, et al. Renal ultrasonography. Emerg Med Clin N Am 1997;15:887; with permission.)
distended by homogeneous black anechoic branching areas that represent uid in the renal pelvicalyceal space [15]. Hydronephrosis must be distinguished from renal cysts that also appear as anechoic collections of uid within the kidney (Fig. 9). Scanning through the kidney in real time can make this distinction. Hydronephrosis is distributed within the collecting
Fig. 7. Transverse (short-axis) view of normal kidney. Note the echogenic white Gerotas fascia, the grainy gray renal cortex, and the central echogenic pelvicalyceal structures. (From Brown DFM, et al. Renal ultrasonography. Emerg Med Clin N Am 1997;15:887; with permission.)
653
Fig. 8. Hydronephrosis. Longitudinal view of right kidney with dilated collecting system.
system, whereas cysts will be located focally within the renal cortex, with a normal-appearing collecting system. It also is important to distinguish hydronephrosis from an extrarenal pelvis, a commonly seen normal variant. In these patients, the collecting system generally is outside the kidney, appearing as a medial anechoic uid collection that can be mistaken for a sign of obstruction. The EP also should look for a perinephric uid collection, which may represent calyceal rupture and extravasation of urine
Fig. 9. Renal cyst. Longitudinal scan of right kidney demonstrates intraparenchymal, smoothwalled, uid-lled structure. (Courtesy of D. Riley, MD, St. Lukes/Roosevelt Hospital, New York).
654
resulting from high-grade obstruction. The opposite kidney must be scanned to distinguish unilateral from bilateral hydronephrosis, which has a different differential diagnosis that usually is related to bladder outlet obstruction. There are some common pitfalls in renal US when assessing for hydronephrosis. Specically, false-negative scans are encountered when scanning a markedly dehydrated patient because the presence of hydronephrosis commonly is masked in the dehydrated patient [16,40]. Falsepositive ndings are associated with conditions such as pregnancy, polycystic kidney disease (Fig. 10) [45], vesicoureteral reux (Fig. 11) [45], and an overdistended bladder. These conditions can mimic bilateral hydronephrosis [16]. Overdistended bladders easily are avoided by having the patient void before evaluating renal architecture if possible.
Clinical algorithm for bedside emergency department renal ultrasound Bedside EP-performed renal US is indicated in all patients presenting with ank pain who are suspected of having renal colic. The classic patient is the patient with ank pain and hematuria, but US can be performed on all patients presenting with ank pain. In addition, all patients presenting with urinary retention and bladder distention can benet from bedside ED renal US that includes an assessment of bladder volume. Fig. 12 is an adapted clinical algorithm for patients with ank pain [16].
Summary Renal US is one of several imaging modalities available to the EP in the evaluation of patients with acute urologic disorders. It oers excellent
Fig. 10. Adult polycystic kidney disease. Longitudinal scan of right kidney demonstrates the classic appearance of multiple cysts of varying shape and size. (From Scheible W, Talner LB. Gray scale ultrasound and the GU tract. Radiol Clin N Am 1979;17:288; with permission.)
655
Fig. 11. Vesicoureteral reux with ultrasound scan false-positive for obstructive hydronephrosis. (A) Longitudinal view of right kidney demonstrates dilated collecting system (arrows). (B) Cystogram reveals massive bilateral reux as the cause of the collecting system dilatation. (From Amis ES, Hartman DS. Renal ultrasonography 1984: a practical overview. Radiol Clin N Am 1984;22:323; with permission.)
anatomic detail without exposure to radiation or contrast agents but is limited in its assessment of renal function. It is an important alternative to helical CT scanning for evaluating renal colic, especially in children and pregnant women. It has an important role in excluding bilateral renal obstruction as the cause of acute renal failure. It is likely that Doppler renal US also will take on a prominent role in the evaluation of renal vascular
656

Patient with flank pain, flank pain and hematuria or bladder distention
Focused Renal Ultrasound
Severe hydronephrosis
Mild to moderate hydronephrosis
No hydronephrosis
Assess Aorta
Assess Aorta
Repeat US after hydration
Normal aorta, Assess bladder volume
Abnormal aorta, Consider CT if stable Consult surgery, OR if unstable
Normal aorta, Assess bladder volume
No hydro, Consider alternative diagnosis
Normal bladder volume, Consider CT and urology consult/admit
Increased bladder volume, foley and reassess
Normal bladder volume, Treat clinically for renal colic
Improved, Discharge for urology follow-up with strainer
Not improved, consider CT and urology consult/admit
Fig. 12. Clinical algorithm for patients with ank pain. (Adapted from Swadron S, Mandavia DP. Renal ultrasound. In: Ma OJ, Mateer JR, editors. Emergency ultrasound. New York: McGraw Hill Professional; 2002; p. 199.)
disorders. It already has become the standard of care in the management of renal transplant patients [46]. Bedside emergency renal US performed and interpreted by EPs with limited training and experience is increasing in use and gaining acceptance. At present, the primary role of renal US is to identify hydronephrosis in patients with renal colic or acute renal failure
657
but, in the future, its role likely will expand as technology advances and its use increases. In many patients, bedside renal US may obviate the need for further diagnostic workup and speed the diagnosis and treatment of an emergency patient. References
[1] Gerlach AT, Pickworth KK. Contrast medium-induced nephrotoxicity: pathophysiology and prevention. Pharmacotherapy 2000;20:5408. [2] Waybill MM, Waybill PN. Contrast media-induced nephrotoxicity: identication of patients at risk and algorithms for prevention. J Vasc Interv Radiol 2001;12:39. [3] Burge HJ, Middleton WD, McClennan BL, et al. Ureteral jets in healthy subjects and in patients with unilateral ureteral calculi: comparison with color Doppler US. Radiology 1991;180:43742. [4] Rosen P, Barkin R, Ling LJ. Emergency medicine: concepts and clinical practice. 4th edition. St. Louis (MO): Mosby; 1998. [5] Chen MY, Zagoria RJ. Can noncontrast helical computed tomography replace intravenous urography for evaluation of patients with acute urinary tract colic? J Emerg Med 1999;17:299303. [6] Colistro R. Unenhanced helical CT in investigation of acute ank pain. Clin Radiol 2002; 57(6):43541. [7] Fielding JR, Steele G, Fox LA, et al. Spiral computerized tomography in the evaluation of acute ank pain: a replacement for excretory urography. J Urol 1997;157:20713. [8] Fowler KA. US for detecting renal calculi with nonenhanced CT as a reference standard. Radiology 2002;222(1):10913. [9] Sheley RC, Semonsen KG, Quinn SF. Helical CT in the evaluation of renal colic. Am J Emerg Med 1999;17:27982. [10] Sheafor DH, Hertzberg BS, Freed KS, et al. Nonenhanced helical CT and US in the emergency evaluation of patients with renal colic: prospective comparison. Radiology 2000;217:7927. [11] Smith RC, Roseneld AT, Kyuran AC. Acute ank pain: comparison of non-contrast enhanced CT and intravenous urography. Radiology 1995;194:78994. [12] Smith RC, Verga M, McCarthy S, et al. Diagnosis of acute ank pain: value of unenhanced helical CT. AJR Am J Roentgenol 1996;166:97101. [13] Swanson SK, Heilman RL, Eversman WG. Urinary tract stones in pregnancy. Surg Clin N Am 1995;75(1):12342. [14] Wong SK. Acute renal colic: value of unenhanced spiral CT compared with intravenous urography. Ann Acad Med Singapore 2001;30(6):56872. [15] Coleman BG. Ultrasonography of the upper genitourinary tract. Urol Clin N Am 1985;12: 63344. [16] Swadron S, Mandavia DP. Renal ultrasound. In: Ma OJ, Mateer JR, editors. Emergency ultrasound. New York: McGraw Hill Professional; 2002. pp. 197220. [17] Kiely EA, Hartnell GG, Gibson RN, et al. Measurement of bladder volume by real-time ultrasound. Br J Urol 1987;60:335. [18] Sinclair D, Wilson S, Toi A, et al. The evaluation of suspected renal colic: ultrasound scan vs excretory urography. Ann Emerg Med 1989;18:5569. [19] Svedstrom E, Alanen A, Nurmi M. Radiologic diagnosis of renal colic: the role of plain lms, excretory urography and sonography. Eur J Radiol 1990;11:1803. [20] Dalla Palma L, Stacul F, Bazzocchi M, et al. Ultrasonography and plain lm versus intravenous urography in ureteral colic. Clin Radiol 1993;47:3336. [21] Smith SL. The role of the plain radiograph and renal tract ultrasound in the management of children with renal tract calculi. Clin Radiol 2000;55(9):70810.
658
[22] Ireton RC, Krieger JN, Cardenas DD, et al. Bladder volume determination using a dedicated, portable ultrasound scanner. J Urol 1990;143:90911. [23] Mainprize TC, Drutz HP. Accuracy of total bladder volume and residual urine measurements: comparison between real-time ultrasound and catheterization. Am J Obstet Gynecol 1989;160:10136. [24] Kawashima A. Radiologic evaluation of patients with renal infections. Infect Dis Clin N Am 2003;17(2):43356. [25] Allen HA, Walsh JW, Brewer WH, et al. Sonography of rmphysematous pyelonephritis. J Ultrasound Med 1984;3:5337. [26] Huang JJ, Tseng CC. Emphysematous pyelonephritis: clinicoradiological classication, management, prognosis and pathogenesis. Arch Intern Med 2000;160:797805. [27] Thurston W, Wilson S. The urinary tract. In: Rumack C, Wilson S, Charboneau J, editors. Diagnostic ultrasound. St. Louis (MO): Mosby; 1997. pp. 32998. [28] Williamson M. Renal ultrasound. In: Williamson M, editor. Essentials of ultrasound. Philadelphia: WB Saunders; 1996. pp. 56279. [29] Chan H. Noninvasive bladder volume measurement. J Neurosci Nurs 1993;25:309. [30] Chang SD, Hricak H. Radiological evaluation of the urinary bladder, prostate and urethra. In: Grainger RG, Allison DJ, editors. Diagnostic radiology: a textbook of medical imaging. 4th edition. London: Churchill Livingstone; 2001. pp. 161552. [31] Brenner BM, Rector FC. Renal vein thrombosis. In: Brenner BM, Rector FC, editors. The kidney. Philadelphia: WB Saunders; 2000. [32] Mandavia DP, Pregerson B, Henderson SO. Ultrasonography of ank pain in the emergency department: renal cell carcinoma as a diagnostic concern. J Emerg Med 2000; 18:836. [33] Smith SJ, Bosniak MA, Megibow AJ, et al. Renal cell carcinoma: earlier discovery and increased detection. Radiology 1989;170:699703. [34] Tosaka A, Ohya K, Yamada K, et al. Incidence and properties of renal masses and asymptomatic renal cell carcinoma detected by abdominal ultrasonography. J Urol 1990; 144:10979. [35] Ueda T, Mihara Y. Incidental detection of renal carcinoma during radiological imaging. Br J Urol 1987;59:5135. [36] Lanoix R. A preliminary evaluation of emergency ultrasound in the setting of an emergency medicine training program. Am J Emerg Med 2000;18(1):415. [37] Buchberger W, Penz T, Wicke K, et al. Diagnosis and staging of blunt kidney trauma. A comparison of urinalysis, IV urography, sonography and computed tomography. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 1993;58:507. [38] Furtschegger A, Egender G, Jaske G. The value of sonography in the diagnosis and followup of patients with blunt renal trauma. Br J Urol 1988;62:1106. [39] American College of Emergency Physicians. Policy statement: use of ultrasound imaging by emergency physicians. Ann Emerg Med 1997;30:3645. [40] Mandavia DP. Ultrasound training for emergency physiciansa prospective study. Acad Emerg Med 2000;7(9):100814. [41] Rosen CL, Brown DFM, Sagarin MJ, et al. Ultrasonography by emergency physicians in patients with suspected ureteral colic. J Emerg Med 1998;16:86570. [42] Henderson SO, Hoffner RJ, Aragona JL, et al. Bedside emergency department ultrasonography plus radiography of the kidneys, ureters and bladder vs intravenous pyelography in the evaluation of suspected ureteral colic. Acad Emerg Med 1998;5(7): 66671. [43] Heller M, Jehle D. Primary applications of ultrasound. In: Heller M, Jehle D, editors. Ultrasound in emergency medicine. Philadelphia: WB Saunders; 1995. pp. 41104. [44] Brandt TD, Neiman HL, Dragowski MJ, et al. Ultrasound assessment of normal renal dimensions. J Ultrasound Med 1982;1:49.
659
[45] Amis ES, Cronan JJ, Pster RC, et al. Ultrasonic inaccuracies in diagnosing renal obstruction. Urology 1982;19(1):1015. [46] ONeill WC, Baumgarten DA. Sonography in transplantation: essential component of management of renal transplantation. Am J Kidney Dis 2002;39(4):66378. [47] Chang TS, Lepanto L. Ultrasonography in the emergency setting. Emerg Med Clin N Am 1992;10:125. [48] Kellet MJ. The genitourinary tract. In: Grainger RG, Allison DJ, editors. Diagnostic radiology: a textbook of medical imaging. 4th edition. London: Churchill Livingstone; 2001. pp. 148996. [49] Spencer BA, Wood BJ, Dretler SP. Helical CT and ureteral colic. Urol Clin N Am 2000;27: 23141. [50] Zubarev AV. Ultrasound of renal vessels. Eur Radiol 2001;11(10):190215.
Hepatobiliary ultrasound
Kaushal Shah, MDa,*, Richard E. Wolfe, MDb
a
St. LukesRoosevelt/Columbia University, Department of Emergency Medicine, University Hospital of Columbia University College of Physicians and Surgeons, 1111 Amsterdam Avenue, New York, NY 10025, USA b Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, One Deaconess Road, West Campus Clinical Center 2, Boston, MA 02115, USA
The use of ultrasound technology in the emergency department (ED) is a recent and fast-growing phenomenon. One of the most valuable uses of ultrasound in the ED is to image the hepatobiliary systemspecically, the gallbladder (GB). Ultrasound is an extremely valuable tool for the evaluation of GB disease in the ED for several reasons: this disease is a common medical problem, cholecystitis can present in dierent ways clinically, the nature of the GB allows it to be well visualized by ultrasound, and ultrasound has many benets and few complications associated with its use. It is estimated that gallstones are present in approximately 10% to 15% of Americans and that more than 20% of women and 8% of men over 40 years of age have gallstones [15]. Among patients with gallstones, it is estimated that new biliary pain will occur in 10% at 5 years, 15% at 10 years, and 18% at 15 to 20 years [4]. There are 500,000 cholecystectomies performed each year and approximately 6000 to 10,000 deaths associated with GB disease annually in the United States [3,5,6]. GB disease (ie, biliary colic, cholecystitis) can present to the ED in a variety of ways: right upper quadrant pain, epigastric pain, abdominal pain, right ank pain, right shoulder pain, nausea/vomiting, and sepsis without a source. The emergency physician (EP) must be able to narrow a large dierential diagnosis to make a diagnosis and disposition quickly and eciently. Realtime ultrasound allows the EP to assess the GB to determine whether the best rst diagnostic test should be formal ultrasound or abdominal CT and,
* Corresponding author. E-mail address: kshah@chpnet.org (K. Shah). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.015
662
K. Shah, R.E. Wolfe / Emerg Med Clin N Am 22 (2004) 661673
potentially, to provide earlier consultation, earlier pain medication, earlier antibiotics, and earlier denitive diagnosis or disposition [710]. The GB is ideal for ultrasound detection and disease diagnosis. It located fairly supercially in the body under the liver, which is an excellent acoustic window, meaning that it enhances resolution of ultrasound waves. The GB itself is a cystic structure that is lled only with echo-free substance (bile) in the normal state. Gallstones, the most common sign of GB disease, are identied easily because they are echogenic and cast an acoustic shadow. With real-time ultrasound, the movement of stones can be appreciated with changes in position of the patient; gallstones should settle in dependent areas. Ultrasound is the test of choice to diagnose acute cholelithiasis, with an accuracy of 90% to 95% [1113]. The benets of bedside ultrasound in the ED are safety (no contrast material or radiation is required), speed (no signicant preparation is required, such as oral contrast load or transport to another location in the hospital), cost (other than the initial purchase and minimal maintenance costs of an ultrasound machine, there is no additional cost to the physician or patient), and accuracy, especially to rule out disease. Limited, goal-directed emergency ultrasound The eld of emergency medicine has promoted the use of ultrasound with a specic, unique philosophy: limited, goal-directed ultrasonography. The EP is not expected to identify subtle abnormalities in the hepatobiliary system. The goal is to answer specic binary (yes or no) questions to help determine the presence or absence of disease (eg, Are there gallstones present? and Is there a sonographic Murphys sign?). If the pretest probability for GB disease is increased signicantly, then the EP not only can order a formal ultrasound but also initiate subsequent steps that normally would be delayed until after a formal ultrasound, such as surgical consultation, antibiotics, and pain medication. If the pretest probability is reduced signicantly, then the rst diagnostic test may be abdominal CT scan instead of formal ultrasound. In indeterminate cases (GB not visualized or tender GB without stones), there is only a small loss of time to perform the bedside ultrasound; in general, the time investment is worth the potential benet of ultrasound. The accuracy of EP ultrasonography of the hepatobiliary system has been investigated. This article reviews the focused examination of the GB, with specic attention to test characteristics (sensitivity, specicity, positive predictive value, and negative predictive value) when performed by EPs in the ED. Technical ultrasound considerations Although the GB can be assessed well with ultrasound, it is one of the more dicult structures to locate; therefore, a methodological approach is
663
recommended. Using a 3.5- 5.0-MHz ultrasound probe, the right upper quadrant of the abdomen is scanned in the longitudinal plane under the costal margin, initially using the liver as an acoustic window to identify the GB. If there is diculty, the patient can be asked to take a slow deep breath because the GB moves signicantly with respiration. Changing the position of the patient also can be helpful. To avoid bowel gas and rib shadows, the probe can be positioned in the intercostal space. After the GB is identied, conrmation is needed that the structure is indeed the GB. Associated structures must be identied because the GB easily can be mistaken for a vessel (eg, inferior vena cava), duct in the liver, or loop of intestine on cross-section. Helpful associated structures include (1) gallstones within the lumen of the GB that are echogenic with echolucent shadows; (2) the main lobar ssure of the liver (linear echogenic structure) in sagittal section points toward the GB neck and also connects the portal vein; (3) the common bile duct (CBD) usually runs between the GB and the portal vein and usually is seen when the probe is parallel to the subcostal margin (transverse or slightly oblique view); and occasionally (4) the hepatic artery visible anteromedial to the portal vein in addition to the CBD visible anterolateral to the portal vein, forming the Mickey Mouse sign (Fig. 1). Identifying these structures is essential to prove that the image is the GB and useful when taking pictures to convince consultants and colleagues that the structure is the GB. It is important always to visualize the GB in various planes (eg, longitudinal and transverse) to see the entire extent of the structure.
Fig. 1. The Mickey Mouse sign is comprised of the portal vein (PV), common bile duct (CBD), and hepatic artery (HA) as labeled on the ultrasound scan.
664
Gallstones or polyps occasionally can be seen in only certain views. Also, small stones impacted in the cystic duct can be overlooked. When gallstones are suspected, the position of the patient should be changed from lying at to left lateral recumbent to see the stones shift position. Gallstones should settle in dependent regions. If the stones do not move and there is no acoustic shadow, then they may not be stones at all. Rarely, the GB can be scarred down on a large gallstone, resulting in an atypical appearance of the GB. There will be no standard pear-shaped GB and no lumen. There may be only a hyperechoic area with distal shadowing. Focus emergency medicine examination As discussed previously, the goal of the EP is to identify the presence or absence of specic ndings to raise the suspicion of cholecystitis. The following is a review of the most common ndings and signs that should be sought during an ultrasound of the right upper quadrant. Gallstones/cholelithiasis The presence or absence of gallstones is critical because it is the primary sonographic criteria for the diagnosis of acute cholecystitis [12,14]. Most patients (approximately 95%99%) who have cholecystitis have gallstones [14,15]. It should be noted that stones also can be present in the setting of biliary colic or found incidentally (asymptomatic gallstones); however, in the ED ultrasound evaluation to rule out cholecystitis, every eort should be made to identify gallstones. As mentioned earlier, there are three important characteristics of gallstones: they appear as an echogenic focus within the GB, they may cast an acoustic shadow, and they exhibit gravitational dependency. Note that very small stones (usually less than 3 mm) may not cast a shadow due to the nature of ultrasound waves, especially with low-frequency transducers (Fig. 2) [12]. In an early prospective study by Jehle et al [16] in 1989, EPs moderately trained in sonography successfully identied all 21 patients with gallstones among 48 suspected of biliary tract disease. Subsequent studies also have found high rates of gallstone identication by EPs: Kendall and Shimp [8] reported successful identication of 49 of 51 patients with gallstones, yielding a sensitivity of 96%, and Rosen et al [9] reported successful identication of 60 of 65 patients with gallstones, yielding a sensitivity of 92%. The specicity in both studies was slightly lower: 88% and 78%, respectively. It is now generally accepted that moderately trained EPs can identify gallstones by ultrasonography with reasonable accuracy. Sonographic Murphys sign A sonographic Murphys sign is positive when maximal tenderness exists where pressure is applied with the ultrasound probe directly over the
665
Fig. 2. Echogenic gallstones casting an acoustic shadow are clearly visible.
sonographically visualized GB. It is, theoretically, more specic than the standard Murphys sign because tenderness is elicited from the GB specically and not just the right upper quadrant. It is important to test for sonographic Murphys sign because a positive sign in addition to the presence of gallstones has a positive predictive value greater than 90% for cholecystitis [12,14]. In one formal radiology study, sonographic Murphys sign was found to be 86% sensitive for acute cholecystitis [17]. In a study of 109 limited right upper quadrant ultrasounds performed by EPs, 54 patients had a sonographic Murphys sign that correlated with a 75% sensitivity for a pathologic diagnosis of acute cholecystitis; interestingly, the sensitivity was only 45% when elicited by formal ultrasound technicians [8]. In another study, combining the presence of gallstones and a positive sonographic Murphys sign yielded a sensitivity of 91% in the hands of EPs [9]. A sonographic Murphys sign is a valuable test to diagnose cholecystitis and can be elicited reliably by EPs. Gallbladder wall thickening GB wall thickening is a secondary sign of acute cholecystitis [18,19]. In general, however, it is a nonspecic nding because it may appear thickened in conditions other than GB disease. Other common medical problems that increase GB wall thickness are congestive heart failure, renal disease, hepatitis, ascites, and alcoholic liver disease (for a more complete list, see Box 1). Even in an otherwise healthy person, the GB wall may appear thickened when the GB is contracted, particularly after a large, fatty meal (Fig. 3). To avoid measuring artifact, the anterior wall always should be measured perpendicular to the ultrasound beam. The normal, noncontracted GB wall
666
Box 1. Differential diagnosis of thickened gallbladder wall Adenomyomatosis Alcoholic liver disease Ascites Cholecystitis Congestive heart failure Gallbladder tumors Hepatitis Hypoalbuminemia Hypoproteinemia Multiple myeloma Pericholecystic abscess Renal disease Systemic venous hypertension
Adapted from Simon B, Snoey E. Ultrasound in emergency and ambulatory medicine. St. Louis (MO): Mosby; 1997; with permission.
usually measures less than 3 mm; the mean GB wall thickness in acute cholecystitis is 9 mm [12,20]. For purposes of ED ultrasonography, a GB wall greater than 3 mm is considered abnormal. Ultrasound measurements correlate within 1 mm to pathology specimen ndings after surgery in greater than 90% of cases and within 1.5 mm in 100% of cases [18]. Clinical correlation is critical when using GB wall thickness to diagnose cholecystitis. In the appropriate clinical setting, a thickened GB wall in addition to gallstones can raise the positive predictive value greater than 90% [12,14]. Data on the correlation of EP measurement of GB wall thickness compared with formal ultrasonography measurements or pathology specimen measurements is sparse. A thick-walled, tender GB without gallstones should not be disregarded. Among 31 cases identied by a radiology department in Sweden, 14 were found to have stones and cholecystitis on further study, 7 had another etiology (concomitant disease) for the thickened GB wall, 7 had acalculous cholecystitis, and 3 did not have further evaluation [21]; clearly, stones can be elusive even to highly trained sonographers, and secondary ndings of cholecystitis in the right clinical setting are clues to the rare diagnosis of acalculous cholecystitis. Pericholecystic uid Pericholecystic uid is a sign of GB inammation and appears as uid (echolucent) within or around the GB wall. Identifying the presence of
667
Fig. 3. A contracted gallbladder can appear to have a thickened gallbladder wall. Also visible in this scan are the portal vein (PV) and, vaguely, the common bile duct (CBD) between the portal vein and the gallbladder.
pericholecystic uid is useful because it is highly specic for GB disease (either acute cholecystitis, GB perforation, or pericholecystic abscess) but seen only occasionally [14,20]. When there is enough inammation to produce pericholecystic uid around the entire GB, a halo or echolucent space develops around the external wall of the GB [22]. In one retrospective review, this halo sign was seen in 26% of patients with acute cholecystitis [23]. This sign is believed to be very specic in a noncontracted GB. Common bile duct dilatation Measurement of the CBD is recommended as part of the right upper quadrant ultrasound because a dilated CBD suggests obstruction most commonly due to choledocholithiasis (a stone in the biliary ducts). Depending on the location of the lodged stone, there can be inammation of the GB or pancreas. In cases of acute biliary pancreatitis, ultrasound is the initial test of choice because a gallstone is identied in 60% to 80% of cases [24,25]. Dilatation of the CBD greater than 6 mm is considered abnormal, but the CBD usually is less than one tenth of the patients age [22]. Data from ultrasound and autopsy studies clearly demonstrate that the size of the CBD increases with age [2628]. When scanning an elderly patient, the common rule of thumb is to add 1 mm the standard 6 mm for each decade over 60 years to determine the upper limit of normal; however, a large prospective study of elderly patients found that 98% had a CBD less than 6 mm and 99% had a CBD less than 7 mm.
668
Ultrasound is sensitive for the detection of CBD dilatation; however, even in the hands of formal ultrasonographers, detection of the underlying problem (choledocholithiasis, pancreatic mass, stricture, and so forth) is inaccurate [22]. The EP should attempt to identify and measure the CBD in every patient being evaluated for hepatobiliary disease but is not expected to determine the underlying cause based on ultrasonography. Other signs Gallbladder size In general, the GB size is not important or helpful. In an otherwise normal person, the size of the GB correlates best with the time and quality of the last meal. Normal GBs can be as large as 11 to 13 cm [10]. On the other hand, approximately 87% of infected GBs (cholecystitis) are larger than 4 to 5 cm in transverse diameter [20]; size may be somewhat sensitive but certainly not specic for acute cholecystitis. Biliary sludge On occasion, gallstones are not identied; however, low-amplitude echoes in the dependent portion of the GB that do not produce signicant acoustic shadowing (sludge) are present. Can this sludge block the cystic duct or be the nidus for acute cholecystitis? This question has been the source of some debate. It is believed that biliary sludge is echogenic bile that develops due to stasis [10]. The clinical signicance of this nding is unclear because biliary sludge is present in 50% of patients with cholecystitis and commonly present in nursing home patients and those who are fasting or have poor nutrition [10]. Eberle and Rettenmaier [29] found no biliary tract symptoms in a study of 82 patients with sludge, whereas Allen et al [30] reported 2 cases of gangrenous cholecystitis in a study of 97 patients with sludge. Over the course of 6 years, one institution identied 87 patients with biliary sludge without gallstones and 11 were diagnosed with GB disease; interestingly, 5 of the 11 patients had gallstones found after cholecystectomy [31]. In a prospective longitudinal study of 96 patients with sludge, 14 (14.5%) developed gallstones and 6 required cholecystectomy in the subsequent 3 years [32]. Clearly, the nding of biliary sludge alone is not diagnostic; however, it may be a useful secondary nding, especially in the right clinical setting. Intramural/intraluminal gas Although rare, identication of intramural/intraluminal gas in the GB is important because it reects the presence of gas-forming organisms that lead to relatively high mortality due to gangrene of the GB called acute emphysematous cholecystitis. The literature contains many case reports of successful identication of gas in the GB by radiology departments. It is not a necessary part of the ED bedside ultrasound but, given the associated increased mortality, it is worth mentioning and remembering.
669
Intramural gas is seen as an area of high reectivity in the GB wall, with distal reverberations that change position with movement of the patient; intraluminal gas is seen as a dense band of hyper-reective echoes, causing reverberation and shadowing that often obscures the posterior wall of the GB [3336]. These ultrasonographic ndings are a result of the behavior of sound waves against focal areas of gas. Wall, echo, and shadow sign It is dicult to assess a contracted GB for signs of cholecystitis because the GB wall measurements often are falsely high, stones are not as clearly visible, and occasionally, it is dicult to nd the GB at all. There is, however, a nding called the WES sign (WES is an acronym for wall, echo, and shadow) that is a specic sonographic sign of gallstones in a contracted GB. When ultrasonography demonstrates only the anterior GB wall, the echogenicity of the stone, and a shadow, it can be fairly certain that a contracted GB containing a stone has been identied [37,38]. Incidental ndings: phrygian cap, septa, Heisters valves It is useful to know about a few incidental ndings that may be encountered so as not to confuse them with true pathology. The distal segment of the GB will be found to fold over on itself in approximately 4% of the population; this is called a phrygian cap (Fig. 4) and has no clinical signicance [22]. On occasion, there are benign junctional folds called septa that extend from the GB wall and can be seen to traverse the entire GB (complete septa) or only partially (incomplete septa) [22]. It also is important to know about the
Fig. 4. A pharygian cap is a normal variant of gallbladder wall appearance on ultrasonography.
670
existence of Heisters valves because they can be mistaken for an impacted stone at the neck of the GB [22]. Clinical correlation The most important question remains: Can EP-performed ultrasound diagnose cholecystitis? It has been well demonstrated in multiple studies that ED bedside ultrasound aids in ruling in and ruling out the diagnosis of acute cholecystitis. During a 1-year proctoring period of EPs by the medical imaging department at one institution, the accuracy of ED GB ultrasound yielded a sensitivity of 86%, a specicity of 97%, and a positive predictive value of 97% [39]. Rosen et al [9] studied the accuracy of EPs who made the diagnosis of cholecystitis based on the presence of gallstones and a positive sonographic Murphys sign in the rst 3 months after the introduction of an ultrasound machine into the ED; the sensitivity was high (91%), but the specicity was relatively low (66%), leading the investigators to conclude that EPs with limited ultrasound experience should order a conrmatory test before cholecystectomy. Residents also can perform ultrasound of the GB successfully. A prospective study at an inner-city emergency medicine residency program over the course of 2.5 years yielded a sensitivity of 90% and a specicity of 85% in diagnosing acute cholecystitis by ultrasound in the ED [40]. The sensitivities and specicities generated by EP-performed ultrasound to diagnose cholecystitis are comparable to those of formal ultrasound test characteristics found in the radiology literature; sensitivity ranges from 84% to 98% and specicity ranges from 90% to 99% [11,4144]. Making the diagnosis of cholecystitis requires more than simply identifying gallstones. Gallstones alone (or any ultrasonographic sign alone) cannot make the diagnosis of an infected GB. Rosen et al [9] postulated that the low specicity of bedside ultrasound by EPs likely was due to not using secondary ndings. Although EPs have the advantage of knowing the history and having the ability to do a physical examination, it is critical to look for secondary signs of cholecystitis in addition to gallstones and a sonographic Murphys sign. Fig. 5 shows the recommended approach to bedside ultrasound of the GB. Machine matters Although many EDs are obtaining ultrasound machines for use by EPs, the quality of the machines generally is inferior to those used by the radiology department. The reasons for this discrepancy often involve cost and portability/size. Many practices use the Sonosite (Sonosite, Inc., Bothell, Washington) machines that allow quick and easy mobility from
671
Gallstones?
yes no
Sonographic Murphys sign?

yes no
Sonographic Murphys sign?

yes no
Other signs?
yes no yes
Other signs?
no yes
Other signs?
no
High prob. cholecystitis
Moderate prob. cholecystitis
Low prob. cholecystitis Consider other diagnoses also
No cholecystitis
Fig. 5. A suggested approach to using bedside ultrasound to evaluate the gallbladder for possible cholecystitis.
room to room at the expense of image quality. Few studies have compared operator accuracy between two dierent-quality machines. Kendall and Shimp [8] found that sensitivity and specicity for detection of gallstones were higher with the Toshiba (Toshiba America, Inc., New York, New York) 140A (97% and 93%, respectively) ultrasound machine compared with the Toshiba Capasee (92% and 78%, respectively) in their study containing 51 patients with gallstones.
Time and money EP use of bedside GB ultrasound has been shown to decrease patient length of stay in the ED. In comparing discharged patients who had bedside ultrasounds with those who did not, Blaivas et al [7] demonstrated a decrease in length of stay of 10% (30 minutes); similarly, those who presented to the ED after hours had a length of stay shortened by 15% (58 minutes) when they were scanned by an EP. For patients suspected of GB disease, the use of ultrasound likely improves patient ow. Given that formal ultrasound is not available 24 hours per day at most institutions, it has been suggested that ultrasonography in the hands of EP is accurate enough to rule out acute cholecystitis [9,16]. In fact, Durston et al [45] demonstrated that ED ultrasound not only is reasonably accurate but also improves quality of care and decreases cost of care (the cost of having and maintaining an ED ultrasound was oset by the cost of calling in a formal ultrasound technician in the middle of the night). The amount of time required of the EP to perform a biliary ultrasound also is a critical issue. One study demonstrated that most (83%) focused
672
right upper quadrant ultrasounds performed by EPs were completed in a reasonable amount of time (less than 10 minutes), and no scans took longer than 20 minutes [8].
Summary Ultrasound is an extremely valuable tool for the evaluation of GB disease in the ED.
References
[1] Traverso LW. Clinical manifestations and impact of gallstone disease. Am J Surg 1993; 165(4):4059. [2] Warwick DJ, Thompson MH. Six hundred patients with gallstones. Ann R Coll Surg Engl 1992;74(3):21821. [3] Cucchiaro G, Watters CR, Rossitch JC, Meyers WC. Deaths from gallstones. Incidence and associated clinical factors. Ann Surg 1989;209(2):14951. [4] Gracie WA, Ransoho DF. The natural history of silent gallstones: the innocent gallstone is not a myth. N Engl J Med 1982;307(13):798800. [5] Vitetta L, et al. Gallstones at autopsy and cholecystectomy: a comparative study. Aust N Z J Surg 1988;58(7):5618. [6] Hopper KD, et al. The prevalence of asymptomatic gallstones in the general population. Invest Radiol 1991;26(11):93945. [7] Blaivas M, Harwood RA, Lambert MJ. Decreasing length of stay with emergency ultrasound examination of the gallbladder. Acad Emerg Med 1999;6(10):10203. [8] Kendall JL, Shimp RJ. Performance and interpretation of focused right upper quadrant ultrasound by emergency physicians. J Emerg Med 2001;21(1):713. [9] Rosen CL, et al. Ultrasonography by emergency physicians in patients with suspected cholecystitis. Am J Emerg Med 2001;19(1):326. [10] Heller M, Jehle D. Ultrasound in emergency medicine. Philadelphia: W.B. Saunders; 1995. [11] Marton KI, Doubilet P. How to image the gallbladder in suspected cholecystitis. Ann Intern Med 1988;109(9):7229. [12] Cooperberg PL, Gibney RG. Imaging of the gallbladder 1997. Radiology 1987;163(3): 60513. [13] Walker J, Chalmers RT, Allan PL. An audit of ultrasound diagnosis of gallbladder calculi. Br J Radiol 1992;65(775):5814. [14] Ralls PW, et al. Real-time sonography in suspected acute cholecystitis. Prospective evaluation of primary and secondary signs. Radiology 1985;155(3):76771. [15] Babb RR. Acute acalculous cholecystitis. A review. J Clin Gastroenterol 1992;15(3):23841. [16] Jehle D, et al. Emergency department sonography by emergency physicians. Am J Emerg Med 1989;7(6):60511. [17] Bree RL. Further observations on the usefulness of the sonographic Murphy sign in the evaluation of suspected acute cholecystitis. J Clin Ultrasound 1995;23(3):16972. [18] Engel JM, Deitch EA, Sikkema W. Gallbladder wall thickness: sonographic accuracy and relation to disease. AJR Am J Roentgenol 1980;134(5):9079. [19] Sanders RC. The signicance of sonographic gallbladder wall thickening. J Clin Ultrasound 1980;8(2):1436. [20] Simon B, Snoey E. Ultrasound in emergency and ambulatory medicine. St. Louis (MO): Mosby; 1997.
673
[21] Ekberg O, Weiber S. The clinical importance of a thick-walled, tender gall-bladder without stones on ultrasonography. Clin Radiol 1991;44(1):3841. [22] Roy S. Hepatobiliary. In: Ma OJ, Mateer JR, editors. Emergency ultrasound. New York: McGraw-Hill Professionals; 2002. [23] Elyaderani MK, Gabriele OF. Cholecystosonography in detection of acute cholecystitis: the halo signa signicant sonographic nding. South Med J 1983;76(2):17480. [24] Pezzilli R, et al. Ultrasonographic evaluation of the common bile duct in biliary acute pancreatitis patients: comparison with endoscopic retrograde cholangiopancreatography. J Ultrasound Med 1999;18(6):3914. [25] Neoptolemos JP, et al. The urgent diagnosis of gallstones in acute pancreatitis: a prospective study of three methods. Br J Surg 1984;71(3):2303. [26] Perret RS, Sloop GD, Borne JA. Common bile duct measurements in an elderly population. J Ultrasound Med 2000;19(11):72730 [quiz: 731]. [27] Kaude JV. The width of the common bile duct in relation to age and stone disease. An ultrasonographic study. Eur J Radiol 1983;3(2):1157. [28] Wu CC, Ho YH, Chen CY. Eect of aging on common bile duct diameter: a real-time ultrasonographic study. J Clin Ultrasound 1984;12(8):4738. [29] Eberle F, Rettenmaier G. [Gallbladder sludge: a sonographically recognizable stage of lithogenesis]. Z Gastroenterol 1984;22(2):827. [30] Allen B, et al. Sludge is calcium bilirubinate associated with bile stasis. Am J Surg 1981; 141(1):516. [31] Ohara N, Schaefer J. Clinical signicance of biliary sludge. J Clin Gastroenterol 1990; 12(3):2914. [32] Lee SP, Maher K, Nicholls JF. Origin and fate of biliary sludge. Gastroenterology 1988; 94(1):1706. [33] Bloom RA, et al. The ultrasound spectrum of emphysematous cholecystitis. J Clin Ultrasound 1989;17(4):2516. [34] Bloom RA, et al. Shifting intramural gasa new ultrasound sign of emphysematous cholecystitis. J Clin Ultrasound 1984;12(1):402. [35] Blaquiere RM, Dewbury KC. The ultrasound diagnosis of emphysematous cholecystitis. Br J Radiol 1982;55(650):1146. [36] Nemcek AA Jr, et al. The eervescent gallbladder: a sonographic sign of emphysematous cholecystitis. AJR Am J Roentgenol 1988;150(3):5757. [37] Rybicki FJ. The WES sign. Radiology 2000;214(3):8812. [38] MacDonald FR, Cooperberg PL, Cohen MM. The WES triada specic sonographic sign of gallstones in the contracted gallbladder. Gastrointest Radiol 1981;6(1):3941. [39] Schlager D, et al. A prospective study of ultrasonography in the ED by emergency physicians. Am J Emerg Med 1994;12(2):1859. [40] Lanoix R, et al. A preliminary evaluation of emergency ultrasound in the setting of an emergency medicine training program. Am J Emerg Med 2000;18(1):415. [41] Berk RN, et al. The radiological diagnosis of gallbladder disease. An imaging symposium. Radiology 1981;141(1):4956. [42] Detwiler RP, Kim DS, Longerbeam JK. Ultrasonography and oral cholecystography: a comparison of their use in the diagnosis of gallbladder disease. Arch Surg 1980;115(9): 10968. [43] McAvoy JM, et al. Role of ultrasonography in the primary diagnosis of cholelithiasis: an analysis of fty cases. Am J Surg 1978;136(3):30912. [44] McKay AJ, et al. A prospective study of the clinical value and accuracy of grey scale ultrasound in detecting gallstones. Br J Surg 1978;65(5):3303. [45] Durston W, et al. Comparison of quality and cost-eectiveness in the evaluation of symptomatic cholelithiasis with dierent approaches to ultrasound availability in the ED. Am J Emerg Med 2001;19(4):2609.
Ultrasound detection of abdominal aortic aneurysm

Adam Z. Barkin, MD, Carlo L. Rosen, MD*
Beth Israel Deaconess Medical Center, Harvard Aliated Emergency Medicine Residency, Harvard Medical School, West Clinical Center 2, One Deaconess Road, Boston, MA 02215, USA
One of the most useful and potentially life-saving applications of ultrasound (US) is the rapid diagnosis of an abdominal aortic aneurysm (AAA). The incidence of AAA (aortic diameter greater than 3.0 cm) is increasing as the population ages [1]. Approximately 15,000 people die of an AAA each year in the United States [2]. Rupture of an AAA in the out-ofhospital setting may carry a mortality rate of 90% or higher [3,4], yet patients with rupture who survive to the operating room have a 50% mortality [3,5]. Delay in diagnosis of an AAA can be fatal, whereas early recognition and surgical intervention can prevent morbidity and mortality. Plummer et al [6] demonstrated improved outcome and decreased time to diagnosis in patients with ruptured AAAs when bedside US was used (5.4 minutes) compared with traditional diagnostic modalities (83 minutes). In this study, the time to disposition for patients requiring operative intervention decreased from 90 to 12 minutes with the use of bedside US. Detection of AAA is one of the uses of bedside US that makes a critical dierence in patient care in the ED. Similar to the unstable trauma patient, the hypotensive patient with abdominal or ank pain benets tremendously from this technique. Furthermore, bedside US allows emergency physicians to make the diagnosis of AAA in normotensive patients with ank or abdominal pain.
Clinical presentation of abdominal aortic aneurysm Patients with AAAs arrive in the ED with a myriad of chief complaints and clinical presentations. Symptoms are caused by rupture or rapid
* Corresponding author. E-mail address: crosen2@bidmc.harvard.edu (C.L. Rosen). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.009
676
A.Z. Barkin, C.L. Rosen / Emerg Med Clin N Am 22 (2004) 675682
expansion of the aneurysm. The classic triad of hypotension, back pain, and pulsatile abdominal mass as a presentation for a ruptured AAA may be present only 50% of the time [7]. Misdiagnoses are common and include gastrointestinal bleed, nephrolithiasis, and diverticulitis. Patients can present with hypovolemia and shock; however, symptoms vary because retroperitoneal rupture may be contained within the retroperitoneal space. The resulting tamponade eect will limit the extent of the bleeding, and these patients may be normotensive [3]. Signs and symptoms of AAA include pain in the back, abdomen, chest, ank, buttock, leg, groin, or scrotum. Patients also may present with syncope [1]. Symptoms may be present for days to weeks before patients present to an ED. Compression of bowel loops may cause a functional bowel obstruction and induce nausea and vomiting. Retroperitoneal hematomas also may compress the ureter, causing obstruction and hydronephrosis. When the AAA ruptures, it will rupture into the retroperitoneal or the intraperitoneal cavity. Intraperitoneal rupture is a rapidly fatal event, and these patients usually will not make it to the ED alive, whereas retroperitoneal rupture is more likely in patients arriving in the ED. Up to 30% of the time, the diagnosis of AAA initially is missed. Marston et al [8] reported on 152 patients with a ruptured AAA and found that the most common misdiagnosis was renal colic (24%), followed by diverticulitis (13%) and gastrointestinal bleeding (13%). Other misdiagnoses included acute myocardial infarction, back pain, and sepsis. The high rate of misdiagnosis occurs because over 75% of patients with an AAA are not aware that they have this condition when they present with complications [1] and because of the vague symptoms and normal hemodynamic status that many patients have. Particular attention must be focused on elderly patients who frequently present with nonspecic complaints.
Making the diagnosis Physical examination is notoriously unreliable in detecting AAAs. In one study by Fink et al [9], the overall sensitivity and specicity of detecting an AAA by physical examination were 68% and 75%, respectively. The accuracy of the physical examination improves with increasing size of the AAA and smaller abdominal girth but still is not satisfactory in making the diagnosis of a life-threatening condition. A meta-analysis by Lederle and Simel [10] also concluded that physical examination and abdominal palpation are not enough to exclude the diagnosis of AAA. In most cases, an imaging study is required to denitively make or exclude the diagnosis of AAA. The primary diagnostic modalities used by emergency physicians in patients with suspected AAA are CT and US. The advantages of CT are that it helps to dene the entire extent of an AAA, provides information about thrombus formation within the aneurysm, and provides views of the
677
retroperitoneum in cases of retroperitoneal leakage [11]. CT is very accurate for detecting the size of the aorta and the rupture; however, it is not always available, which necessitates patient transport out of the safe environment of the ED.
Ultrasound detection of abdominal aortic aneurysm The primary benet of US is the opportunity to perform an accurate, sensitive, and rapid examination of the abdominal aorta at the bedside. The other key advantage is that it can be used as part of an abdominal US examination to evaluate stable patients with ank or abdominal pain to exclude other diagnoses such as cholecystitis and renal colic. Other imaging modalities have signicant limitations in making the diagnosis of AAA in an unstable patient. It may not be safe to send a patient to the ED radiology suite where monitoring and resuscitation are dicult. Multiple studies have demonstrated the sensitivity of US for detecting an AAA to be 100%. It is accurate for determining the size of the aorta [1114]. Several studies have demonstrated that emergency physicians with limited formal training in US can diagnose AAAs. An early study in 1989 by Jehle et al [15] reported a small series of patients with AAA successfully diagnosed by EP-performed US and later conrmed by surgical intervention or by other imaging modalities. Kuhn et al [14] studied a convenience sample of 68 patients greater than 50 years of age with abdominal pain. The investigators demonstrated 100% sensitivity, specicity, and positive and negative predictive values in the US evaluation of aortas by emergency medicine faculty and senior residents after a 3-day US course. It was determined that emergency bedside US would have improved the care of approximately two thirds of patients enrolled without any adverse eects. Schlager et al [13] also demonstrated 100% sensitivity, specicity, and negative and positive predictive values in the detection of AAA in a study of 167 patients. In a recent prospective, consecutive study, Tayal et al [16] reported on 114 patients presenting to an ED over a 2-year period and demonstrated that US performed by emergency medicine faculty and senior residents was 100% sensitive for AAA compared with US interpreted by a radiologist, CT scan, MRI, and operative ndings. The investigators reported a specicity of 96%, a positive predictive value of 95%, and a negative predictive value of 100%. Although these studies included a small number of patients, the data suggest that emergency physicians can detect AAA accurately with bedside US. Because the test is operator dependent and may be technically dicult in certain patient groups, caution should be used in interpreting a negative result. In patients with a concerning presentation for AAA, a negative US should be conrmed by CT or radiology-interpreted US. In a study undertaken to train internal medicine residents in the evaluation of AAA by US, Bailey et al [17] showed that physicians with
678
less than 3 hours of formal training in US can eectively visualize the aorta and determine whether an aneurysm is present. Despite the high sensitivity, prior studies in the radiology literature noted that some ruptured aneurysms will be impossible to see because of obesity, guarding, bowel gas, or loss of integrity of the aneurysm. US can provide important information for the diagnosis and treatment of critically ill patients with AAA. Hendrickson et al [18] reported a case of pulseless electrical activity presumably caused by AAA rupture and diagnosed by ED US. When a patient presents with abdominal or back pain and has hypotension or shock of unknown etiology, US should be used as the primary diagnostic tool. Evaluation of the aorta with US can provide a rapid diagnosis of an AAA and expedite surgical consultation and management. The indications for US of the abdominal aorta should include patients who are suspected of having an AAA who have abdominal, ank, or back pain or syncope of unknown origin. Patients with unexplained hypotension or pulseless electrical activity also should undergo an US examination of the aorta. In patients with pulseless electrical activity, cardiac US should be performed for cardiac activity in addition to an examination of Morisons pouch and the aorta to detect treatable causes.
Technical considerations and ndings US of the abdominal aorta should be performed with a 2.5- to 3.5-MHz transducer to maintain adequate resolution and penetration. The anteriorto-posterior diameter in the transverse plane is the most accurate measurement of the aorta. This measurement should be made from outer wall to outer wall. The entire aorta from the xiphoid process to the bifurcation of the iliac vessels should be visualized and the aorta measured at the maximum diameter. It is critical to image the entire extent of the aorta down to the bifurcation because 95% of AAAs are infrarenal [19]. It may be easier to follow the aorta down to its bifurcation with the probe in the longitudinal plane. Measurements also should be taken in the transverse plane by rotating the probe 90( so as to not underestimate the size of the aorta. As the operator moves the probe from the epigastric area caudad, in the longitudinal plane, the celiac artery comes o rst and then the superior mesenteric artery will be visualized parallel to the aorta as it branches o of the aorta. The inferior vena cava may be confused with the aorta. The inferior vena cava is located to the patients right of the aorta, is thin walled, and will collapse easily when the patient takes a deep inspiration or snis (the sni test). During the examination, gentle pressure may be needed to move bowel gas out of the way. The aorta will be visualized above the hyperechoic spine in the transverse view. An alternate approach is to visualize the aorta throughout the right anterior axillary line in the subcostal
679
area with the patient in the left lateral decubitus position. This view takes advantage of the liver as an acoustic window. In a patient in cardiac arrest with suspected AAA, a view of Morisons pouch also should be obtained to document the presence of intraperitoneal rupture. The US ndings consistent with an AAA are an enlarged abdominal aorta greater than 3 cm (or an increase in the normal diameter of greater than 50%) or focal dilatation of the aorta. The incidence of rupture of an aneurysm does not become signicant until at least 4 cm [3]. The risk of rupture is less than 2% when the diameter is less than 4 cm [20,21]. Thus, a clinically signicant AAA likely will be greater than 4 cm. US is insensitive for retroperitoneal bleeding, so it should not be used to detect the presence of rupture. The sensitivity of US for detecting extraluminal blood is as low as 4% [22]. An intramural thrombus will appear as an echoic substance within the lumen of the aorta (Fig. 1). In patients with intraperitoneal rupture, uid will be seen in Morisons pouch, the splenorenal space, or on the suprapubic view; however, this is rarely clinically useful because these patients usually are in cardiac arrest. There are limitations to the use of US for AAA detection in certain types of patients. A tortuous aorta may be dicult to visualize and track with the US probe, thus yielding an inaccurate measurement. Patient characteristics
Fig. 1. Transverse image of AAA measuring 6.14 cm with thrombus.
680
such as obesity and bowel gas may not allow an adequate window for visualization [1]. Furthermore, aortic branch aneurysms conned to the iliac vessels or renal arteries or within the liver may not be visualized on US [1]. If aneurysms in these locations are suspected, then CT may be warranted to provide additional diagnostic information. In the sagittal plane, US may underestimate the size of the aorta by the cylinder tangent eectthe maximum diameter is not captured by the US probe [23]. If the US beam is o center, then the diameter measured will not be the maximal diameter. The nal limitation of US is that it is accurate for detecting the size of the aorta but will not be able to detect the presence or absence of rupture. The primary use of US is to determine the diameter of the aorta and whether or not an aneurysm exists.
Ultrasound as a screening test for abdominal aortic aneurysm In the future, US screening for asymptomatic AAAs may become the standard of care. This practice, which has been studied in the primary care setting, would result in early detection and elective repair before rupture. Ashton et al [24] published a large United Kingdom study involving over 67,000 men aged 65 to 74 years who were randomized to US screening or to no US screening. Patients with an AAA of 3 cm or greater received a followup US. The investigators reported a 42% risk reduction in mortality due to AAA in men who were screened versus those who were not screened. The mortality rates of elective and emergency surgery were 6% and 37%, respectively. Salen et al [25] studied the feasibility of US screening for AAA in an ED setting. Hemodynamically stable men greater than 65 years of age who presented without symptoms consistent with an AAA were screened. If the aorta was measured to be greater than 3 cm, then the patient was referred for formal US. Among the 103 patients enrolled, 6 had conrmed AAA. The positive predictive value of screening US compared with radiologyinterpreted US was 75%. A limitation of this study was the lack of follow-up for patients identied with a normal aorta on the screening examination. These data are promising and suggest that US screening for AAA may become the standard of care in the primary care setting and in the ED in the future.
Summary US is becoming a widely used diagnostic and therapeutic tool in emergency medicine training and practice. Its use in detecting an AAA is ecient and practical and can occur concurrently with resuscitation. US of the abdominal aorta has been shown to be highly accurate for the detection of AAA and to decrease the time to operative repair of ruptured AAA. The emergency
681
physician should consider using US to search for an AAA in any patient presenting with abdominal or back pain or with shock of unknown etiology.
References
[1] Rose WM III, Ernst CB. Abdominal aortic aneurysm. Comp Ther 1995;21(7):33943. [2] Minino AM, Arias E, Kochanek KD, Murphy SL, Smith BL. Deaths: nal data for 2000. Natl Vital Stat Rep 2002;50(15):1119. [3] Ernst CB. Abdominal aortic aneurysm. N Engl J Med 1993;328(16):116772. [4] Johansson G, Swedenborg J. Ruptured abdominal aortic aneurysms: a study of incidence and mortality. Br J Surg 1986;73(2):1013. [5] Greenhalgh RM. Prognosis of abdominal aortic aneurysm. BMJ 1990;301(6744):136. [6] Plummer D, Clinton J, Matthew B. Emergency department ultrasound improves time to diagnosis and survival in ruptured abdominal aortic aneurysm [abstract]. Acad Emerg Med 1998;5:417. [7] Rohrer MJ, Cutler BS, Wheeler HB. Long-term survival and quality of life following ruptured abdominal aneurysm. Arch Surg 1988;123(10):12137. [8] Marston WA, Ahlquist R, Johnson G Jr, Meyer AA. Misdianosis of ruptured abdominal aortic aneurysms. J Vasc Surg 1992;16(1):1722. [9] Fink HA, Lederle FA, Roth CS, Bowles CA, Nelson DB, Haas MA. The accuracy of physical examination to detect abdominal aortic aneurysm. Arch Intern Med 2000;160(6): 8336. [10] Lederle FA, Simel DL. The rational clinical examination. Does this patient have an abdominal aortic aneurysm. JAMA 1999;281(1):7782. [11] LaRoy LL, Cormier PJ, Matalon TA, Patel SK, Turner DA, Silver B. Imaging of abdominal aortic aneurysms. Am J Roentgenol 1989;152(4):78592. [12] Maloney JD, Pairolero PC, Smith SF, Hattery RR, Brakke DM, Spittell JA Jr. Ultrasound evaluation of abdominal aortic aneurysms. Circulation 1977;56(3 Suppl 2):II805. [13] Schlager D, Lazzareschi G, Whitten D, Sanders AB. A prospective study of ultrasonography in the ED by emergency physicians. Am J Emerg Med 1994;12:1859. [14] Kuhn M, Bonnin RLL, Davey MJ, Rowland JL, Langlois SLP. Emergency department ultrasound scanning for abdominal aortic aneurysm: accessible, accurate and advantageous. Ann Emerg Med 2000;36(3):21923. [15] Jehle D, Davis E, Evans T, Harchelroad F, Martin M, Zaiser K, et al. Emergency department sonography by emergency physicians. Am J Emerg Med 1989;7(6):60511. [16] Tayal VS, Graf CD, Gibbs MA. Prospective study of accuracy and outcome of emergency ultrasound for abdominal aortic aneurysm over two years. Acad Emerg Med 2003;10(8): 86771. [17] Bailey RP, Ault M, Greengold NL, Rosendahl T, Cossman D. Ultrasonography performed by primary care residents for abdominal aortic aneurysm screening. J Gen Intern Med 2001;16(12):8459. [18] Hendrickson RG, Dean AJ, Costantino TG. A novel use of ultrasound in pulseless electrical activity: the diagnosis of an acute abdominal aortic aneurysm rupture. J Emerg Med 2001;21(2):1414. [19] Isiko MB, Hill MC. Sonography of the renal arteries: left lateral decubitus position. Am J Roentgenol 1980;134(6):11779. [20] Cronenwett JL, Murphy TF, Zelenock GB, Whitehouse WM Jr, Lindenauer SM, Graham LM, et al. Actuarial analysis of variables associated with rupture of small abdominal aortic aneurysms. Surgery 1985;98(3):47283. [21] Ouriel K, Green RM, Donayre C, Shortell CK, Elliot J, DeWeese JA. An evaluation of new methods of expressing aortic aneurysm size: relationship to rupture. J Vasc Surg 1992; 15(1):128.
682
[22] Shuman WP, Hastrup W Jr, Kohler TR, Nyberg DA, Wang KY, Vincent LM, et al. Suspected leaking abdominal aortic aneurysm: use of sonography in the emergency room. Radiology 1988;168(1):1179. [23] Plummer D. Abdominal aortic aneurysm. In: Emergency ultrasound. New York: McGraw-Hill; 2003. p. 12942. [24] Ashton HA, Buxton MJ, Day NE, Kim LG, Marteau TM, Scott RA, et al. Multicentre Aneurysm Screening Study Group The Multicentre Aneurysm Screening Study (MASS) into the eect of abdominal aortic aneurysm screening on mortality in men: a randomized control trial. Lancet 2002;360(9345):15319. [25] Salen P, Melanson S, Buro D. ED screening to identify abdominal aortic aneurysms in asymptomatic geriatric patients. Am J Emerg Med 2003;21(2):1335.
Gynecologic ultrasound in emergency medicine

Michael J. Lambert, MD, RDMS*, Martha Villa, MD
7435 West Talcott, Chicago, IL 60631, USA
Ultrasound, without question, is the gold standard imaging modality in the pregnant patient presenting to the emergency department (ED) with abdominal pain or vaginal bleeding. Its ability to dierentiate complications of early pregnancy and assist in diagnosing possible late pregnancy problems are documented in articles elsewhere in this issue and throughout the literature. This goal of this article, however, is to focus on the use of ultrasound in the nonpregnant female patient who presents to the ED with acute abdominopelvic pain.1 Abdominal pain is a very common presenting problem in everyday practice. As a whole, these patients create management dilemmas for most emergency physicians (EPs) who provide their care. This quandary usually stems from the inability of the physical examination to dierentiate a potential surgical problem (ie, appendicitis, ovarian torsion, pelvic abscess) from a nonsurgical etiology (ie, ovarian cyst, pelvic inammatory disease [PID], diverticulitis) reliably. This dierentiation has signicant implications, in that most patients with nonsurgical causes of pain frequently can be managed as outpatients. Historically, many EPs have enlisted the help of colleagues from surgery and gynecology to help resolve this predicament. It is unfortunate that consultations with these colleagues characteristically leave EPs with the same impasse and more calls to make because their clinical examinations have the same shortcomings. Help beyond history and physical examination is needed to manage these patients. Faced with this dilemma, EPs frequently use some type of imaging modality to help direct the management of these patients and to stay the surgeons hand. Ultrasound currently is the best initial imaging modality for
* Corresponding author. Resurrection Medical Center 7435 W. Talcott Chicago, IL 60631. 1 Throughout this article, any reference to abdominal pain will serve as the abbreviated form of abdominopelvic pain. E-mail address: mlambert62@mac.com 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.016
684
M.J. Lambert, M. Villa / Emerg Med Clin N Am 22 (2004) 683696
the gynecologic patient presenting to the ED with acute abdominal pain. It is quick, accurate, cost eective, and able to be performed at the patients bedside. Emergency ultrasound indications Unlike the obstetric patient, the gynecologic indications for ultrasound in the nonpregnant patient presenting to the ED are limited. The gynecologic disorders for which it is believed that ultrasound plays a crucial role in the emergent diagnosis are limited to acute ovarian torsion, pelvic abscesses, and ruptured ovarian cyst. Ovarian torsion and pelvic abscess are considered emergent conditions, and their ultrasound ndings and criteria are covered in detail later. Ovarian cysts also are covered due to the frequency with which this entity presents as abdominal pain to the ED. Although this list might seem scant, these are not emergent indications in the medical or reproductive sense as much as they are urgent indications. Likewise, other imaging modalities such as CT or procedures such as laparoscopy may be more available or appropriate. Ovarian cysts Denition Ovarian cysts are by far the most common gynecologic mass to occur in the pelvis [1,2]. The overwhelming majority is physiologic in origin, unilocular, and non-neoplastic. The term ovarian cyst frequently is misused to describe normal ovarian follicles or corpus lutea seen by ultrasound. Although there is not a specic size criterion to accurately dene the dierence between an ovarian cyst and normal physiologic structures within the ovary, there are measurements to use as guidelines. The rational for these measurements is based on ultrasound ndings during a normal menstrual cycle in which it is common for cystic structures to develop within the ovary [3]. During the course of this menstrual cycle (Fig. 1), a mature graaan follicle or the corpus luteum ordinarily may reach a size of up to 2.5 cm. The term ovarian cyst, therefore, should be reserved to describe cystic structures that attain a size greater than 2.5 cm in length, height, or width. The most common type of ovarian cysts in women of reproductive age is a follicular cyst [2,4]. Most follicular cysts are produced by changes in hormone levels that happen in the course of the menstrual cycle with the production and release of eggs from the ovaries. Many experts theorize that the occurrence of a follicular cyst is merely a mishap of the normal ovarian cycle in which there is failure of the follicle to rupture or regress [5]. Although follicular cysts are the most frequent type of unilocular cysts found in women of reproductive age, several dierent types of ovarian cysts are identiable by pelvic ultrasound within and adjacent to the ovary (Box 1) [4].
685
LUTEAL PHASE
Corpus Luteum
OVULATION
Germinal epithelium Primodial Follicles
Mature Graafian Follicle
FOLLICULAR PHASE
Fig. 1. Diagram of the normal menstrual cycle.
Clinical presentation Ovarian cysts can develop at any time throughout a womans life. They usually are diagnosed in one of two ways: a routine gynecologic examination reveals a palpable mass in an asymptomatic patient or the patient complains of abdominal pain that leads their physician to search for a cause. Most symptomatic women diagnosed with an ovarian cyst who present to the ED are in their reproductive years. The most common presenting symptom is that of abdominal pain. The onset of pain typically is sudden (ie, ruptured ovarian cyst) or subacute (large ovarian cyst). Rupture of an ovarian cyst
Box 1. Types of ovarian cysts Simple Functional Parovarian Dermoid Endometrial Serous cystadenoma Mucinous cystadenoma Hydrosalpinx Other benign Borderline Cancerous
686
may cause signicant peritonitis, hypotension, abdominal distention, and bleeding. Ultrasound is helpful in hemodynamically stable patients to identify an adenexal mass and establish the presence of intra-abdominal uid [6,7]. Ultrasound ndings Ovarian cysts usually are identied easily by transabdominal or endovaginal ultrasound. Both probes oer advantages in evaluation of possible ovarian pathology. The transabdominal approach allows a better overall view of the pelvis and the abdomen. This overall view is helpful for visualizing very large ovarian masses and subsequent complications of their shear size (ie, hydronephrosis secondary to obstructed ureter) or free uid in the peritoneal cavity (ie, ruptured ovarian cyst). The benet of the endovaginal probe is its proximity to the ovaries. By allowing the footprint of the probe to be close to the area of interest, clinicians can take advantage of a high-frequency probe. This probe, in turn, allows superior imaging capabilities by virtue of its enhanced axial and lateral resolution capabilities. To understand the sonographic morphology of ovarian cysts, it is helpful to compare their features with those of a normal adult ovary. Typically, the normal-sized adult ovary averages 2.5 to 5 cm long, 1.5 to 3 cm wide, and 0.6 to 1.5 cm thick [8]. If the follicular phase is used as an example, several follicles usually are visible within the ovarian tissue by ultrasound (Fig. 2). Ovarian cysts classically are rounded, anechoic, smoothly marginated, unilocular, and displace adjacent ovarian tissue to the periphery (Fig. 3).
Fig. 2. Endovaginal scan in an oblique view through the left ovary. This still image reveals a normal ovary anterior and medial to a branch of the left iliac vein. The large central anechoic saclike structure is a mature follicle. Several peripheral follicles are visible at the perimeter of the ovary (white arrowheads).
687
Fig. 3. Oblique view through the right adnexa reveals a 3.2-cm ovarian cyst. The cyst is well circumscribed within the ovary. Notice the stromal tissue (asterisks) and follicles (white arrow) at the periphery. Posterior acoustic enhancement is evident deep to the cyst (between black arrows).
They range in size from 2.5 cm to over 15 cm in length, width, and height. These cysts typically are low-attenuating tissues with good through-sound transmission. When compared with the adjacent tissue deep to the actual cyst, posterior enhancement (increased echogenicity) usually can be demonstrated (Fig. 4). Depending on the size of the cysts, the surrounding ovarian stromal tissue may not be recognizable by ultrasound. Hemorrhage into an ovarian cyst is a common complication. Classically, fresh blood appears anechoic on ultrasound. This appearance progresses to mixed echogenicity in the subacute phase and then returns to being anechoic. These changeable sonographic ndings are related to the temporal sequence of clot formation, lysis, and retraction [9]. This variability also can be attributed to the technique of the sonologist. Many factors (gain,
Fig. 4. Posterior acoustic enhancement is evident deep to the cyst (between black arrows).
688
frequency, contrast, artifacts, and so forth) obviously can aect the appearance of blood on ultrasound. The fundamental sonographic features of a hemorrhagic ovarian cyst are increased through-sound transmission. (Fig. 5) This feature distinguishes a cystic mass from a solid mass in almost all cases. The most common appearance is a heterogeneous mass that predominantly is anechoic [10]. Other sonographic features include thick rim, septations, and associated uid in the cul-de-sac [11]. Ovarian torsion Clinical presentation Ovarian torsion represents the fth most common surgical gynecologic emergency. It typically occurs in the early reproductive years with regular development of a corpus luteum in the second half of the menstrual cycle. Fertile women especially are susceptible because of the regular development of a corpus luteum in the second half of the menstrual cycle. Twenty percent of all ovarian torsion cases occur in pregnant women, and a mass is discovered in half of these cases [12]. The mass is believed to be a fulcrum around which the ovary or entire adnexa revolves. Literature reviews consistently detail how dicult and subtle the diagnosis of ovarian torsion can be. The classic presentation begins with severe, sharp pain in the lower abdomen that worsens intermittently over several hours. Case reports and large-scale reviews, however, show that the presentation usually is more innocuous. In one 15-year review from two urban teaching hospitals, Houry and Abbott [13] documented ndings associated with surgically proved torsion in 87 women. The average duration of pain lasted from several hours to weeks. Most patients presented
Fig. 5. Endovaginal scan through the ovary revealing a heterogeneous ovarian mass that is predominately anechoic. Note the increased through-sound transmission.
689
with minimal ndings on abdominal palpation. Only 3 had documented peritoneal signs. A small minority had a white blood cell count greater than 15,000. Most were afebrile and without nausea or vomiting. Perhaps most compelling is the fact that the diagnosis of ovarian torsion was considered in the admitting dierential diagnosis in less than half of the patients (47%). The investigators commented in their discussion, This paucity of objective ndings may be a clue that ovarian torsion should be considered [13]. Ultrasound ndings Ultrasound is considered the imaging modality of choice in the diagnosis of ovarian torsion. Visualization and interpretation of pelvic pathology has been an area of increasing interest since the advent of endovaginal sonography in the mid 1980s. Characterization of ovarian torsion in grayscale imaging and on duplex Doppler and color ow Doppler imaging is discussed in the following sections. The most common sonographic nding in torsion is ovarian enlargement. Normally, the ovary measures 2 3 cm. In torsion, interstitial pressures lead to impaired venous and lymphatic drainage and, subsequently, an increase in the size of the ovary. The ovary also often demonstrates large, peripheral follicles. Although uncommon, visualization of both ovaries in a longitudinal plane likewise is suspicious for torsion (Fig. 6). As discussed previously, ovarian masses often are associated with torsion. These masses frequently are non-neoplastic or hemorrhagic cysts [14]. On ultrasound, the masses have a variable appearance, ranging from cystic or solid to having a complex pattern, with debris and septations. When a cyst is present, it sometimes has a thickened wall. The use of duplex and color Doppler sonography is a logical application of ultrasound technology in cases of possible torsion. These sophisticated modalities were developed specically to detect the presence and direction of arterial and venous blood ow to organs. Application of Doppler and its interpretation in ovarian imaging, however, has been called confusing and inconsistent [12]. Indeed, numerous studies and reports document the presence of blood ow in surgically proved cases of ovarian torsion [13,1519]. Doppler examination of ovarian vascularity is a challenge for several reasons. First, ovaries have a dual blood supply. The adnexal branch of the uterine artery and the ovarian artery feeds them. This vascular redundancy might contribute to the perception of arterial ow by the sonographer. Second, the clinical presentation sometimes includes an intermittent quality of pain. In these cases, the ovary possibly is torsing and detorsing. If a scan is performed while the ovary is in a normal position, then the diagnosis will be more dicult to make. Third, the presence of a large mass or intraovarian hemorrhage can complicate thorough Doppler interrogation. From a technical perspective, sonographers should be mindful of settings for wall lter, color sensitivity, and scale [20]. Setting the wall lters and
690
Fig. 6. (A) Transabdominal view of two large cystic masses in a longitudinal plane. LO, left ovary; RO, right ovary. (B) Endovaginal view of the same patient with both ovaries visible in a longitudinal plane. FF, free uid; LO, left ovary; RP, right ovary; UT, uterus.
color velocity scale setting too high or the color gain too low can lead to false-positive diagnosis for torsion. One researcher specically cautions that false-positive diagnoses can result from too high a lter pulse repetition frequency setting or incorrect gray-scale priority settings [15]. Although Doppler examination admittedly is challenging, the EP still should try to characterize ow. Several guidelines can be helpful. Vascular interrogation always should take place across multiple sonographic planes because this will increase the chances of detecting ow, if present. Venous ow is compromised rst due to the fact that the ovary represents a low pressure/ow organ. Detection of abnormal venous outow may be even more important than loss of both arterial and venous ow in early detection of torsion [12]. Color ow Doppler imaging also has a role in cases of diagnosing ovarian torsion. Specically, it has been studied as a means for predicting ovarian viability. Researchers in one small study found an absence of central ow in nonviable ovaries. In viable ovaries, however, some central venous ow was depicted [15].
691
Another nding of note is the status of the vascular pedicle. Ultrasound literature describes the identication of a twisted vascular pedicle as being suggestive of torsion. In a retrospective study by Lee et al [17], the nding of a twisted pedicle had a diagnostic accuracy of 88% for torsion. These investigators identied the torsed pedicle by nding coiled vessels in the rounded or elliptical mass of the pedicle. They referred to this as the whirlpool sign (Fig. 7). As a nal technical matter, one might consider transrectal scanning if an endovaginal study is dicult or impossible. These situations might arise in cases of vaginal atresia or in patients with severe pain due to vaginal scarring from radiation for treatment of cervical cancer. Transrectal scanning was found to be superior to transabdominal scanning in a study of 31 patients. With proper patient selection and counseling, researchers found that transrectal scanning was completed without signicant patient discomfort [21].
Tubo-ovarian abscess Tubo-ovarian abscess is not a disease unto itself but rather a nding along the spectrum of PID. Therefore, the various stages of PID are reviewed to appreciate the sonographic features of tubo-ovarian abscess. PID is an infection of the upper genital tract. A combination of clinical examination and laboratory data establishes the diagnoses of PID. The Centers for Disease Control and Prevention divides the diagnosis of PID into major and minor criteria (Box 2). All three major criteria and one minor criterion are needed to establish the diagnosis of PID. The spectrum of clinical ndings of PID is divided into four distinct stages, according to the International Disease Society for Obstetrics and GynecologyUSA (Box 3) [22]. Although this staging provides a scale of reference for severity of PID, its utility as a management tool in the ED is limited. The key question facing the EP is whether or not this patient should be treated as an outpatient or an inpatient. Short of laparoscopy, pelvic ultrasound is the gold standard in delineating answers regarding management and disposition. Pelvic ultrasound has a sensitivity of 93% and a specicity of 98% in the diagnosis of tubo-ovarian abscess [22]. Although 60% to 80% of tubo-ovarian abscesses resolve with medical management [23], consultation with a gynecologist regarding the possible need for surgical management is warranted. Familiarity with key sonographic features of PID and tubo-ovarian abscess can help expedite diagnosis. PID demonstrates multiple ndings on ultrasound. The endometrium can show thickening and heterogeneity. As pus or mucous lls the fallopian tubes, they become distended. In addition to seeing an increase in total wall thickness, the cogwheel sign might be visualized in cross-section (Fig. 8).
692
Fig. 7. (A) Ovarian mass. This enlarged ovary was also torsed. (B) Whirlpool sign. Endovaginal scan of adnexal mass with color ow Doppler imaging. The vascular whirlpool sign of the pedicle is suggestive of torsion. (C) Adnexal mass. Same scan of adnexal mass with addition of spectral Doppler to evaluate blood ow velocities (Figures courtesy of Dr. Arthur Fleischer, Vanderbilt Medical Center).
693
Box 2. Pelvic inammatory disease criteria Major Abdominal pain Cervical motion tenderness Adnexal tenderness Minor Fever greater than 101(F Abnormal cervical/vaginal discharge Elevated Erythrocyte sedimentation rate and/or C-reactive protein Positive cervical cultures with Neisseria gonorrhoea or Chlamydia trachomatis
This sign occurs when the center of the tube displays relatively anechoic uid in association with prominent inner folds of the fallopian tubes walls. A study by Timor-Tritsch et al [24] showed that the cogwheel sign was present in 86% of patients with acute disease. The sonographic nding of an incomplete septa can also be visualized when the tubes became occluded and, subsequently convoluted. This nding was present in 92% of study participants who had acute and chronic inammation. Therefore, although helpful, the sonographic ndings of an incomplete septa cannot dierentiate whether the inammatory process is acute or chronic. Finally, ovaries are usually enlarged, ill-dened, and immediately adjacent to the uterus [25]. The above descriptions occur in ultrasound examinations in which a distinction can be made between ovaries and fallopian tubes. While the ovaries and fallopian tubes are easily discernable anatomically, their borders are less easily dened by ultrasound. Sonographically they are therefore commonly referred to as a tubo-ovarian complex [24]. In tubo-ovarian abscess, however, pus-lled, edematous, inamed tissue contributes to the deterioration of the normal boundaries between tissues. Bacteria can gain entry into the ovary when the corpus luteum has ruptured. Ultrasound shows a large adnexal mass, with poor dierentiation between structures, often in association with free uid. Changes related to infection and
Box 3. Pelvic inammatory disease stage ndings Stage 1: PID, as dened in Box 1 Stage 2: PID with peritonitis Stage 3: presence of mass on examination or documented on ultrasound Stage 4: abscess rupture
694
Fig. 8. Cogwheel sign. Note the anechoic uid in association with prominent inner folds of the fallopian tubes wall (Figure Courtesy of Dr. I.E. Timor-Tritsch, New York University Medical Center).
inammation often aect one adnexal structure rst; therefore, imaging of both adnexa may show them to be out of sync. Ultrasound pitfalls 1. Bladder: when scanning the ovaries from an endovaginal approach, a lled bladder can obscure the area of interest, cause side lobe artifacts, and diminish the quality of the ultrasound images. From a transabdominal approach, an empty bladder creates a potential space for bowel to ll the anterior cul-de-sac, thus obliterating the acoustic window. 2. Bowel: from an endovaginal approach, bowel interposed between the ovary and the tip of the endovaginal probe obscures image quality. Gentle pressure applied toward the ovary can displace this interfering bowel. 3. Ovarian vascularity: the presence of spectral, color ow, or power Doppler within ovarian tissue cannot rule out ovarian torsion. 4. Blood: fresh blood typically is anechoic sonographically. Pooled blood or stagnant blood that has had time to coagulate, however, often appears with internal echoes (even homogeneous) and can be confused with a solid mass. Summary The true value of ultrasound in acute abdominal pain lies in its ability to detect gynecologic disorders and eectively rule out other causes of acute abdominal pain that require surgical repair. Although the emergent gynecologic indications discussed in this article are few in number, this
695
does not suggest that the nonpregnant patient presenting to the ED with abdominal pain should not receive an ultrasound examination. On the contrary, the author believes that in a perfect world, ultrasound should be the initial imaging study in most of these patients. The reality is that it is dicult to convince radiology colleagues to call in a sonologist in the middle of the night for any indication other than ovarian torsion when CT scans can diagnose ovarian cysts and tubo-ovarian abscess. As was pointed out in the section on ovarian torsion, even adequate ovarian blood ow does not rule out this diagnosis. Ideally, an ultrasound of the pelvis could be undertaken at the time of the pelvic examination, adding as little as 5 to 10 minutes. If a gynecologic disorder could be conrmed, other imaging studies might be unnecessary, thereby reducing cost (potential savings on laboratory tests, cervical cultures, or CT scans), length of stay, and adverse complications of CT (contrast material reactions, and radiation exposure). Emergency medicine ultrasound continues to grow at a rapid pace. We are working toward a time when most EPs will be competent and comfortable performing bedside ultrasound examinations in a limited number of applications. The gynecologic application of ultrasound, however, requires skill beyond the level of the primary applications of emergency medicine ultrasoundspecically, mastering Doppler ultrasound. Although ultrasound has proved to be a valuable imaging modality in the nonpregnant patient with acute abdominal pain when performed by a seasoned sonographer, the role of ED ultrasound has been limited to those EPs with signicantly more training. The author believes that even limited expertise in gynecologic ultrasound is valuable in helping direct the management of these patients. Further research by skilled EP sonographers eventually will help dene the role of EPs in this particular application of ultrasound. EPs should not be discouraged from developing expertise in this examination when a conrmatory study in radiology will be performed. Miles on the ultrasound odometer will not only sharpen sonographic skills but also will help the EP to better communicate with nonpregnant patients presenting with abdominal pain. There is a fairly specic barometer already in place to gauge ones gynecologic ultrasound skills: a seasoned EP sonographer never skips over the chart of a young woman with right lower quadrant pain.
References
[1] Blaustein A. Nonneoplastic cysts of the ovary. In: Blaustein A, editor. Pathology of the female genital tract. New York: Springer-Verlag; 1977. p. 393. [2] Morgan A. Adnexal mass evaluation in the emergency department. Emerg Med Clin N Am 2001;19:799816. [3] Drake J. Diagnosis and management of the adenexal mass [abstract]. Am Fam Physician 1998;57:24716.
696
[4] Ekerhovd E, Wienerroith H, Staudach A, Granberg S. Preoperative assessment of unilocular adnexal cysts by transvaginal ultrasonography: a comparison between ultrasonographic morphologic imaging and histopathologic diagnosis. Am J Obstet Gynecol 2001;184:4854. [5] Bajallan N. Pathology of adenexal masses. Infertility Reprod Med Clin N Am 1995;6: 46192. [6] Kamin RA, Nowicki TA, Courtney DS, Powers RD. Pearls and pitfalls in the emergency department evaluation of abdominal pain. Emerg Med Clin N Am 2003;21:6172. [7] Mortele K, Cantisani V, Brown D, Ros P. Spontaneous intraperitoneal hemorrhage: imaging features. Radiol Clin N Am 2003;41:1183201. [8] Hall DA. Sonographic appearance of normal ovary, of polycystic disease, and of functional ovarian cysts. Semin Ultrasound 1983;4:149. [9] Arger P. Asymptomatic palpable adnexal masses. In: Bluth E, editor. Ultrasonography in urology: a practical approach to clinical problems. New York: Thieme; 2001. p. 10917. [10] Laing F, Brown D, DiSalvo D. Gynecologic ultrasound. Radiol Clin N Am 2001;39: 52340. [11] Baltarowich OH, Kurtz AB, Pasto ME, Rifkin MD, Needleman L, Goldberg BB. The spectrum of sonographic ndings in hemorrhagic ovarian cysts. AJR Am J Roentgenol 1987;148:9015. [12] Ignacio EA, Hill MC. Ultrasound of the acute female pelvis. Ultrasound Q 2003;19:8698. [13] Houry D, Abbott JT. Ovarian torsion: a fteen-year review. Ann Emerg Med 2001;38: 1569. [14] Filly RA. Ovarian masses. . .what to look for. . .what to do. In: Cullen PW, editor. Ultrasonography in obstetrics and gynecology. Philadelphia: WB Saunders Co.; 1994. [15] Fleischer AC, Lin EC. Ovarian torsion. Emedicine. Available at: http://www.emedicine. com/radio/topic509.htm. Accessed August 2002. [16] Stark JE, Siegel MJ. Ovarian torsion in prepubertal and pubertal girls: sonographic ndings. AJR Am J Roentgenol 1994;163:147982. [17] Lee EJ, Kwon HC, Joo HJ, Suh JH, Fleischer AC. Diagnosis of ovarian torsion with color Doppler sonography: depiction of a twisted vascular pedicle. J Ultrasound Med 1998;17: 839. [18] Pena JE, Ufberg D, Cooney N, Dennis AL. Usefulness of Doppler sonography in the diagnosis of ovarian torsion. Fertil Steril 2000;73:104750. [19] Willms AB, Schlund JF, Meyer WR. Endovaginal doppler ultrasound in ovarian torsion: a case series. Ultrasound Obstet Gynecol 1995;5:12932. [20] Pellerito JS, Troiano RN, Quedens-Case C, Taylor KJ. Common pitfalls of endovaginal color Doppler ow imaging. Radiographics 1995;15:3747. [21] Timor-Tritsch IE, Monteagudo A, Rebarber A, Goldstein S, Tsymbal T. Transrectal scanning: an alternative when transvaginal scanning is not feasible. Ultrasound Obstet Gynecol 2003;21:4739. [22] Zeger W, Holt K. Gynecologic infections. Emerg Med Clin N Am 2003;21:63148. [23] Rich RS. Pelvic inammatory disease. In: Cline DM, Ma OJ, Tintinalli JL, Kelen GD, Stapczynski JS, editors. Emergency medicine: a comprehensive study guide companion handbook. 5th edition. New York: McGraw-Hill; 2000. pp. 33740. [24] Timor-Tritsch IE, Lerner JP, Monteagudo A, Murphy KE, Heller DS. Transvaginal sonographic markers of tubal inammatory disease. Ultrasound Obstet Gynecol 1998;12: 5666. [25] Mudgil S. Pelvic inammatory disease/tubo-ovarian abscess. Emedicine. Available at: http://www.emedicine.com/radio/topic 543.htm. Accessed November 22, 2002.
Ultrasound in pregnancy
Christopher Moore, MD, RDMS, RDCSa, Susan B. Promes, MD, FACEPb,*
a
Section of Emergency Medicine, Yale University School of Medicine, PO Box 208062, Suite 260, New Haven, CT 06519, USA b Division of Emergency Medicine, Duke University, Erwin Road, Duke North Room 0681, Box 3096, Durham, NC 27710, USA
Goal-directed pelvic sonography in pregnant women is one of the primary indications for emergency department (ED) ultrasound. It can be used to conrm the presence of an intrauterine pregnancy (IUP) and rule out an ectopic pregnancy in a woman with a positive pregnancy test and abdominal pain or vaginal bleeding. After an IUP has been conrmed, fetal age may be estimated. Bedside ultrasound also can be used to evaluate the fetus and placenta of the pregnant trauma patient to look for an abruption and signs of fetal cardiac activity and fetal movement.
Ectopic pregnancy Ectopic pregnancy, also known as extrauterine gestation (EUG), continues to increase in incidence in the United States and occurs in approximately 2% of all pregnancies [1]. In the ED, the prevalence of EUG in pregnant patients is approximately four times that of the general population and is estimated to occur in 8% of pregnant ED patients [2 4]. Although the death rate from EUG is approximately 1 in 1000, it remains a top cause of maternal mortality and has a relative risk of death 10 times higher than childbirth [5]. Although various algorithms that incorporate clinical and laboratory variables have been proposed, ultrasound remains the test of choice in the evaluation of the pregnant patient who presents with abdominal pain or vaginal bleeding.
* Corresponding author. E-mail address: prome001@mc.duke.edu (S.B. Promes). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.005
698
C. Moore, S.B. Promes / Emerg Med Clin N Am 22 (2004) 697722
Risk factors for ectopic pregnancy include prior pelvic surgery, previous EUG, intrauterine device use, previous genital infections, and smoking, among others [1]. Nearly half of ectopic pregnancies, however, have none of these risk factors. Although a reported 97% of pregnant women with EUG complain of abdominal pain and 87% have vaginal bleeding, history and physical examination alone are not sensitive or specic for the diagnosis [3]. Before the more widespread availability of ED ultrasound, it has been estimated that more than 40% of EUGs were missed on initial presentation [6]. With liberal ultrasound use, sensitivity and specicity should exceed 90% [7].
Who performs the ultrasound? Ultrasound has been called the stethoscope of the future [8] and no longer can be considered to be the domain of any one specialty [9]. Ultrasound for ED patients with suspected EUG may be performed by a technician, radiologist, obstetrician/gynecologist (OB/GYN), or emergency physician (EP). Even within a particular ED, performance of the ultrasound examination may vary depending on time of day and availability of personnel [10]. As this technology continues to become more portable and less expensive, it increasingly is being used by the EP for diagnosis at the bedside. In 2002, 92% of emergency medicine residencies reported performing bedside ultrasound, increased from 50% in 1996 [1113]. In 2002, 78% of emergency medicine programs reported that they trained residents to perform transabdominal rst-trimester obstetric examinations and 59% reported that residents were trained to perform transvaginal obstetric examinations. With proper training, it has been shown that bedside EP-performed ultrasound is safe and may provide advantages in terms of length of stay, overall cost, and clarifying a patients diagnosis before discharge [1417]. Debate continues, however, as to what constitutes adequate training [18 22], and it should be kept in mind that performance of ultrasound by practitioners of any specialty with inadequate training may risk misdiagnosis [23,24]. Typically, the EP approaches this examination with the specic question: Is there an IUP? Although denitive diagnosis of an IUP does not completely rule out coexisting EUG, the presence of an IUP in a patient without risk factors renders the probability small. For the patient in whom the pre-existing probability of EUG was not very high (ie, mild pain or bleeding, hemodynamically stable, not on fertility agents), verication of IUP may place the patient in a range that the EP will be comfortable with outpatient followup. Although it is recommended that the EP who is performing ultrasound attempt to image the adnexa for the presence of any obvious pathology, it is beyond the scope of most EPs to perform detailed ultrasound examinations of the adnexae and ovaries.
699
Ultrasounds performed by a technician in conjunction with a radiologist frequently have the advantage of more time, better equipment, and more extensive training in this modality. In many centers, EPs perform an initial ultrasound, with patients being referred for more denitive testing when these results are questionable or not denitive [16].
b-Human chorionic gonadotropin Although this article primarily addresses ultrasound in pregnancy, understanding the relationship between b-Human chorionic gonadotropin (b-hCG) levels and ultrasound ndings is important. b-hCG is a hormone secreted by the placenta that rises predictably in early pregnancy, nearly doubling every 48 hours. Pregnancy often is rst detected using a qualitative urine b-hCG test, with a serum b-hCG test being performed when the quantitative level is desired. Although urine pregnancy tests used by hospitals and those available over-the-counter dier in sensitivity, most are sensitive and the best detect levels as low as 1 mIU/mL [25,26]. A negative urine pregnancy test, however, does not completely exclude pregnancy. Studies that correlate b-hCG levels with ultrasound examinations have resulted in the concept of the discriminatory zone, which is the b-hCG level at which an IUP should be visualized by ultrasound with near 100% sensitivity [27]. This level generally is agreed to be between 1000 and 2000 mIU/mL and frequently is cited to be 1500 mIU/mL, although it may vary between dierent institutions [2,4,2729]. The discriminatory zone for transabdominal sonography is less well dened, but has been suggested to be between 4000 and 6500 mIU/mL [30,31]. There is no minimum b-hCG level at which EUG rupture may occur. In one series of 38 patients with ruptured EUG, the b-hCG level ranged from 10 mIU/mL to 189,720 mIU/mL [4]. Although the risk of rupture peaks when the b-hCG level approaches 1000 mIU/mL [32], in a series of 131 patients with ruptured EUG, 57% ruptured at a level below 1000 mIU/mL, with 8% below 300 mIU/mL and 2% below 100 mIU/mL [33]. Rupture of an EUG at lower levels typically indicates an isthmic EUG as opposed to an ampullary EUG. Although it is infrequent, EUG rupture can occur and present dramatically, even when the urine b-hCG is negative [34]. When an EUG is suspected, it is reasonable to obtain an ultrasound, regardless of b-hCG level [2,35]. In a study of 111 patients with a b-hCG level of less than 1000 mIU/mL, Dart et al [32] found that ultrasound identied 39% of EUGs in the group and that more than half of these patients had b-hCG levels less than 500 mIU/mL. Although some clinicians recommended withholding ultrasound in stable patients who have a b-hCG level of less than 1500 mIU/mL [36], the largest published protocol using this algorithm discharged over a third of the patients with ectopic
700
pregnancies (69/167) without having ultrasound performed [4,37]. Although sensitivity was claimed to be 100%, the median time to diagnosis was 5 days, 5 patients had rupture at the time of intervention, and 5% of all patients were lost to follow-up. Although ultrasound certainly is less sensitive with a lower b-hCG level, it is likely that more than a third of these ectopic pregnancies could have been identied before discharge and others could have had an IUP conrmed. The patient who presents with a b-hCG level below the discriminatory zone and an indeterminate ultrasound could represent any of three possibilities: EUG, early IUP, or spontaneous abortion. The clinical picture and consultation with an OB/GYN to assure continuity of care are important. In the stable patient, rechecking the b-hCG level 48 hours later in an outpatient setting may be indicated [2]. The b-hCG level in a patient with a normal IUP will double during this time period. A decline in b-hCG of more than 50% indicates a low likelihood of EUG and, most likely, represents a spontaneous abortion. A slowly increasing b-hCG level (\66%), especially in the setting of an empty uterus by ultrasound, is highly indicative of an EUG (odds ratio, 24.7) [29,38]. When a b-hCG level is above the discriminatory zone and no IUP is visualized, many OBs may opt to perform a dilatation and curettage. Presence of chorionic villi provides evidence that this pregnancy represents a spontaneous abortion, whereas absence indicates likely EUG. Some recommend that all women undergo a dilatation and curettage before initiation of methotrexate therapy for EUG, although this is debatable [27,39].
Progesterone Progesterone is another hormone that occasionally has been used for verication of ectopic pregnancy but is not widely used by EPs. Progesterone rises in a normal IUP, although not at the same rate that b-hCG does. Levels below 5 ng/mL are associated with nearly 100% nonviability of the pregnancy (including ectopic), whereas levels above 25 ng/mL are 97.5% predictive of a normal IUP. Levels between 5 and 25 ng/mL are not considered diagnostically helpful [3].
Approach: transabdominal versus transvaginal The two approaches to pelvic imaging (transabdominal and transvaginal) have advantages and disadvantages. Although there is no doubt that the invasive transvaginal approach provides superior imaging of the uterus and adnexae when done correctly, it is not always necessary. The transabdominal approach can provide a broader view, identifying pathology that may be missed with the transvaginal approach alone [40]. The
701
transabdominal approach is noninvasive, easier to learn, and requires less in terms of setup and infection-control precautions. It is the authors recommendation that all pelvic ultrasounds include an initial transabdominal approach, including imaging of Morisons pouch (hepatorenal space) for free abdominal uid. A full bladder is helpful in transabdominal imaging of the pelvis but a hindrance to transvaginal imaging. During transabdominal imaging, a full bladder acts as an acoustic window, allowing better visualization of the pelvic organs. During transvaginal scanning, however, a full bladder forces the uterus posteriorly out of the eld of view and can cause signicant artifacts. If possible, have the patient empty her bladder before transvaginal imaging. Transabdominal imaging typically is done with a curvilinear probe in the frequency range of 2.5 to 5.0 MHz. Frequency is a trade-o: with higher frequency comes better resolution but decreased penetration. Transvaginal scanning is typically done with a tight-radius, small-footprint curvilinear probe in the higher frequency range of 7.0 to 10 MHz. Perhaps counterintuitively, the transvaginal examination may be more comfortable physically for the patient than the transabdominal examination. The combination of pregnancy, a full bladder, and someone pressing on the suprapubic area can be uncomfortable compared with transvaginal imaging and an empty bladder.
Performance of a pelvic ultrasound The following sections describe the use of ED transabdominal and transvaginal pelvic ultrasound to identify an IUP. Organizations and hospitals dier in their requirements for credentialing or privileging physicians to perform ultrasounds but most agree that there is no substitute for proctored instruction and experience. Even if the EP is not performing the ultrasound primarily, it is helpful to understand the process used to obtain the images. This understanding can facilitate patient preparation and communication with the sonographer after images have been acquired. Transabdominal approach and ndings Ideally, a transabdominal pelvic scan is performed with a full bladder, although as the pregnancy progresses, this is less of an issue. A 2.5- to 5MHz ultrasound probe is placed just cephalad to the symphysis pubis with the probe indicator oriented to the patients right side. The probe is angled inferiorly so that the bladder is visualized, with the vaginal stripe and cervix posterior to the bladder (Fig. 1). This is a transverse view and may be considered akin to a CT scan, in which the image is viewed as if from the patients feet upward, with structures on the patients right seen on the left side of the screen as the sonographer looks at it. The probe is then tilted
702
Fig. 1. Transabdominal scan, transverse view. The bladder is anterior, with patients right to the left side of the screen as you look at it. The lower part of the uterus/upper cervix is seen posteriorly.
upward, keeping the indicator to the right, with appearance of the cervix and uterine fundus. The probe should continue to be tilted upward until the top of the uterus disappears from view. After completing a transverse scan of the uterus, the probe should be rotated clockwise so that the probe indicator is toward the patients head, providing a longitudinal (or sagittal) view. In this view, the bladder is seen anteriorly and inferiorly, with the vaginal stripe posterior (Fig. 2).
Fig. 2. IUP. Transabdominal sagittal view shows bladder anteriorly, vaginal stripe posteriorly, and fetal pole within a gestation sac located in the anteverted uterus.
703
Transvaginal approach and ndings The bladder should be as empty as possible before attempting to obtain transvaginal images. A small-footprint, high-frequency 7-MHz or greater intracavitary probe with cover or condom is used. It is important to have gel between the cover and the transducer and outside the cover. The probe should be held with the thumb placed on an indicator that corresponds to the left side of the screen as you face it. Probe orientation may be veried prior to inserting the probe by placing gel on the side of the indicator and making sure that it corresponds to the left side of the screen. Orientation in transvaginal scanning is more challenging due to the fact that the probe is internal. In the longitudinal plane, the indicator (thumb) should be up (toward the ceiling). Looking at the screen in this orientation, the left side of the screen is anterior and the right side is posterior. As the probe is inserted in the vaginal vault, the cervix should come into view. Typically, the end of the probe will be on the cervix or in one of the fornices. In most women, the uterus is anteverted and, thus, the cervix will curve toward the uterine fundus from right to left on the screen (Fig. 3). If the bladder is visualized at all, it will be anterior and on the left side of the screen near the transducer. Posterior to the cervix is the pouch of Douglas (rectouterine pouch), the most dependent area of the pelvis and where pelvic uid will collect if present. The longitudinal view should include the uterus to the top of the fundus and then be swept from left to right to image the entire uterus. Following longitudinal imaging, the probe should be rotated counterclockwise so that the examiners thumb is to the patients right. This provides a transverse or coronal image. In this orientation, the left side of the screen is the patients right, similar to a CT scan or transabdominal
Fig. 3. Empty uterus. Transvaginal sagittal scan shows the uterine fundus anteriorly (to the left of the screen) with endometrial stripe. Bladder is empty and not visualized.
704
Fig. 4. Transvaginal scan, coronal view. The bladder is seen on both sides of the image, anterior to the uterus.
image (Fig. 4). Again, the probe should be swept, this time from top to bottom, to image the uterus from fundus to cervix.
Ovaries and adnexa Detailed imaging of the ovaries and adnexa is beyond the scope of this article. Although these structures occasionally are seen on transabdominal imaging, transvaginal imaging is far superior for imaging them. Ovaries typically appear as somewhat ovoid, generally homogenously echogenic with scattered cystic areas, and have been likened to the appearance of a chocolate chip cookie. They are located anterior and medial to the internal iliac vessels, which should be used as landmarks. Ultrasound ndings Regardless of the ultrasound approach used (although the transvaginal approach obviously is more sensitive), ndings may be categorized into one of three major categories: IUP, visualized ectopic, or indeterminate (also termed no denitive IUP). IUP may be categorized further into live, normal, or abnormal. Presence or absence of free pelvic or abdominal uid may modify ultrasound ndings further. Morisons pouch The hepatorenal space, also known as Morisons pouch, is the most sensitive sonographic area for detection of hemoperitoneum. Because rupture of EUG with signicant bleeding frequently will cause uid to collect in the peritoneal and pelvic spaces [41], it is recommended that
705
Fig. 5. Free uid in the abdomen. Morisons pouch (hepatorenal space) shows an area of anechogenicity (blackness), indicating free uid in the abdomen between the kidney and liver. This case was a ruptured cornual pregnancy that went to the operating room shortly after this scan.
imaging of Morisons pouch be incorporated as a routine part of the ultrasonographic examination for EUG. The presence of uid in the peritoneal space, positive b-hCG, and no IUP can be used as strong, indirect evidence for an EUG [42]. In a retrospective study of 37 ED patients who received operative intervention for ruptured EUG, the presence of a positive Morsions pouch by EP-performed ultrasound was associated with a reduction of time to operative intervention of over 3 hours (from 322 to 111 minutes) [43]. This view typically is obtained in a coronal plane at the right ank with a 2.5 to 5-MHz curvilinear transducer. Presence of an anechoic area between the liver and kidney on this view is indicative of at least 250 to 500 mL of uid, usually indicating signicant blood in the peritoneal space (Fig. 5). Fluid in the cul-de-sac Also known as the pouch of Douglas, the rectouterine pouch or cul-desac is the most dependent portion of the female pelvis; when free uid is present in the pelvis, it tends to collect here. Performance of a culdocentesis, a procedure rendered nearly obsolete by the proliferation of ultrasound [42,44,45], is intended to withdraw uid from this area. On ultrasound, uid in the cul-de-sac appears as an anechoic area posterior to the cervix with or without the presence of echogenic material. It generally is rated subjectively as small, moderate, or large (Figs. 68). The presence of moderate anechoic uid collection in the absence of IUP in symptomatic patients presents a relative risk for EUG of 5.2, whereas echogenic or large uid presents a relative risk of 9.1 [46].
706
Fig. 6. Large free uid in the pelvis. Transabdominal sagittal scan showing free uid posterior to the uterus in the pouch of Douglas (rectouterine space).
Gestational sac The gestational sac is a circular, anechoic intrauterine area that should be visualized transvaginally by about 4 to 5 weeks (b-hCG level of 10002000 mIU/mL) or transabdominally by 5 to 6 weeks. It should be surrounded by a slightly thickened echogenic rim that represents the border of the chorionic cavity (Fig. 9). Absence of this characteristic border may indicate a pseudogestational sac, which can develop with EUG [47]. Although a true gestational sac is indicative of IUP, dierentiation between a true gestational sac and a pseudogestational sac sometimes is dicult. It is
Fig. 7. Pelvic free uid. Transverse abdominal view of the pelvis with large free uid (arrows). Bladder is anterior, with uterus posterior.
707
Fig. 8. Moderate pelvic uid. Transvaginal sagittal scan with empty uterus and moderate free uid in the pouch of Douglas.
recommended that presence of a gestational sac alone should not be considered denitive evidence of IUP, especially in the hands of a novice sonographer [48]. Double decidual sac sign The double-sac sign rst was described by Bradley et al [49] in 1982 as an ultrasound nding used to diagnose early IUP in the absence of embryonic ndings. Although the hormones involved in EUG can cause
Fig. 9. Gestational sac (GS). Transvaginal coronal scan of the uterus showing gestational sac with faint image of the fetal pole. Notice the characteristic hyperechoic double decidual sac sign surrounding the anechoic sac.
708
Fig. 10. IUP. Transabdominal sagittal scan shows the uterus superior to the bladder with an intrauterine gestational sac and yolk sac. Notice the regular circular size of the yolk sac.
thickening of the endometrial lining (decidual cast), a true IUP should involve the development of the decidua parietalis and the deciduas capsularis, creating a pattern of concentric rings in the endometrial cavity. Although the double-sac sign still is considered by many to be the earliest ultrasonographic sign of IUP, it is not 100% sensitive or specic for IUP and should be interpreted with caution [50]. Yolk sac The yolk sac is the rst true embryonic landmark, appearing by about 5 weeks with transvaginal ultrasound. The yolk sac has a characteristic appearance of an echogenic ring ($5 mm in diameter) within the gestational sac (Fig. 10). In a rst-trimester pregnancy, between 5 and 10 weeks estimated gestational age, the upper limit of normal for a developing yolk sac is 5.6 mm [51]. In a normal IUP, a yolk sac should be visualized by the time the gestational sac reaches 8 to 10 mm. Visualization of a yolk sac is the rst denitive evidence of IUP [48]. Fetal pole, heart tones A fetal pole should be visualized by the time the gestational sac reaches 16 mm (Fig. 11). After the fetal pole has reached 5 mm, a cardiac icker should be visualized by transvaginal ultrasound [52]. M-mode (motion mode) sonography may be useful in measuring fetal heart rate. M-mode is a mode in which a one dimensional line (cursor) is placed over the two dimensional image, then displayed over time. This allows visualization and measurment of physical motion, such as that seen in the fetal heart (Fig. 12). Fetal heart rate increases with increasing size, and fetal bradycardia may represent impending fetal loss. A normal fetus of less than 5 mm, 5 to 9 mm,
709
Fig. 11. IUP. Transvaginal coronal scan shows the bladder anteriorly and an intrauterine fetal pole with yolk sac.
and 10 to 15 mm should have a heart rate of at least 100, 110, and 120 beats per minute, respectively [51]. Abnormal pregnancy As the gestational sac enlarges, the yolk sac should be visualized, followed by the fetal pole with cardiac activity. With transvaginal ultrasound, a gestational sac over 10 mm with no yolk sac or a gestational sac over
Fig. 12. Fetal heart tones, measured using measurement mode (m-mode). The two-dimensional (B-mode) image is shown in the upper portion of the screen, with the m-mode cursor across the fetal pole/cardiac icker. The m-mode tracing in the bottom portion shows this onedimensional cursor over time, and depicts the measurement of fetal heart tones (FHT) over two cycles.
710
16 mm with no fetal pole can be considered abnormal. This abnormality typically is called an anembryonic gestation or blighted ovum [52]. Subchorionic hemorrhage Subchorionic hemorrhage is the most common cause of vaginal bleeding in patients with intrauterine gestations [52]. It appears as a hypoechoic or anechoic crescent adjacent to the rim of the gestational sac and represents bleeding at the placental margin [51]. The overall rate of miscarriage with subchorionic hemorrhage is approximately 9%, increasing with maternal age and size of the hematoma [53]. Extrauterine gestationectopic pregancy Although EUG may be suggested by an ultrasound that demonstrates an empty uterus with the presence of free pelvic uid, more denitive evidence is provided by direct visualization of the ectopic pregnancy. The most frequent sonographic evidence of ectopic pregnancy is a complex adnexal mass in the absence of an IUP (Figs. 13 and 14). Occasionally, a true EUG with yolk sac or fetal pole with cardiac icker may be seen. In skilled hands, a live fetus may be seen in approximately one fourth of EUGs [51]. More commonly, a tubal ring is visualized in the adnexa, indicating an embryo developing in a thickening fallopian tube [54]. In addition to true EUGs, it is important to keep in mind that eccentrically placed gestations (see later section on Interstitial pregnancy) are a possibility and may present even more dramatically than tubal gestations. Gestational trophoblastic disorder Gestational trophoblastic disorders, also known as molar pregnancies, may be partial, complete, or invasive (choriocarcinoma). The most common
Fig. 13. EUG. Transabdominal scan showing the bladder anteriorly, uterus posteriorly, with adnexal mass on the right side.
711
Fig. 14. Ectopic gestation. Sagittal transabdominal view with ectopic gestation (later shown to be cornual) at the upper edge of the uterine fundus.
type, a complete mole, represents a pregnancy whereby all embryonic DNA is paternal in origin. Sonographically, the typical appearance of a complete mole is that of an enlarged uterus with multiple small cystic areas, often referred to as a cluster of grapes (Fig. 15). Transabdominal sonography, with its wider eld of view, may provide an easier view of this disorder. Typically associated with extremely high b-hCG levels, molar pregnancies should undergo dilatation and curettage and monitoring to watch for progression to invasive forms.
Fig. 15. Molar pregnancy. Transverse transabdominal view showing the cystic cluster of grapes pattern.
712
Indeterminate Frequently, the ultrasound does not show denitive evidence of IUP, which for purposes of this review is considered to be the presence of an intrauterine gestational sac with yolk sac or fetal pole. The spectrum of indeterminate ndings includes empty uterus, nonspecic uid, echogenic intrauterine material, and normal or abnormal intrauterine sac. In a review of 635 patients with these indeterminate ndings, frequency of EUG was as follows: 36 of 259 with empty uterus (13.7%), 6 of 127 with nonspecic uid collection (4.7%), 4 of 93 with echogenic material (4.3%), and none of 156 with normal or abnormal intrauterine sac [55]. Pitfalls Although these pitfalls apply primarily to sonographers with less experience, experienced sonographers and radiologists are not immune to them [56]. Being aware of pitfalls in the ultrasonographic evaluation of early pregnancy is important whether you perform the ultrasound evaluation or you receive reports that others have performed. Landmarks When imaging a pregnancy by ultrasound, it is tempting to place the probe on the patients abdomen, see a fetus with cardiac activity, and call it an IUP. Although a fetus large enough to have a heartbeat is unlikely to be extrauterine, the literature is replete with cases of extrauterine pregnancies presenting into the second and even third trimesters. It is essential that landmarks be visualized (urinary bladder and vaginal stripe on transabdominal scan, cervix and bladder on transvaginal scan) and that any pregnancy be localized in the correct portion of the uterus. Early intrauterine pregnancy versus ectopic gestation Early intrauterine gestations may be dicult to separate from early ectopic pregnancy. The diagnosis of IUP should be made conservatively to avoid diagnosing an ectopic as an early IUP. The radiology literature describes the double decidual sac sign as the rst ultrasonographic evidence of IUP. This sign appears as two parallel hyperechoic (bright) lines lining the endometrial cavity and represents the closely opposed deciduas vera and deciduas capsularis. Although the double decidual sac sign is reasonably specic for IUP when interpreted by sonogrophers with sucient experience [50], it is somewhat subjective and should not be used solely to determine presence of IUP. Similarly, presence of a gestational sac often is cited as evidence of early IUP but may appear even in ectopic gestation (the pseudogestational sac). The earliest denitive evidence of IUP is a yolk sac appearing within a gestational sac, followed by appearance of the fetal pole with cardiac icker. Avoiding the
713
temptation to interpret the ultrasound as early IUP when yolk sac or fetal pole are not present will avoid false-positive interpretations.
Heterotopic pregnancy Heterotopic pregnancy is the presence of a simultaneous intrauterine gestation and EUG. The exact prevalence of heterotopic pregnancy is unknown, and estimates have ranged from 1 in 2600 to 1 in 30,000 pregnancies [57,58]. Incidence increases with age and multiparity. In the presence of fertility agents, visualization of an IUP cannot be considered adequate to rule out ectopic gestation in the symptomatic patient, and these patients should undergo the most thorough imaging available in consultation with the OB [59]. Even in the absence of fertility agents, a patient with an IUP and no ectopic pregnancy visualized by ultrasound still may harbor a heterotopic pregnancy. All patients presenting for rule-out EUG should be given good discharge instructions and follow-up, and verication of IUP, although reassuring, should not eliminate completely the possibility of EUG [60]. It also is possible for a patient to harbor more than one ectopic pregnancy simultaneously [61].
Interstitial, cornual, subcornual, and cervical pregnancy Pregnancies that are placed eccentrically in relation to the normal low uterine site of implantation represent particular pitfalls for ultrasonographic diagnosis because they may be misinterpreted as an IUP [6,23,62]. In addition, although true interstitial pregnancies represent less than 1% of ectopic gestations, rupture can be catastrophic due to the massive bleeding that may occur if this vascular portion of the fallopian tube at the edge of the uterus is disrupted [63,64]. The myometrial walls of the uterus come together at a junction known as the cornual point, which marks the beginning of the interstitial portion of the fallopian tube. Pregnancies may implant at any portion of this continuum: subcornual, cornual, and interstitial. A pregnancy in the corneal region easily can be misdiagnosed as a normal pregnancy. It is imperative that the sonographer see myometrium surrounding the entire gestation sac and that the myometrial rim be at least 8 mm thick to help rule out a cornual ectopic pregnancy. Management of these pregnancies may vary [65,66]. Although subcornual pregnancies may proceed normally to term, evacuation (dilatation and curettage) may be recommended. Cornual pregnancies typically require removal, which may be accomplished by dilatation and curettage or by laparotomy. True interstitial pregnancies usually are treated by surgical resection and usually in an open manner, although more conservative approaches have been suggested.
714 C. Moore, S.B. Promes / Emerg Med Clin N Am 22 (2004) 697722
Fig. 16. Algorithm for possible ectopic pregnancy. CBC, complete blood count; F/U, Follow-up.
715
An algorithmic approach Fig. 16 represents an algorithmic approach to the patient who presents with a suspicion of EUG. Although this patient typically would be a woman of childbearing age with abdominal pain or vaginal bleeding, other presentations for which an EP would entertain the diagnosis of ectopic pregnancy may be included (lightheadedness, syncope, and so forth). This algorithm incorporates suggested pathways for EP-performed ultrasound, ultrasound performed by radiology, and the complementary use of both. As with any algorithm, there is no substitute for clinical experience and judgment. Treatment of ectopic gestation If an ectopic gestation is discovered or suspected based on ultrasound, serial b-HCG, or dilatation and curettage ndings, management options are of two types: medical or surgical. Emergent surgical intervention is warranted if the patient is unstable, there is a large or increasing amount of pelvic/intraperitoneal uid, or there is an interstitial or large ectopic pregnancy. Conversely, medical management may be more appropriate in a stable patient with an early or suspected ectopic pregnancy and minimal evidence of pelvic or peritoneal bleeding. Methotrexate treatment for ectopic pregnancies may be appropriate in patients with a b-hCG level less than 15,000 mIU/mL, adnexal mass \3.5cm, and minimal free pelvic uid [2]. Lower b-hCG levels (\10,000 mIU/mL) are the strongest predictor of success with this treatment [67]. Typically, an initial dose of 50 mg/m2 of methotrexate is administered, with repeat evaluation (including repeat b-hCG) at 4 days, at which point the dose may need to be repeated. Although generally regarded as safe, this treatment should be administered under direct consultation with an OB/ GYN and with close follow-up and clear discharge precautions [60]. Gestational age After a pregnancy has been identied, reliably assessing gestational age becomes important. Calculating gestational age based on a womans last menstrual period is frequently inaccurate or at times impossible because the woman cannot recall the date of her last menstrual period. Transvaginal ultrasound, as mentioned earlier in this article, allows clinicians to visualize pregnancies early on. Using transvaginal ultrasound, a hypoechoic gestational sac is visible at approximately 4 to 5 weeks gestation and a denitive pregnancy, dened as a yolk sac seen within the gestational sac, typically is visualized with transvaginal ultrasound at approximately 5 to 6 weeks gestation. The next identiable structure to develop is the embryo, followed by the identication of cardiac activity. After an embryo can be identied,
716
the gestational age of the fetus is calculated by measuring the crownrump length, the most accurate sonographic technique for establishing gestational age in the rst trimester. Typically, these measurements for gestational age can be done using transabdominal ultrasound. Ultrasound machines contain software that automatically calculates the gestational age based on the distance between the top of the fetal head to the outer fetal rump (excluding the fetal limbs or yolk sac) as marked on a frozen image. Cardiac activity should be seen when the crownrump length exceeds 7 mm. The fetal heart rate can be measured using a split ultrasound screen with a combination of B-mode and measurement-mode ultrasound. After a good image that shows fetal cardiac activity is identied in B-mode, the cursor is placed over the heart. The split screen, in measurement mode, will show waveforms corresponding with the fetal heartbeat. By using calipers to mark the distance between waveforms, a fetal heart rate is displayed on the screen. In the second trimester, the biparietal diameter is the method commonly accepted to estimate fetal age. To ensure a proper measurement, the biparietal diameter is measured at the level of the thalamus, and the head should be ovoid. After the appropriate image is obtained and frozen on the screen, the ultrasound machine is put into measurement mode. A caliper is placed on the near-eld edge of the parietal portion of skull (outer border) and the track ball is used to move the marker to the near-eld portion of the opposite parietal bone (inner border). Measuring the head circumference of the fetus is another tool that can be used to calculate gestational age. Most ultrasound machines have calipers that open up to outline the fetal head from the same frozen image that is used to measure biparietal diameter. Be aware that abnormal head shapes and prenatal compression of the fetal skull can skew the results of biparietal diameter and head circumference calculations. Biparietal diameter or head circumference can be used to estimate gestational age during the second and third trimesters; the key is to get the best measurement possible. Comparing the results of the two measurements can be helpful because all methods have inherent error. Another site to measure to estimate gestational age is bone length; in particular, femur length. The sonographer rst must identify the fetal bladder and iliac crests and then rotate the probe to visualize the full length of the femur with distinct proximal and distal ends and an acoustic shadow distal to the length of the bone. The femur is measured from the greater trochanter to distal femoral condyle. EPs must remember that their measurements should be thought of as rough estimates, and all pregnant women should have formal ultrasounds performed to document true gestational age because slight errors in measurements can lead to gross over- or underestimations of actual gestational age. Placental abruption In addition to estimating gestational age, ultrasound can be used to assess the placenta; namely, its location to identify a placenta previa and its
717
integrity to identify signs of abruption. A placental abruption, by denition, is a premature separation of the placenta from the uterine wall or myometrium and occurs in less than 1% of all pregnancies but accounts for more than 25% of all perinatal mortalities [68,69]. Risk factors include short umbilical cord, diabetes, cocaine use, smoking, vascular disease, hypertension, and trauma [69]. The classic triad associated with a placental abruption is abdominal pain, vaginal bleeding, and uterine contractions. A patient who presents classically with an obvious abruption should be taken directly to the operating room and the baby delivered immediately without performing an ultrasound. Not all patients with abruptio placenta, however, present classically. For example, depending on the timing and location of the abruption, the patient may not have vaginal bleeding and, in almost 50% of cases, abdominal pain is absent [68,69]. Uterine irritability, manifested by contractions, is present in most cases. In all second- or third-trimester pregnancies in which the clinician is entertaining a diagnosis of placental abruption, the mother should be placed on cardiotocographic monitoring and assessed for uterine contractions and fetal distress. Cardiotocographic monitoring is the key to identifying signs of abruption. Occasionally an abruption can be visualized using ultrasound, however ultrasound is not suciently sensitive to rule out the diagnosis. In the absence of visualized abruption, clinical symptoms and cardiotocographic monitoring over time should be used. Sonographically, the appearance of an abruption depends on the length of time it has been present and its location and size, and can be hyperechoic or isoechoic with respect to the placenta. After a signicant placental abruption is identied, the treatment is immediate cesarean section. If the abruption is small and the mother and fetus are doing well, as demonstrated by cardiotographic monitoring, it may be possible to delay delivery and continue to monitor the patient and her baby while stopping contractions chemically using tocolytics.
Trauma in pregnancy In the past, the leading cause of maternal death was attributed to complications of pregnancy or childbirth. Now, however, trauma has emerged as the leading cause of death during pregnancy and, in one series, accounted for nearly 50% of maternal deaths [70]. According to estimates, 6% to 7% of pregnancies are complicated by trauma and less than 0.5% actually require hospitalization [71]. Most pregnant women are injured by blunt trauma rather than by penetrating trauma [72]. In the pregnant patient, ultrasound can be used to identify hemoperitoneum and potentially decrease the risks associated with a diagnostic peritoneal lavage or radiation directed toward the fetus with a CT scan. The authors suggest that a FAST examination be performed on all gravid blunt trauma patients because of
718
the benet demonstrated by Ma et al [73]. The mantra for pregnant trauma patients has always been tend to the mother rst because fetal survival is dependent on maternal well-being. If a surgically correctable maternal injury exists, then surgical intervention should be obtained in a timely manner in consultation with OB. After approximately 22 to 24 weeks gestation, the EP also must consider the fetal welfare separate from the mother because the fetus may survive outside the uterus at this point in its development. After a FAST examination is performed using the 3.5-mHz probe looking for free intraperitoneal uid, the probe should be directed to the gravid uterus as part of the secondary survey. Fetal age and cardiac activity should be assessed in addition to placing the patient on cardiotocographic monitoring, if available. This information can be helpful to the EP if the mother is unable to be resuscitated and the performance of perimortem cesarean section by the EP is being entertained. In addition, an ultrasound also may show signs of uterine rupture, a rare condition consisting of an empty uterus with free intraperitoneal uid and a fetus and placenta seen within the abdomen. Uterine rupture carries with it a risk of mortality to the mother but almost assures death to the fetus. Trauma is one of the causes of placental abruption, and although ultrasound is not the denitive study to identify this disease process, it occasionally can be a useful adjunct in identifying this potentially life-threatening placental abnormality.
Rapid third-trimester fetal assessment When a pregnant female presents to the ED with an abdominal complaint or complaint that can be attributed to her pregnancy, the patient frequently is sent directly to labor and delivery for evaluation, bypassing the ED. If, however, the patient rst is evaluated in the ED, in addition to a typical ED evaluation for the patients presenting complaint, it can be helpful for the EP to perform a quick assessment of the fetus. By placing a transabdominal probe over the gravid uterus, the EP quickly can note fetal activity, fetal heart rate, quantify the amount of amnitotic uid, and note the fetal position (ie, breech). This rapid assessment must not take the place of cardiotocographic monitoring that is typically performed by obstetric sta. This added information acquired from the bedside ultrasound examination can be valuable when consulting with the OB/GYN colleague. Summary Ultrasound represents the cornerstone of the emergent evaluation of pregnancy-related complaints. Knowledge of the potential and the limitations of this imaging modality, regardless of who performs it, is important for physicians who evaluate and manage these patients and their unborn children in the ED.
719
References
[1] Pisarska MD, Carson SA, Buster JE. Ectopic pregnancy. Lancet 1998;351(9109):111520. [2] American College of Emergency Physicians Clinical Policies Subcommittee on Early Pregnancy. Clinical policy: critical issues in the initial evaluation and management of patients presenting to the emergency department in early pregnancy. Ann Emerg Med 2003;41(1):12333. [3] Stovall TG, et al. Emergency department diagnosis of ectopic pregnancy. Ann Emerg Med 1990;19(10):1098103. [4] Barnhart K, Mennuti MT, Benjamin I, et al. Prompt diagnosis of ectopic pregnancy in an emergency department setting. Obstet Gynecol 1994;84(6):10105. [5] Filly RA. Ectopic pregnancy: the role of sonography. Radiology 1987;162(3):6618. [6] Abbott J, Emmans LS, Lowenstein SR. Ectopic pregnancy: ten common pitfalls in diagnosis. Am J Emerg Med 1990;8(6):51522. [7] Durston WE, Carl ML, Guerra W, et al. Ultrasound availability in the evaluation of ectopic pregnancy in the ED: comparison of quality and cost-eectiveness with dierent approaches. Am J Emerg Med 2000;18(4):40817. [8] Greenbaum LD. It is time for the sonoscope [comment]. J Ultrasound Med 2003;22(4):3212. [9] American Medical Association. Publication H-230.960. 2001. Available at: http://www. ama-assn.org/apps/pf_new/pf_online. Accessed May 21, 2004. [10] Heller M, Crocco T, Patterson J, et al. Emergency ultrasound services as perceived by directors of radiology and emergency departments. Am J Emerg Med 1995;13(4):4301. [11] Counselman FL, Saunders A, Slovis CM, et al. The status of bedside ultrasonography training in emergency medicine residency programs. Acad Emerg Med 2003;10(1): 3742. [12] Cook T, Roepke T. Prevalence and structure of ultrasound curricula in emergency medicine residencies. J Emerg Med 1998;16(4):6557. [13] Heller MB, Mandavia D, et al. Residency training in emergency ultrasound: fullling the mandate. Acad Emerg Med 2002;9(8):8359. [14] Shih CH. Eect of emergency physician-performed pelvic sonography on length of stay in the emergency department. Ann Emerg Med 1997;29(3):34851[discussion: 352]. [15] Mateer JR, Valley VT, Aiman EJ, et al. Outcome analysis of a protocol including bedside endovaginal sonography in patients at risk for ectopic pregnancy. Ann Emerg Med 1996; 27(3):2839. [16] Durham B, Lane B, Burbridge L, Balasubramaniam S, et al. Pelvic ultrasound performed by emergency physicians for the detection of ectopic pregnancy in complicated rsttrimester pregnancies. Ann Emerg Med 1997;29(3):33847. [17] Durham B. Emergency medicine physicians saving time with ultrasound. Am J Emerg Med 1996;14(3):30913. [18] Timor-Tritsch I, Greenidge S, Admon D, Reuss ML. Emergency room use of transvaginal ultrasonography by obstetrics and gynecology residents. Am J Obstet Gynecol 1992; 166(3):86672. [19] Zegel HG, Chong WK, Pasto ME. US in the emergency department: our experience and proposed resolution of a conict between emergency medicine and academic radiology [comment]. Acad Radiol 1999;6(10):5928. [20] Witting MD, Euerle BD, Butler KH. A comparison of emergency medicine ultrasound training with guidelines of the Society for Academic Emergency Medicine [comment]. Ann Emerg Med 1999;34(5):6049. [21] Mateer J, Plummer D, Heller M, et al. Model curriculum for physician training in emergency ultrasonography. Ann Emerg Med 1994;23(1):95102. [22] Lanoix R, Leak LV, Gaeta T, Gernsheimer JR. A preliminary evaluation of emergency ultrasound in the setting of an emergency medicine training program. Am J Emerg Med 2000;18(1):415.
720
[23] Binder DS. Use of ultrasonography in the emergency department: time for a reappraisal [comment]. Ann Emerg Med 2003;41(5):7556. [24] Timor-Tritsch IE. Transvaginal sonography in gynecologic oce practice. Curr Opin Obstet Gynecol 1992;4(6):91420. [25] Alfthan H, Bjorses UM, Tiitinen A, Stenman UH. Specicity and detection limit of ten pregnancy tests. Scand J Clin Lab Invest Suppl 1993;216:10513. [26] Asch RH, Asch B, Asch G, et al. Performance and sensitivity of modern home pregnancy tests. Int J Fertil 1988;33(3):15461. [27] Barnhart KT, Katz I, Hummel A, Gracia CR. Presumed diagnosis of ectopic pregnancy [comment]. Obstet Gynecol 2002;100(3):50510. [28] Ankum WM, Vander Veen F, Hamerlynck JV, Lammes FB. Suspected ectopic pregnancy. What to do when human chorionic gonadotropin levels are below the discriminatory zone. J Reprod Med 1995;40(7):5258. [29] Mol BW, Hajenius PJ, Engelsbel S, et al. Serum human chorionic gonadotropin measurement in the diagnosis of ectopic pregnancy when transvaginal sonography is inconclusive. Fertil Steril 1998;70(5):97281. [30] Kadar N, DeVore G, Romero R. Discriminatory hCG zone: its use in the sonographic evaluation for ectopic pregnancy. Obstet Gynecol 1981;58(2):15661. [31] Chambers SE, Muir BB, Haddad NG. Ultrasound evaluation of ectopic pregnancy including correlation with human chorionic gonadotrophin levels. Br J Radiol 1990; 63(748):24650. [32] Dart RG, Kaplan B, Cox C. Transvaginal ultrasound in patients with low beta-human chorionic gonadotropin values: how often is the study diagnostic? [comment] Ann Emerg Med 1997;30(2):13540. [33] DiMarchi JM, Kosasa TS, Hale RW. What is the signicance of the human chorionic gonadotropin value in ectopic pregnancy? Obstet Gynecol 1989;74(6):8515. [34] Kalinski MA, Guss DA. Hemorrhagic shock from a ruptured ectopic pregnancy in a patient with a negative urine pregnancy test result. Ann Emerg Med 2002;40(1):1025. [35] Kaplan BC, Dart RG, Moskos M, et al. Ectopic pregnancy: prospective study with improved diagnostic accuracy [comment]. Ann Emerg Med 1996;28(1):107. [36] Barnhart KT, Simhan H, Kamelle SA. Diagnostic accuracy of ultrasound above and below the beta-hCG discriminatory zone [comment]. Obstet Gynecol 1999;94(4):5837. [37] Barnhart K, Coutifaris C. Diagnosis of ectopic pregnancy [comment]. Ann Emerg Med 1997;29(2):2956. [38] Dart RG, Mitterando J, Dart LM. Rate of change of serial beta-human chorionic gonadotropin values as a predictor of ectopic pregnancy in patients with indeterminate transvaginal ultrasound ndings. Ann Emerg Med 1999;34(6):70310. [39] Barnhart KT, Katz I, Hummel A, Gracia CR. Presumed diagnosis of ectopic pregnancy. Obstet Gynecol 2002;100(3):50510. [40] Maliha WE, Gonella P, Degnan EJ. Ruptured interstitial pregnancy presenting as an intrauterine pregnancy by ultrasound [comment]. Ann Emerg Med 1991;20(8):9102. [41] Popat RU, Adams CP. Diagnosis of ruptured ectopic pregnancy by bedside ultrasonography. J Emerg Med 2002;22(4):40910. [42] Atri M, Valenti DA, Bret PM, Gillett P. Eect of transvaginal sonography on the use of invasive procedures for evaluating patients with a clinical diagnosis of ectopic pregnancy. J Clin Ultrasound 2003;31(1):18. [43] Rodgerson JD, Heegaard WG, Plummer D, et al. Emergency department right upper quadrant ultrasound is associated with a reduced time to diagnosis and treatment of ruptured ectopic pregnancies. Acad Emerg Med 2001;8(4):3316. [44] Stovall TG, Ling FW. Ectopic pregnancy. Diagnostic and therapeutic algorithms minimizing surgical intervention. J Reprod Med 1993;38(10):80712. [45] Kim DS, Chung SR, Park MI, Kim YP, et al. Comparative review of diagnostic accuracy in tubal pregnancy: a 14-year survey of 1040 cases. Obstet Gynecol 1987;70(4):54754.
721
[46] Dart R, McLean SA, Dart L. Isolated uid in the cul-de-sac: how well does it predict ectopic pregnancy? Am J Emerg Med 2002;20(1):14. [47] Benacerraf B, Parker-Jones K, Schi I. Decidual cast mimicking an intrauterine gestational sac and fetal pole in a patient with ectopic pregnancy. A case report. J Reprod Med 1984; 29(7):498500. [48] Nyberg DA, Mack LA, Harvey D, Wang K. Value of the yolk sac in evaluating early pregnancies. J Ultrasound Med 1988;7(3):12935. [49] Bradley WG, Fiske CE, Filly RA. The double sac sign of early intrauterine pregnancy: use in exclusion of ectopic pregnancy. Radiology 1982;143(1):2236. [50] Nyberg DA, Laing FC, Filly RA, et al. Ultrasonographic dierentiation of the gestational sac of early intrauterine pregnancy from the pseudogestational sac of ectopic pregnancy. Radiology 1983;146(3):7559. [51] Lyons EAL, Cliord S, Dashefsky, et al. In: Rumack CM, editor. The rst trimester, in diagnostic ultrasound. St. Louis (MO): Mosby; 1998. p. 9751011. [52] Albayram F, Hamper UM. First-trimester obstetric emergencies: spectrum of sonographic ndings. J Clin Ultrasound 2002;30(3):16177. [53] Bennett GL, Bromley B, Lieberman E, Benacerraf BR. Subchorionic hemorrhage in rsttrimester pregnancies: prediction of pregnancy outcome with sonography. Radiology 1996; 200(3):8036. [54] Fleischer AC, Pennell RG, McKee MS, et al. Ectopic pregnancy: features at transvaginal sonography. Radiology 1990;174(2):3758. [55] Dart R, Howard K. Subclassication of indeterminate pelvic ultrasonograms: stratifying the risk of ectopic pregnancy. Acad Emerg Med 1998;5(4):3139. [56] James AE Jr, Fleischer AC, Sacks GA, Greeson T. Ectopic pregnancy: a malpractice paradigm. Radiology 1986;160(2):4113. [57] Robert W, DeVoe JHP. Simultaneous intrauterine and extrauterine pregnancy. Am J Obstet Gynecol 1948;56(6):111925. [58] Stephen R, Richards LES, Carlton BD. Heterotopic pregnancy: reappraisal of incidence. Am J Obstet Gynecol 1982;142(7):92830. [59] Thomsen T, Brown DF, Nadel ES. Abdominal pain in rst trimester pregnancy. J Emerg Med 2003;24(1):558. [60] American College of Emergency Physicians. Clinical policy for the initial approach to patients presenting with a chief complaint of vaginal bleeding. Ann Emerg Med 1997;29(3): 43558. [61] OBrien MC, Rutherford T. Misdiagnosis of bilateral ectopic pregnancies: a caveat about operator expertise in the use of transvaginal ultrasound. J Emerg Med 1993;11(3):2758. [62] DeWitt C, Abbott J. Interstitial pregnancy: a potential for misdiagnosis of ectopic pregnancy with emergency department ultrasonography. Ann Emerg Med 2002;40(1): 1069. [63] Chen GD, Lin MT, Lee MS. Diagnosis of interstitial pregnancy with sonography. J Clin Ultrasound 1994;22(7):43942. [64] de Boer CN, Van Dongen PW, Willemsen WN, Klapwijk CW. Ultrasound diagnosis of interstitial pregnancy. Eur J Obstet Gynecol Reprod Biol 1992;47(2):1646. [65] Sagiv R, Golan A, Arbel-Alon S, Glezerman M. Three conservative approaches to treatment of interstitial pregnancy. J Am Assoc Gynecol Laparosc 2001;8(1):1548. [66] Habana A, Dokras A, Giraldo JL, Jones EE. Cornual heterotopic pregnancy: contemporary management options [comment]. Am J Obstet Gynecol 2000;182(5): 126470. [67] Lipscomb GH, McCord ML, Stovall TG. Predictors of success of methotrexate treatment in women with tubal ectopic pregnancies [comment]. N Engl J Med 1999;341(26):19748. [68] Van De Kerkhove K, Johnson TRB. Bleeding in the second half of pregnancy: maternal and fetal assessment. In: Pearlman MD, Tintinalli JE, editors. Emergency care of the woman. New York: McGraw-Hill; 1998. p. 7798.
722
[69] Clark S. Placenta previa and abruptio placentae. In: Creasy F, Resnick S, editors. Maternal-fetal medicine. 4th edition. Philadelphia: WB Saunders; 2000. p. 61663. [70] Fields J, Reed L, Jones N, et al. Trauma: the leading cause of maternal death. J Trauma 1992;32:6435. [71] Lavery JP, Staten-McCormick M. Management of moderate to severe trauma in pregnancy. Obstet Gynecol Clin N Am 1995;22(1):6990. [72] Kissenger DP, Rozycki GS, Morris JA Jr, et al. Trauma in pregnancy: predicting pregnancy outcome. Arch Surg 1991;126:107986. [73] Ma OJ, Mateer JR, DeBehnke DJ. Use of ultrasonography for the evaluation of pregnant trauma patients. J Trauma 1996;40:6658.
Testicular ultrasound
Michael Blaivas, MD, RDMS*, Larry Brannam, MD, RDMS
Section of Emergency Ultrasound, Department of Emergency Medicine, Medical College of Georgia, 1120 15th Street, AF-2039, Augusta, GA30912-4007, USA
Acute scrotal pain makes up approximately 0.5% of all complaints presenting to an emergency department [1]. Some of the most common diagnoses for this complaint are testicular torsion and epididymitis [2]. Misdiagnosing testicular torsion can lead to organ loss, cosmetic deformity, and compromised fertility [3]. Imaging services are not always available to emergency physicians to provide the needed studies to make the diagnosis [4]. This lack of services is due mostly to funding cutbacks experienced throughout the eld of medicine and a shortage of ultrasound technologists who perform most of the radiology ultrasound examinations in this country. The history and the physical examination often fail to make a reliable diagnosis in many cases of acute scrotal pain [2,5]. Even when reliable patients are questioned, the history obtained frequently overlaps for various etiologies of scrotal pain [6]. Epididymitis often is remembered by patients as having a sudden onset. Similarly, epididymitis or orchitis can cause enough diuse pain and swelling that the entire hemiscrotum is painful on examination and the testicular lie is ambiguous. Severe epididymitis actually can cause testicular torsion [7]. For those who rely on urine analysis to distinguish between epididymitis or orchitis and torsion, it is important to keep in mind that the urine will be normal in 50% of patients with epididymitis and orchitis and can be seen as positive in patients with torsion [8]. Studies on emergency physician use of testicular ultrasonography have been limited to several case reports and retrospective studies that have compared its accuracy with surgical ndings, nuclear medicine studies, and radiology ultrasound examinations [810]. The radiology literature is more extensive but still lacks many large prospective or blinded studies that evaluate critical pathways or outcome changes. It is clear, however, that the
* Corresponding author. E-mail address: blaivas@pyro.net (M. Blaivas). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.002
724
M. Blaivas, L. Brannam / Emerg Med Clin N Am 22 (2004) 723748
modern ultrasound examination of the scrotum is the test of choice for acute scrotal pathology and yields high accuracy compared with surgical exploration [1118]. Prompt intervention is required in cases of testicular torsion, rupture, or incarcerated hernias, and this application has high utility for emergency physicians seeing acute scrotal complaints with any frequency.
Technology A broadband linear ultrasound transducer capable of high-resolution imaging (ideally up to 10 MHz) that can perform power and spectral Doppler ultrasonography is critical. Power Doppler probably is of greater importance than color Doppler. Color Doppler is a color-based display of blood ow in vessels of the scrotal contents. It assigns one color to ow toward the transducer and another to ow away from the transducer. The color mapthe scale that dictates what colors are seencan be changed on many machines. This mode is sensitive to direction of blood ow. Power Doppler traditionally is direction insensitive, making it more likely that blood ow is registered. The ability to measure the power of the Doppler signal rather than the Doppler frequency shift signicantly enhances its sensitivity [19]. The end result is that power Doppler is the mode of choice over color Doppler to pick up slow-moving blood such as seen in ovaries and testicles. It is not uncommon to scan a patients testicle using color Doppler and initially be left wondering whether there is any ow in the organ. After changing to power Doppler, however, the amount of ow that is seen almost always increases in the normal testicle. Power Doppler is up to ve times more sensitive to blood ow than regular color Doppler. Stando pads, once recommended for testicular examinations, rarely should be needed with high-frequency transducers that are able to resolve well in the near eld [20].
Technique Typically, the unaected hemiscrotum is scanned rst to familiarize the patient with the process and decrease anxiety regarding discomfort. The scrotum should be scanned in at least two planes (short and long axes). Using the highest resolution transducer reasonable, the scan is begun in a longitudinal axis to the testicle, with the direction indicator toward the head showing a long axis cut through the testicle with the epididymis on the left side of the ultrasound screen (Fig. 1). The testicle is then scanned from one extreme to another, noting the echotexture and any abnormalities. It is important to scan through the entire testicle, thereby obtaining an eective three-dimensional examination of the organ so that an area of hemorrhage, mass, or other pathology that is limited to a small section of the testicle will
725
Fig. 1. Image of a normal testicle with the epdididymal head (E) on the left and body of testicle on the right (T).
not be missed. The scan is then repeated with the probe turned 90( toward the patients right to obtain a short axis cut through the testicle. It is helpful to compare both testicles in gray-scale ultrasonography side by side in short axis (Fig. 2). The remaining scrotal contents also are imaged to evaluate for extratesticular pathology such as an abscess or hernia. A key component of the testicular examination is use of power and spectral Doppler. Examination of the acute scrotum should not be undertaken unless Doppler capability is available because the evaluation of blood ow is such an important part not only of testicular torsion diagnosis but also of orchitis,
Fig. 2. Image of right and left testicles in short axis compared on one screen.
726
epididymitis, trauma, and hemorrhage into a mass. Power Doppler typically is used after the gray-scale ultrasonographic examination of the scrotum is complete. The unaected side is imaged rst to obtain accurate Doppler settings. For instance, the wall lter, scale, and gain may need to be adjusted to pick up the most blood ow without signicant artifact. Power Doppler is prone to motion artifact to a greater degree than color Doppler, and a moving target will make the test almost uninterpretable. Blood ow can be seen throughout the testicle and should be checked for in the epididymis as well (Fig. 3). On most machines, the Doppler gate can be moved around the eld of view so that the hand can be kept steady to reduce artifact. The transducer then can be shifted to another portion of the testicle and the Doppler interrogation performed again. Comparing both testicles on the screen at the same time, with the Doppler gate covering a portion of each of them in cross-section, can help to pick up dierences in ow. Whereas power Doppler allows for detection of blood ow in the testicle, the spectral (pulsed wave) Doppler allows identication of the ow (ie, whether it is venous or arterial) (Fig. 4). This point will be emphasized more in the torsion section; however, torsion should not be ruled out without the use of spectral Doppler to document both venous and arterial blood ow. Typically, power and spectral Doppler can be performed at the same time on the same ultrasound window. This technique helps to aim the spectral gate, which often will be two small parallel lines or a tiny rectangle. The goal is to place this gate into an area of blood ow as denoted by the power Doppler. If the area of the Doppler gate is too large in comparison to the area of blood ow shown by power Doppler, then areas of no ow will be averaged with the area of ow and a poor wave form will be obtained. It is important, therefore, to adjust the sampling gate size to the appropriate size. This adjustment is possible on most modern ultrasound machines. Unlike
Fig. 3. Testicular blood ow shown by power Doppler ultrasonography (arrows).
727
Fig. 4. (A) An example of a venous waveform. Note the at uniform appearance (arrows). (B) An arterial waveform with a systolic peak and diastolic baseline. Note the diastolic peaks of the arterial waveform (arrows).
spectral-Doppler interrogation of the carotids or femoral veins, angle adjustment for direction of the vessel will not be possible in many cases because vessels are tortuous and rarely seen in signicant length.
Normal sonographic anatomy Normal scrotal wall thickness ranges from 2 to 8 mm but depends on cremasteric muscle contraction [21]. The normal testicle is roughly oval in shape. Average measurements are 4 cm 3 cm 2.5 cm on ultrasonography, and weight ranges from 10 to 19 g. The testes are surrounded by the tunica
728
albuginea that is enveloped by the tunica vaginalis. Multiple septations arise from the tunica albuginea and run through the testis. These septations result in the separation of the testis into multiple lobules. The epididymis is an extratesticular structure and is made up of the head, body, and tail. The tail of the epididymis turns into the vas deferens as it travels superiorly out of the scrotum. The vas deferens, in turn, travels up the spermatic cord out of the scrotum. The spermatic cord contains the testicular, cremasteric, and deferential arteries, lymphatic structures, and the genitofemoral nerve. Most of the testicle is supplied with blood by way of the testicular artery that originates from the abdominal aorta. The deferential and cremasteric arteries supply the extratesticular structures of the scrotum, including the epididymis, but also are responsible for a small portion of the arterial supply to the testicle. The gray-scale ultrasonographic appearance of the normal testicle is one of homogeneity and medium echotexture. Vessels may be seen running through the center of the testicle (see Fig. 1). The echogenicity of the testes sometimes is compared with that of the liver. Prepubertal testes typically have lower echogenicity than adult testicles [22]. Many structures such as the tunica albuginea are not seen under normal circumstances. The epididymal head readily can be identied from the rest of the testis in normal and pathologic instances. It has similar echogenicity to the testis but can appear slightly brighter. The body and tail may be harder to dierentiate when no inammation or scrotal uid is present. The appendix testis is a small oval structure that normally is hidden by the epididymal head, making it nearly impossible to dierentiate in normal examinations unless it is surrounded by uid. The mediastinum testis is seen as an echogenic band extending through the center of the testicle in a caudocranial direction. Its thickness and length can vary from individual to individual (Fig. 5). If a hydrocele is present, then the appendix testis often becomes outlined by the uid and is seen as a dened structure (Fig. 6). Rete testis can be identied with sensitive transducers in just less than 20% of patients. When seen in normal subjects, they appear as hypoechoic, tubular areas lying adjacent to each other near the mediastinum testis. Pathologic tubular ectasia of the rete testis can be seen elsewhere in the testicle, as discussed briey later [23]. Color and power Doppler easily can detect blood ow in these vascular structures in normal and most pathologic states (see Fig. 3).
Pathologic ndings Testicular torsion Testicular torsion is an organ- and fertility-threatening surgical emergency. In the past, nuclear medicine studies were considered more accurate, but ultrasonography now is the test of choice for diagnosing and ruling out testicular torsion, largely due to technologic improvements [2426]. Nuclear medicine studies are limited by lack of anatomic denition. Thus, unsuspected
729
Fig. 5. Mediastinum testes (arrows).
pathology, other than testicular torsion, can be missed [27]. Further, in the presence of torsion, hyperemia of the scrotal skin can give a false-negative result because it takes up atypical amounts of tracer [28]. Again, the unaected testicle is scanned rst not only to familiarize the patient with the procedure but also to calibrate the power Doppler settings on the healthy hemiscrotum. Power Doppler gain, lter, and scale should be adjusted to maximize sensitivity to blood ow without creating artifact. The testicle is scanned from side to side and from top to bottom. The eected testicle is scanned in the same manner. In complete torsion, no blood ow will be seen (Fig. 7). If the torsion has existed long enough, then the echotexture of the testicle will begin to change, with areas of inhomogeneity
Fig. 6. Appendix testis (A) is outlined by a hydrocele (F).
730
Fig. 7. A testicle with no ow on power Doppler ultrasonography is shown. The contralateral side showed normal blood ow.
(Fig. 8). The testicle will be hypoechoic compared with the unaected side due to edema and, eventually, necrosis. The overall size of the testicle also will increase due to edematous changes. It is useful to compare both testicles side by side by scanning across the scrotum. This practice allows a comparison of echotexture, size, and color Doppler between the two sides. Animal studies have shown that complete torsion of the testicle occurs at approximately 450( of rotation of the spermatic cord [29]. In cases where torsion is early or not complete, some blood ow may be noted on the power Doppler sonogram [10]. Although a dierence between the two sides
Fig. 8. A missed testicular torsion is shown. No blood ow was detected on power Doppler ultrasonography, and edematous changes are noted.
731
may be obvious, it is critical to make use of spectral Doppler to look for arterial and venous waveforms [30,31]. Due to its low-pressure system, the venous circulation is compromised rst in the torsion process. It is possible, therefore, to see ow on a power Doppler sonogram and mistakenly assume that no torsion is present. A careful examination with spectral Doppler, however, will reveal that venous waveforms are absent while arterial ones are present. This nding is an indication of early or partial torsion, and surgical intervention is warranted. It is tempting to assume that a search for waveforms with spectral Doppler can be stopped after venous waveforms are detected because the venous system is compromised rst, and if such waves are present, then obviously there can be no torsion. An interesting phenomenon can be seen in incomplete torsion, however, in which the venous system already is obstructed and the arterial ow is dampened signicantly but not altogether absent. In such cases, dampened arterial waveforms can appear very similar to venous ones and can result in a falsenegative examination (Fig. 9). Venous and arterial waves, therefore, should be documented to avoid such pitfalls. It is important to be aware of the dierences between the two testicles and perfusion among adults, prepubertal children, and infants. Testicular perfusion is much more dicult to detect before puberty. Typically, higherresolution transducers and more patience are required to detect ow and a potential dierence between two testicles. Some imaging specialists prefer to image prepubertal testicles with nuclear medicine scans, and even then, ow may be dicult to detect in infants. Scrotal masses Scrotal masses fall under acute scrotal pain complaints in a number of dierent ways. Herniation of bowel can develop suddenly and present with
Fig. 9. An example of partial torsion with what appears as venous ow (arrows). In actuality, this was dampened arterial ow.
732
signicant pain. The swelling, however, may be months or even years old. Clinical history and physical examination usually make the diagnosis of intrascrotal inguinal hernia. Ultrasonography, however, can be of critical help when the onset of pain and swelling is acute, good history is not available, or when physical ndings are equivocal. Most commonly, herniation into the scrotum involves bowel, with omentum being the second-most likely culprit (Fig. 10) [32]. Rarely, other abdominal organs can herniate into the scrotum. In many cases of bowel herniation, there may be a small bulge that is easily reducible, representing a small portion of a bowel loop. Occasionally, however, a signicant amount of small bowel can be seen in the scrotum (Fig. 11). Bowel strangulation is less common with direct hernias than with indirect hernias. Color or power Doppler may identify ow in the bowel wall, supporting continued perfusion. A signicant increase in ow to bowel wall, however, may suggest early strangulation. Similarly, continued peristalsis should decrease the likelihood of incarceration at that point in time. Studies show that identication of an akinetic, dilated loop of bowel on scrotal ultrasonography has a 90% sensitivity and a 93% specicity for bowel strangulation [33]. Normally, a small amount of uid is located between the parietal and visceral layers of the tunica vaginalis. This small amount of uid can be visualized in over 80% of normal male patients [34]. When more than minimum uid is seen, a hydrocele has accumulated. Most hydroceles are anechoic in nature and, thus, appear as black around the testicle (Fig. 12). Hydroceles are most commonly seen in the anterolateral aspect of the scrotum. They may be bilateral or unilateral. Hydroceles may result from
Fig. 10. Omentum (O) is seen protruding down the inguinal canal (arrows). The testicle is just o the screen on the right.
733
Fig. 11. Two loops of bowel are seen with good clarity in the scrotum (B). Some fecal material is seen in the top loop of bowel (F).
trauma, infection, neoplasm, radiation therapy, undiagnosed torsion, or congenital anatomic variation [35]. Hydroceles can be seen as an isolated nding or in conjunction with acute or chronic pathology. Infrequently, a hydroceles is the sole nding in a patient who complains of acute scrotal pain. Hydrocele can contain loculations, especially when an inammatory process is present. Most acute hydroceles that form in reaction to infection (such as epididymitis or orchitis) are small. In cases of torsion or trauma, uid collections also are small, except when signicant trauma has occurred
Fig. 12. An example of a hydrocele. F, uid; T, testicle.
734
(Fig. 13). Chronic hydroceles that have developed slowly from a secondary process are larger and may have irregular septations, which reect previous hemorrhage or infection. Pyoceles occasionally are seen with untreated infections and appear as complex lesions with septations and echogenic uid. Varicoceles are an abnormal dilation and tortuosity of the veins in the pampiniform plexus located in the spermatic cord. They can occur in up to 15% of adult men [36]. The varicoceles dilate when the patient stands up or performs the Valsalva maneuver [37]. This nding is even more obvious on color or power Doppler than on physical examination (Fig. 14). Almost all cases of varicoceles are left-sided due to a longer left testicular vein, its angle of entry into the left renal vein, and its risk for compression by the left renal vein and descending colon [38]. Performing the ultrasound examination with the patient standing and lying will help to identify varicoceles along with use of the Valsalva maneuver. Patients in their late teens or early twenties may present with a complaint of acute scrotal mass and considerable anxiety. Ultrasound examination of the testes alone will not discover the varicosities along with, therefore, looking above the epididymal head toward the cord is advisable. After ow in the varicoceles is conrmed on power Doppler, reassurance is the main treatment, with outpatient surgical follow-up. Patients should be made aware that a link between infertility and varicoceles exists [39]. If dilated veins in the cord do not have ow in them with adequate Valsalva or do not collapse easily, then urologic consultation is warranted. Testicular masses Identication of testicular cancer typically should not be considered the goal of emergency ultrasonography. Testicular tumors, however, can
Fig. 13. A hematocele is shown. B, blood.
735
Fig. 14. Varicocele with ow in dilated veins (V). The testicle is just outside of the eld of view on the right.
present to the emergency department, even with allegations that it just appeared. Any focal testicular lesion should be referred for later evaluation for cancer. Presentations can range from large tumors in which little of the testicle is left to focal lesions that will not be found unless the entire testicle is scanned in two axes. Tumors may be cystic (Fig. 15) or solid (Fig. 16) in nature. Hemorrhage in a mass often will appear as an echogenic area within the mass itself. Power Doppler may pick up blood ow in the periphery of such a mass but blood ow will not be seen in the area of hemorrhage. Benign entities such as a dilated rete testis often can take on the appearance of tumors to the inexperienced eye (Fig. 17). If any doubt remains, however, follow-up must be stressed. Testicular microlithiasis is a relatively uncommon condition that may be discovered during examination of the testicle with ultrasonography (Fig. 18). Typically, small areas of calcication are scattered throughout the testicle and are thought to be relatively normal [40]; however, this is not entirely agreed on, and focal calcication can mean the presence of testicular cancer [41]. Testicular trauma Testicular trauma is another potential reason for emergent surgical intervention. Trauma is a common cause of acute scrotal pain and can result from assault, athletic injury, or motor vehicle crashes. Most injuries result in a small contusion, but small and large hematomas, testicular fracture, or even rupture are possible Approximately half of all testicular ruptures result from strikes to the groin during sporting activities. Perhaps surprisingly, up to one fth of all testicular ruptures may result from motor vehicle crashes [42]. When rupture of the testicle occurs, fertility can be compromised as
736
Fig. 15. A large cystic testicular mass (M) is shown. Little of the original testicular tissue is left, except on the periphery of the testicle (T).
a result of an autoimmune reaction. Early exploration results in improved outcome in over 80% of ruptured testicles [42]. It is important to take a thorough look through the scrotal contents. Findings can range from large hematoceles and testicular rupture to focal hematoma of the testicle or point hemorrhages (Fig. 19). Rupture typically is seen when testicle borders are irregular (Fig. 20). Typically, there will be a concomitant hematocele, a hypoechoic area or areas in the testicle, and scrotal thickening [43]. Power
Fig. 16. A solid, septated testicular tumor (M) is shown surrounded by normal testicle (T). The testicle is surrounded by a hydrocele (F).
737
Fig. 17. An example of rete testis (arrows).
Doppler can help to dene areas of focally absent blood ow in the testicle or on its capsule. Actual fracture lines through the testicle are seen in less than 20% of cases. Of greatest concern is a disruption of the testicle that could lead to exposure of testicular contents to the individuals immune system and resultant sterility, even if the aected testicle heals [43]. Unless minimal or no changes are seen in the testicle, urologic follow-up is recommended because focal areas of a hematoma may be small tumor nodules that have hemorrhaged.
Fig. 18. Small areas of calcications (small bright echoes) seen in microlithiasis.
738
Fig. 19. A large hemorrhage with signicant clot (C) next to the testicle (T) is seen. Also note the inhomogeneity of the testicle itself.
Orchitis Orchitis is dened as an acute infection of the testis that usually results from an initial epididymitis [44]. The prevalence of orchitis and epididymitis is rising [14,45]. Clinically, orchitis results in a painful testicle that must be dierentiated from other pathologic conditions, such as an acute torsion, in a timely manner. History and physical examination alone are inadequate in making this distinction. With the use of bedside ultrasonography, emergency
Fig. 20. An ill-dened testicular border (arrows) is seen deep to the testicle after signicant trauma. Surgical exploration revealed rupture.
739
physicians were shown to have a sensitivity of 95% and a specicity of 94% in the diagnosis of patients presenting with acute scrotal pain [10]. In addition, Serra et al [46] found the frequency of inconclusive clinical plus ultrasound evaluation of scrotal lesions to be very low (1.4%). On gray-scale ultrasonography, the inammation and edema of the testicle in orchitis lead to the appearance of decreased echogenicity. Other ndings may include swelling of the epididymis, a hydrocele, and separation of the layers of the scrotal wall by edema. On power Doppler, ow within this area of decreased echogenicity classically is increased in comparison with the unaected side, (Fig. 21). In addition, there is increased ow within the tunica vasculosa, which is visible as lines of color signal radiating outward from the mediastinum testis to the periphery [14]. Serial studies and correlation with symptoms will decrease the likelihood of mistaking this increased vascular ow for post-torsion hyperemia seen after intermittent torsion. The combination of history and physical examination with gray-scale ultrasonography and power Doppler imaging results should allow the emergency physician to make an accurate diagnosis and timely disposition of the patient presenting with orchitis [8].
Fig. 21. An example of orchitis with power Doppler ultrasonography. Note the pronounced areas of ow.
740
Epididymitis Epididymitis is the inammation or infection of the epididymis. It is believed that the most common etiology for infection of the scrotal contents is the retrograde spread of infection from the bladder or prostate by way of the vas deferens [8]. Epididymitis is the most common intrascrotal inammatory process and the most common misdiagnosis for testicular torsion [44]. A retrospective review of pediatric patients presenting with epididymitis, testicular torsion, and torsion of the appendix testis demonstrated signicant overlap in signs and symptoms in patients presenting with an acute scrotum [47]. Bedside emergency ultrasonography has been shown to dierentiate accurately between surgical emergencies and other etiologies of scrotal pain and swelling [10]. Sonographic ndings in epididymitis include enlargement of the epididymis, with decreased echogenicity on gray-scale ultrasonography examinations due to associated edema (Fig. 22). Other ndings may include a hydrocele, similar to orchitis or testicular torsion. The key to dierentiation among the possible diagnoses is color Doppler ultrasonographic ow. Blood ow usually is increased due to inammation of the epididymis and easily seen as increased ow compared with the unaected side (Fig. 23). In the past, it was believed that any epididymal blood ow seen on color Doppler was abnormal and indicated hyperemia [48]. With the advancements made in color ow Doppler imaging, however, a more recent study demonstrated that blood ow easily could be detected to the head, body, and tail of the epididymis in 100% of healthy volunteers [49]. Again, the unaected side always should be compared with the side that is symptomatic. As in orchitis, by using the history and physical examination along with
Fig. 22. An enlarged epididymal head (E) is seen in this case of epididymitis. T, testicle.
741
Fig. 23. A slightly oblique view of a testicle with an enlarged epididymis and signicant blood ow on power Doppler ultrasonography (inside Doppler box).
bedside gray-scale ultrasonography and color ow Doppler imaging of the acute scrotum, the emergency physician will be able to dierentiate patients with epididymitis from those who require urgent surgical consultation and have other, similar conditions presenting with overlapping symptoms. In the past, it was common lore that all prepubertal boys who presented with acute scrotal pain required surgical exploration and that because the incidence of torsion in this population was so high, imaging was of little use [50,51]. Several studies, however, have shown that torsion is responsible for less than one fourth of acute testicular pain in boys younger than 17 years. Epididymitis accounts for 31% to 70% of these cases [1,2,47,52]. Torsion of appendix testis Torsion of the appendix testis typically is seen in prepubertal patients but can occur later in life. A blue dot sign is not always present. Depending on the location of the appendix, it may not be well visualized on examination. It also is less obvious in patients with darker skin color. On ultrasonography, the appendix testis normally is not seen unless it is outlined by uid from a hydrocele (see Fig. 5). During torsion of the appendix testis, the resulting inammation often leads to inammation of the overlying head of the epididymis, giving rise to epididymitis in prepubertal boys and nonsexually active men. The edematous appendix testis can be visualized on gray-scale
742
ultrasonography and may be seen on as an area of absent ow on power Doppler (Fig. 24). Scrotal infection Infection of the scrotum itself is fairly uncommon and typically thought of in association with diabetes mellitus. Scrotal edema and erythema may be caused by a number of processes including insect or human bites, contact dermatitis, congestive heart failure, hypoalbuminemia, generalized anasarca, and fungal and bacterial infections or may be idiopathic in nature. The major concern is the formation of a scrotal abscess or Fourniers gangrene. Scrotal abscess involving the skin typically is from an infected hair follicle but can originate from the internal contents of the scrotum such as the testis or epididymis. Sonographic ndings will include a focal uid collection, often of complex uid (mixture of anechoic and echogenic areas). It is important to determine whether such sites are associated with internal scrotal pathology such as pus or an abscess inside the scrotum. Scrotal cellulitis is mostly a clinical diagnosis, except for ultrasonographic verication that no gas or abscess is present. Fourniers gangrene is a polybacterial infection of the scrotum that typically originates from the skin, rectum, or urethra/prostate. The infectious process quickly progresses and leads to thrombosis in the end arteries of scrotal skin and, nally, to necrosis. Although typically associated with diabetes, it can be seen in patients without a history of immune problems. Sonographic evaluation will reveal signicant skin edema throughout the aected area. Gas is easily seen, even before it is
Fig. 24. An enlarged and inamed epididymis is seen. Next to it is a globular area lacking ow which is the torsed appendix testes (A).
743
obvious on physical examination (Fig. 25). It appears as areas of dirty shadowing that can be seen even in small quantities, something that may be dicult to see in plain radiographs of the peritoneum [53]. In nondiabetic patients, the origin of the infection also should be sought by scanning posterior to the scrotum. This process occasionally will reveal a perirectal or perianal abscess. If the origin of the infection is prostatic, then the soft tissue edema and, perhaps in some cases, pus can be traced directly to the prostate (Fig. 26). A number of reports exist that document this nding in the radiology and emergency ultrasound literature [54,55]; however, no large prospective comparisons have been made with CT or plain radiography. In seriously ill patients with scrotal swelling of uncertain origin, scrotal and perineal ultrasonography demonstrated gas in the soft tissue before crepitus was detected on physical examination.
Acute scrotal pain evaluation algorithm Acute scrotal pain often is thought to be pain that started less than 24 hours previously. Due to the stuttering nature of partial torsion or detorsion, however, all testicular pain complaints should be take seriously and evaluated quickly until it can be determined that no surgical emergency is present. An ultrasound examination should be performed as quickly as possible, even before other testing such as blood or urine if there is any delay in those tests being drawn. Because the ultrasound examination is denitive in detecting or ruling out a surgical emergency, the patient should be made as comfortable as possible with parenteral analgesics if necessary. This practice will allow for a more cooperative patient. In most cases, the
Fig. 25. Gas (arrows) is seen in the scrotum in this example of Fourniers gangrene. The gas appears as bright echoes and interferes with imaging of testicle deep to it.
744
Fig. 26. Pus (P) is noted in the scrotum.
emergency physician will be concerned with ruling out testicular torsion, testicular rupture, or the presence of a strangulated hernia. The integrity of the testicle can be veried with a complete survey of the scrotal contents with gray-scale ultrasonography. Findings such as loops of bowel or omental fat can be detected at this point. Power Doppler is performed next to evaluate for blood ow in the aected testicle. If a surgical emergency is conrmed, then immediate urologic consultation is indicated. If surgical detorsion will not occur in less than 6 hours, manual detorsion should be considered. The success of such a maneuver can be veried quickly with the ultrasound examination. Occasionally, the patient will have normal venous and arterial blood ow and nothing else to explain testicular pain, such as epididymitis or orchitis. This subset of patients can be the most challenging because their pain remains a concern for torsion yet ultrasound ndings are normal. The patient can be scanned 30 minutes later to see whether anything has changed. Slowly progressing torsion may show less blood ow than before; conversely, it may be reassuring to see continued normal ow. Finally, if testicular torsion is not believed to be adequately ruled out, then a nuclear medicine scan should be performed or a urologist should be consulted for surgical exploration.
Pitfalls Failure to perform an ultrasound examination One of the most common pitfalls is not performing an ultrasound examination and, thus, failing to diagnose a surgical emergency in a timely
745
manner. This pitfall typically occurs when history or physical examination lead the clinician into a false sense of security regarding a nonemergent diagnosis. Frequently, the history and physical examination do not t into the clinicians paradigm of what testicular torsion should look like. Because ultrasonography is easy to perform at bedside, it behooves the emergency physician to make use of it when considering a surgical emergency in the dierential. Not conrming venous and arterial ow As suggested previously, failure to use spectral Doppler to assure the presence of both arterial and venous blood ow can lead to false-negative ultrasound examinations. It is important visualize and then conrm central
Fig. 27. (A) Testicular blood ow before intravenous contrast administration. (B) Dramatically increased power Doppler ultrasonography detected ow 20 seconds after injection.
746
blood ow within the testicle to avoid mistaking scrotal blood ow for testicular ow. Hyperemia from detorsion When the testicle regains its blood supply in ongoing torsion and detorsion or in spontaneous detorsion, resultant hyperemia is noted. Hyperemia must be noted and dierentiated from orchitis because these patients should still be seen by urology and ideally taken to the operating room for xation because they are at risk for retorsion. Serial examination, unthinkable for the traditional imaging provider, can be accomplished easily at the patient bedside and be surprisingly helpful.
Future developments The most signicant future development will be the continued spread of ultrasound technology that is capable of performing adequate scrotal examinations to accurately rule out testicular torsion. Of interest is intravenous ultrasound contrast. Approved and widely used in Europe and soon to be approved in the United States, the addition of ultrasound contrast greatly enhances Doppler signal detection, allowing the detection of blood ow in a testicle in which little ow otherwise would be noted (Fig. 27).
References
[1] Lewis AG, Bukowski TP, Jarvis PD, Wacksman J, Sheldon CA. Evaluation of acute scrotum in the emergency department. J Pediatr Surg 1995;30:27781. [2] Knight PJ, Vassy LE. The diagnosis and treatment of the acute scrotum in children and adolescents. Ann Surg 1984;200:664. [3] Dunne PJ, OLoughlin BS. Testicular torsion: time is the enemy. Aust N Z J Surg 2000;70: 4412. [4] Blaivas M, Lambert MJ, Harwood RA, Wood JP, Konicki J. Lower-extremity Doppler for deep venous thrombosiscan emergency physicians be accurate and fast? Acad Emerg Med 2000;7:1206. [5] Rosen P, editor. Emergency medicine concepts and clinical practice. 4th edition. St. Louis (MO): Mosby; 1997. [6] Jeerson RH, Perez LM, Joseph DB. Critical analysis of the clinical presentation of acute scrotum: a 9-year experience at a single institution. J Urol 1997;158:1198. [7] Bird K, Roseneld AT. Testicular infarction secondary to acute inammatory disease: demonstration by B-scan ultrasound. Radiology 1984;152:7858. [8] Blaivas M, Sierzenski P. Emergency ultrasonography in the evaluation of the acute scrotum. Acad Emerg Med 2001;8:859. [9] Blaivas M, Batts M, Lambert M. Ultrasonographic diagnosis of testicular torsion by emergency physicians. Am J Emerg Med 2000;18:198200. [10] Blaivas M, Sierzenski P, Lambert M. Emergency evaluation of patients presenting with acute scrotum using bedside ultrasonography. Acad Emerg Med 2001;8:903.
747
[11] Nussbaum Blask AR, Bulas D, Shalaby-Rana E, Rushton G, Shao C, Majd M. Color Doppler sonography and scintigraphy of the testis: a prospective, comparative analysis in children with acute scrotal pain. Pediatr Emerg Care 2002;18:6771. [12] Kravchick S, Cytron S, Leibovici O, Linov L, London D, Altshuler A, et al. Color Doppler sonography: its real role in the evaluation of children with highly suspected testicular torsion. Eur Radiol 2001;11(6):10005. [13] Weber DM, Rosslein R, Fliegel C. Color Doppler sonography in the diagnosis of acute scrotum in boys. Eur J Pediatr Surg 2000;10:23541. [14] Cook JL, Dewbury K. The changes seen on high-resolution ultrasound in orchitis. Clin Radiol 2000;55:138. [15] Lerner RM, Mevorach RA, Hulbert WC, Rabinowitz R. Color Doppler US in the evaluation of acute scrotal disease. Radiology 1990;176:3558. [16] Sanelli PC, Burke BJ, Lee L. Color and spectral doppler sonography of partial torsion of the spermatic cord. AJR Am J Roentgenol 1999;172:4951. [17] Dogra VS, Sessions A, Mevorach RA, Rubens DJ. Reversal of diastolic plateau in partial testicular torsion. J Clin Ultrasound 2001;29:1058. [18] Middleton WD, Middleton MA, Dierks M, Keetch D, Dierks S. Sonographic prediction of viability in testicular torsion: preliminary observations. J Ultrasound Med 1997;16: 237. [19] Albrecht T, Lotzof K, Hussain HK, Shedden D, Cosgrove DO, de Bruyn R. Power Doppler US of the normal prepubertal testis: does it live up to its promises? Radiology 1997;203:22731. [20] Milleret R. Doppler ultrasound diagnosis of testicular cord torsion. J Clin Ultrasound 1976;4:4257. [21] Trainer TD. Testis and the excretory duct system. In: Sternberg S, editor. Histology for pathologists. New York: Raven; 1992. p. 7446. [22] Siegel MJ. The acute scrotum. Radiol Clin N Am 1997;35:95976. [23] Thomas RD, Dewbury KC. Ultrasound appearances of the rete testis. Clin Radiol 1993;47: 1214. [24] Dogra VS, Gottlieb RH, Oka M, Rubens DJ. Sonography of the scrotum. Radiology 2003; 227:1836. [25] Allen TD, Elder JS. Shortcomings of color Doppler sonography in the diagnosis of testicular torsion. J Urol 1995;154:150810. [26] Middleton WD, Siegel BA, Melson GL, Yates CK, Andriole GL. Acute scrotal disorders: prospective comparison of color Doppler US and testicular scintigraphy. Radiology 1990; 177:17781. [27] Frush DP, Sheldon CA. Diagnostic imaging for pediatric scrotal disorders. Radiographics 1998;18:96985. [28] Frush DP, Babcock DS, Lewis AG, Paltiel HJ, Rupich R, Bove KE, et al. Comparison of color Doppler sonography and radionuclide imaging in dierent degrees of torsion in rabbit testes. Acad Radiol 1995;2:94551. [29] Janetschek G, Schreckenberg F, Grimm W, Marberger M. Hemodynamic eects of experimental testicular torsion. Urol Res 1987;15:3036. [30] Sanelli PC, Burke BJ, Lee L. Color and spectral doppler sonography of partial torsion of the spermatic cord. AJR Am J Roentgenol 1999;172:4951. [31] Mevorach RA, Lerner RM, Greenspan BS, Russ GA, Heckler BL, Orosz JF, et al. Color Doppler ultrasound compared to a radionuclide scanning of spermatic cord torsion in a canine model. J Urol 1991;145:42833. [32] Korenkov M, Paul A, Troidl H. Color duplex sonography: diagnostic tool in the dierentiation of inguinal hernias. J Ultrasound Med 1999;18:5658. [33] Ogata M, Imai S, Hosotani R, Aoyama H, Hayashi M, Ishikawa T. Abdominal ultrasonography for the diagnosis of strangulation in small bowel obstruction. Br J Surg 1994;3:4214.
748
[34] Leung ML, Gooding GA, Williams RD. High-resolution sonography of scrotal contents in asymptomatic subjects. AJR Am J Roentgenol 1984;143:1614. [35] Stewart R, Carroll B. The scrotum. In: Rumack CM, Wilson SR, Charboneau JW, editors. Diagnostic ultrasound vol 1. St. Louis (MO): Mosby; 1991. p. 56589. [36] Meacham RB, Townsend RR, Rademacher D, Drose JA. The incidence of varicoceles in the general population when evaluated by physical examination, gray scale sonography and color Doppler sonography. J Urol 1994;151:15358. [37] Wolverson MK, Houttuin E, Heiberg E, Sundaram M, Gregory J. High-resolution realtime sonography of scrotal varicocele. AJR Am J Roentgenol 1983;141:7759. [38] Blaivas M. Testicular ultrasound. In: Ma OJ, Mateer J, editors. Emergency ultrasound. New York: McGraw-Hill Professional; 2002. p. 2218. [39] Pierik FH, Vreeburg JT, Stijnen T, van Roijen JH, Dohle GR, Lameris JS. Improvement of sperm count and motility after ligation of varicoceles detected with colour Doppler ultrasonography. Int J Androl 1998;21:25660. [40] Janzen DL, Mathieson JR, Marsh I, Cooperberg PL, del Rio P, Golding RH, et al. Testicular microlithiasis: sonographic and clinical features. AJR Am J Roentgenol 1992; 158:105760. [41] Backus ML, Mack LA, Middleton WD, King BF, Winter TC III, True LD. Testicular microlithiasis: imaging appearances and pathologic correlation. Radiology 1994;192:7815. [42] Cass AS. Testicular trauma. J Urol 1983;129:299300. [43] Jerey RB, Laing FC, Hricak H, McAninch JW. Sonography of testicular trauma. AJR Am J Roentgenol 1983;141:9935. [44] Rosen P, editor. Emergency medicine concepts and clinical practice. 5th edition. St. Louis (MO): Mosby; 2002. p. 14256. [45] Morbidity Statistics from General Practice. 4th national study, 19911992, Series MB5, no. 3. [46] Serra AD, Hricak H, Coakley FV, Kim B, Dudley A, Morey A, et al. Inconclusive clinical and ultrasound evaluation of the scrotum: impact of magnetic resonance imaging on patient management and cost. Urology 1998;51:101821. [47] Kadish HA, Bolte RG. A retrospective review of pediatric patients with epididymitis, testicular torsion, and torsion of testicular appendages. Pediatrics 1998;102:736. [48] Horstman WG, Middleton WD, Melson GL, Siegel BA. Color Doppler ultrasound of the scrotum. Radiographics 1991;11:94158. [49] Keener TS, Winter TC, Nghiem HV, Schmiedl UP. Normal adult epididymis: evaluation with color Doppler US. Radiology 1997;202:7214. [50] Klein B, Ochsenschlager D. Scrotal masses in children and adolescents: a review for the emergency physician. Pediatr Emerg Care 1993;9:3516. [51] Snyder H, Caldomone A, Duckett J. Scrotal pain. In: Fleisher G, Ludwig S, editors. Textbook of pediatric emergency medicine. 3rd edition. Baltimore (MD): Williams & Wilkins; 1993. p. 3827. [52] Kass EJ, Stone KT, Cacciarelli AA, Mitchell B. Do all children with an acute scrotum require exploration? J Urol 1993;150:6679. [53] Kane CJ, Nash P, McAninch JW. Ultrasonographic appearance of necrotizing gangrene: aid in early diagnosis. Urology 1996;48:1424. [54] Biyani CS, Mayor PE, Powell CS. Case report: Fourniers gangreneroentgenographic and sonographic ndings. Clin Radiol 1995;50:7289. [55] Dogra VS, Smeltzer JS, Poblette J. Sonographic diagnosis of Fourniers gangrene. J Clin Ultrasound 1994;22:5712.
Ultrasound guidance for vascular access

Paul-Andre C. Abboud, MDa, John L. Kendall, MD, FACEPb,*
Department of Emergency Medicine, Kaiser Permanente Medical CenterOakland, 280 West MacArthur Boulevard, Oakland, CA 94611, USA b Division of Emergency Medicine, University of Colorado Health Sciences Center; Assistant Professor, Department of Emergency Medicine, Denver Health Medical Center, 777 Bannock Street, Denver, CO 80204, USA
a
The ability to establish central venous access eciently is a fundamental skill for emergency physicians. Central venous access is essential for hemodynamic monitoring, volume resuscitation, and the delivery of vasoactive drugs [1]. It is important in the management of shock and other conditions such as renal failure and complete heart block because it contributes to temporizing and life-saving therapies. Traditionally, central venous access has been guided only by palpable anatomic landmarks such as bony prominences, muscle surfaces, and arterial pulsations. This blind approach to the central veins assumes anatomic homogeneity, does not account for the possibility of thrombosis, and depends on correct discernment of the relationship among multiple anatomic landmarks [2]. Research in emergency department (ED) and intensive care settings has supported the ecacy of traditional landmark approaches to the internal jugular vein (IJV), subclavian vein (SV), and femoral vein (FV) in adult [3 12] and pediatric patients [1316]. Failure rates, however, have been reported as high as 30% in some series [17]. The failure rate has been demonstrated to be greater for emergent cases and highest for patients in cardiopulmonary arrest [18]. Nonrandomized studies of central venous cannulation specically for critical trauma [4,6,11] or cardiopulmonary resuscitation [8,10,11] have reported success rates ranging from 62% to 99%. One study of failed cardiopulmonary resuscitation cases demonstrated that 31% of attempted FV catheters were not in the FV [19].
* Corresponding author. E-mail address: john.kendall@uchsc.edu (J.L. Kendall). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.003
750
P.-A.C. Abboud, J.L. Kendall / Emerg Med Clin N Am 22 (2004) 749773
Complication rates related to central line placement are reported to range from 0.3% to 18.8%, depending on the site of insertion, patient population, and denition of complications [17,2022]. Acute complications associated with the landmark approach commonly include pneumothorax, arterial puncture, hemothorax, hematoma (subcutaneous or mediastinal), misplaced catheter tip, nerve injury, and dysrhythmia [16,17,2325]. Cases of transient Horner syndrome and dysphonia after IJV catheterization have been reported in some series [21,26]. Death due to complications from a central venous line also has been reported [27]. The complication rate depends on the time needed for catheter insertion [22], the number of needle passes [28], the extremes of body habitus, previous central venous cannulation, prior surgery or radiation therapy in the area of the vein, and operator inexperience [20]. Characteristics associated with dicult or complicated central access include limited sites for access attempts (other catheters already in place, pacemaker, local surgery, or infection), known vascular abnormality, coagulopathy, mechanical ventilation or severely diminished pulmonary function (leading to worse morbidity from a possible pneumothorax), severe peripheral vascular disease, soft tissue edema, chronic intravenous drug use, and patient intolerance of supine position (orthopnea, increased intracranial pressure) [2934]. Under these circumstances in which the margin for error is small, central venous access must be undertaken carefully. Emergency medicine has developed an expanding familiarity with portable two-dimensional (2-D) real-time ultrasound (US) over the past decade [35]. In that time, a body of research has developed that supports US for guidance of central venous cannulation. Descriptions of US guidance for central venous access rst were published in the anesthesiology literature and, subsequently, in the surgery, radiology, nephrology, critical care, and emergency medicine literature. In 1978, Ullman and Stoelting [36] rst described the use of a pencilshaped Doppler probe to identify the windstorm sounds of the IJV to mark the overlying skin site for cannulation. Legler and Nugent [37] published the rst experience with Doppler localization of the IJV before catheterization. In 1986, Yonei et al [38] rst reported the use of 2-D realtime US guidance for cannulation of the IJV. The rst case series of 2-D US for central venous access in the ED was published in 1997 [39]. The reported technique involved two operators: one for line placement and one to hold the US probe. Since then, emergency physicians have published four studies on the use of US for vascular access in the ED [33,4042]. These studies, which are reviewed below, reported favorable experiences and improved success rates for venous access with US guidance. In 1997, the American College of Emergency Physicians (ACEP) published a policy statement on the use of US imaging by emergency physicians. In 2001, a revised policy statement and accompanying ACEP
751
guidelines specically included US guidance for central venous access in a list of primary applications for emergency ultrasound [43].
Evidence for ultrasound-guided central venous access In 2001, the Agency for Healthcare Research and Quality published an evidence-based report on patient safety practices. This report, which has been highly publicized in professional and lay media, includes a chapter on US guidance for central venous access. US guidance for central venous access was listed among 11 practices with the most highly rated strength of evidence for supporting more widespread implementation [44]. This report based its ndings on much of the same literature that previously had been reviewed in a meta-analysis by Randolph et al [1]. The meta-analysis, published in 1996, reviewed eight randomized controlled trials of 2-D or Doppler US guidance versus the landmark method for central venous access. No studies of FV access were included. A signicant decrease in the failure rate, complication rate, and number of attempts for successful access of the SV and the IJV were reported. A subsequent meta-analysis commissioned by the British National Institute for Clinical Excellence (NICE) was published in 2003 [45]. It included 18 randomized controlled trials published through October 2001. These trials compared 2-D real-time or Doppler US with the landmark method for central venous access. The meta-analysis considered risk of failed placement, complications, failure on the rst attempt, number of attempts to successful access, and time to successful access as outcome measures. These outcome measures were analyzed by type of vein studied (IJV, SV, and FV), by US method (2-D and Doppler), and by age category (adult and infant). This meta-analysis concluded that 2-D US guidance was more eective than the landmark method for all outcomes for IJV access in adults. The relative risks of failed attempts, complications, and failed rst attempts were reduced by 86%, 57%, and 41%, respectively. Signicantly fewer attempts were required for success, and the IJV was successfully accessed more quickly when using US. Limited evidence suggested that 2-D US guidance reduced the relative risk of failed access in the SV and FV. The three studies of IJV access in infants included in the meta-analysis were limited by small sample size [4648]; however, the analysis suggested that 2-D US was more eective in these studies. Using US, the relative risk of failed placements and complications in infants was reduced by 85% and 73%, respectively. No studies of SV or FV access among infants were included in the meta-analysis. The investigators [45] also undertook a costeectiveness analysis of 2-D US guidance based on the evidence from their systematic review of the literature. The analysis of a simple decision analytic model suggested that US guidance avoided 90 arterial punctures for every 1000 patients and reduced costs by a negligible amount (approximately $5)
752
per patient. Given the evidence for its superior ecacy, the recent Agency for Healthcare Research and Quality mandate for improving patient safety, and the 2001 ACEP emergency US guidelines, US guidance for central venous access has been transformed from an interesting novelty to an important skill for emergency physicians to acquire.
General technical issues There are several commonly accepted variations of US guidance: indirect, direct or real-time, free-hand, mechanical guide, and Doppler. Choosing among these approaches mostly depends on the location of the vessel to be cannulated and the specic characteristics of the operator, patient, and the equipment at hand. In addition, vessel visualization can be obtained in two dierent ways: in the long axis or the short axis. A solid understanding of these technical issues is necessary to successfully cannulate a vessel, regardless of the approach used or the location of the vessel. Indirect method The indirect method employs the least amount of actual guidance. With this approach, US is used only to identify the vessel and then center it on the US screen (Fig. 1). Next, a temporary mark is placed on the skin that corresponds to the vessels subcutaneous position. This mark is used for the puncture site after US identies the target vessels location, dimensions, and depth below the skin. The easiest way to accurately make this mark is to identify the point where the center of the transducer overlies the skin surface just above the center of the vessel (Figs. 2 and 3).
Fig. 1. IJV and carotid artery viewed in their short axes.
753
Fig. 2. Subcutaneous location of IJV is identied on the skin surface.
There is no direct visualization of the needle as it enters the vessel, however. This technique has been used for localization and cannulation of larger structures but has been criticized for not taking full advantage of the potential of US for greater precision [49]. Manseld et al [20] compared the indirect method of US guidance with the standard landmark approach for SV cannulation. This study was closed after an interim analysis of 824 patients showed that US guidance by the indirect method had no eect on the rate of complications or failures. Real-time visualization The alternative to the indirect method is to perform needle placement under direct, or real-time, US guidance so that the entire procedure is visualized continuously. With this technique, a sterile sheath whose tip is lled with transmission gel is unrolled over the transducer (Fig. 4). The
Fig. 3. Subcutaneous location of IJV is marked on the skin surface.
754
Fig. 4. Sterile barrier covering transducer and cable.
transducer is then placed on the skin and the target vessel is identied and centered on the viewing screen. With the other hand, local anesthetic is injected at a point corresponding to the middle of the US transducer. After anesthesia is achieved, the cannulation needle is advanced through the skin. After the skin has been punctured, the operator can switch visual focus to the US monitor where the needle will appear sonographically as an echogenic line with a ring-down artifact (Fig. 5). Advancement of the needle is then guided by viewing its progression on the US monitor. When the operator visualizes the needle piercing the anterior wall of the vessel (Fig. 6) and after the subsequent ash of blood into the syringe, the transducer is placed aside and the remaining aspects of the procedure can be completed normally. Few studies have compared indirect and real-time US guidance methods for insertion of venous catheters. Nadig et al [50] randomized 73 patients to an external landmark or a real-time US guidance of IJV cannulation. There were 87 unsuccessful attempts among 37 patients in whom cannulation was performed using the indirect method. In comparison, there were only 10
Fig. 5. Advancing needle and ring-down artifact.
755
Fig. 6. Needle visualized advancing through anterior (ant.) vessel wall.
unsuccessful attempts among the 36 patients who underwent real-time US guidance. Mechanical guides A mechanical guide is an attachment to the US transducer that controls the depth, angle, and trajectory of the needle during cannulation (Fig. 7). In addition to venous cannulation, mechanical guides are used for many other US-guided invasive procedures, including amniocentesis, follicle retrieval, cordocentesis, biopsies, and uid aspiration [5154]. The approach uses an attachment to the transducer that provides a xed trajectory for the needle. Advancement of the needle through the designated path ensures a predictable, uniform trajectory of the needle relative to the transducer. This stability may be particularly advantageous for inexperienced operators
Fig. 7. Ultrasound transducer with needle inserted through a mechanical guide.
756
because it incorporates control of the transducers placement with the needles angle of entry [55]. The use of mechanical guides has some notable disadvantages. It requires investing in an additional piece of equipment that makes large, linear transducers even bulkier. Mechanical guides restrict the angle of the needle and the skin entry point. This restriction prevents the operator from continuously redirecting the needle as may be needed in certain cases [56]. Lastly, the xed angle of entry may make some supercial structures dicult to access [57]. The largest study to evaluate needle-guided US was published by Denys et al [58]. This nonrandomized study reported a 100% success rate for US guidance among 928 patients. In comparison, the success rate for the landmark approach was 88% among 302 patients. With needle-guided US, there also was a signicant improvement in venous access time (9.8 versus 44.5 seconds), carotid puncture rate (1.7% versus 8.3%), brachial plexus irritation (0.4% versus 1.7%), hematoma development (0.2% versus 3.3%), and average number of attempts to success (1.3 versus 2.5). Free-hand method The alternative to using a needle-guide system is to perform the procedure using the free-hand technique. In this situation, the transducer and the advancing needle are positioned and stabilized with the operators or an assistants hands (Fig. 8). Continuous ne adjustments can be made in the needles direction and in the transducers view. Although it generally is considered to be a more technically demanding procedure, this approach oers more exibility for the operator. In addition, if an assistant is available, he or she can hold the transducer on a site distant from the needle entry point, thereby removing it from the sterile eld [57] and potentially increasing the speed with which the target vein can be cannulated.
Fig. 8. Free-hand technique.
757
Fig. 9. Short-axis approach to needle insertion.
Short-axis versus long-axis approach Another variable to consider is the axis of visualization for the target vessel. For the short-axis approach, the vessel is identied in the transverse plane and centered under the transducer (see Fig. 1). The midpoint of the transducer then becomes a reference point for insertion of the needle. The needle is inserted at a 45( angle to the transducer (Fig. 9). As the needle is advanced, the tip is visualized as it approaches the anterior wall of the vessel. After contacting the anterior wall of the vessel, further insertion of the needle will cause posterior displacement of the vessel wall (Fig. 10). A ash of blood in the syringe signies that the needle has entered the vessel. At this point, the transducer can be set aside and the rest of the procedure performed normally.
Fig. 10. Posterior displacement of vessel wall by advancing needle viewed in the short axis.
758
Fig. 11. IJV visualized in the long axis.
In contrast, the long-axis approach identies the vessel in its long axis and involves lining up the transducer over the greatest anteriorposterior diameter of the vessel (Fig. 11). The needle is then inserted through the skin just o one end of the transducer in a plane that is in line with the long axis of the transducer and at an approximate 30( angle to the skin surface (Fig. 12). As the needle is advanced, its progress through the subcutaneous tissue is monitored in real-time on the US screen (Fig. 13). After the needle has punctured the anterior wall of the vessel and a ash of blood is apparent in the syringe, the transducer can be set aside and the rest of the procedure completed normally. In the authors experience, the advantages of the short-axis over the longaxis approach are that it is easier to perform in anatomic areas where space
Fig. 12. Long-axis approach to needle insertion.
759
Fig. 13. Real-time visualization of needle through subcutaneous tissue.
is limited (eg, neck), inexperienced users can acquire prociency more quickly, and smaller vessels (eg, deep brachial vein) are more consistently visualized. On the other hand, performing the procedure along the vessels long axis allows much better visualization of the advancing needle tip and, therefore, may avoid inadvertent punctures of the posterior vessel wall (Fig. 14). A study that compared the short- and long-axis approaches with vascular access found that novice users could perform cannulation more quickly using the short-axis approach. There was, however, no statistically
Fig. 14. Needle visualized perforating posterior (post.) vessel wall.
760
signicant dierence in terms of mean diculty, number of skin breaks, and mean number of needle redirections [59]. Doppler method Historically, Doppler US for venous access has not been used in many American EDs, most likely because these EDs purchased real-time 2-D US machines. These multipurpose machines have been favored over purchasing additional equipment that is needed to implement the Doppler method. In the interim, research has shown 2-D real-time US to be superior to Doppler for central venous access [45]. The authors expect most future research on US-guided techniques to focus on 2-D real-time US.
Internal jugular vein approach Anatomic considerations The IJV is the most studied vein for US-guided catheterization. It usually lies anterior and slightly lateral to the carotid artery. It normally is larger in diameter than the carotid artery and expands in diameter with a Valsalva maneuver [60]. Many studies have documented its anatomic variations with regard to potential complications. Yonei et al [38] rst reported the use of real-time US guidance for IJV cannulation in 1986. In this series, central lines were placed successfully in 160 intensive care patients without complication. In 1991, Denys and Uretsky [61] reported a series of 200 patients who underwent IJV cannulation in the cardiac catheterization laboratory, coronary care unit, and intensive care unit. The investigators found anatomic anomalies of the IJV (small diameter, unresponsiveness to Valsalva maneuver, and unexpected lateral or medial displacement) in 8% of patients. Troianos et al [62] reported the largest case series for determining the anatomic relationship between the IJV and carotid artery. Among 1009 patients admitted for surgery, 54% had an IJV overlying the carotid artery, rather than coursing it laterally, as expected. This anomalous anatomy might predispose the patients to arterial punctures if the needle traversed the IJV. In a prospective series of 31 patients with known dicult central venous access, Hateld and Bodenham [29] reported a success rate of 100% using real-time US guidance for 22 patients. Among the remaining 9 patients, for whom indirect US guidance was used, 66% were cannulated successfully within three attempts. Of 23 patients who had been referred specically because of prior diculties with or complications from cannulation, 16 had an anatomic reason for diculty that was determined by US. Docktor et al [63] reported a 100% success rate with real-time US guidance in a prospective series of 150 patients referred for nonemergent central venous access. Using US, the investigators were able to document the phenomenon of double wall puncture among 30 patients. Double wall
761
puncture occurs in cases of an IJV with low pressure. The anterior wall is pushed against the posterior wall before the needle punctures it. If the carotid artery is located underneath the vein along the needle tract, then a double wall puncture can extend into the carotid artery. In this study, the IJV was visualized directly over the carotid artery in 25% of cases. The investigators surmised that this high rate was partially due to variable transducer positioning and dierent degrees of head rotation. The only complications were two cases of carotid artery puncture. These occurred among patients whose IJV was visualized to lie directly over the carotid artery. Based on their ndings, the investigators recommended using realtime US to visualize the anatomic relationship between the IJV and carotid artery and then determining the optimal needle track that would miss the carotid artery in the event of a double wall puncture. Denys et al [58] compared real-time US guidance that used a transducer needle guide with the landmark approach in 1230 patients who had IJV cannulation. Among patients in the US-guidance group, 3.4% had a right IJV that was not visualized or was deemed too small to cannulate. In all of these cases, the left IJV was successfully cannulated. The investigators also recognized the double wall phenomenon, commenting, the IJV is actually compressed completely by the needle before the vessel is penetrated. The needle must be advanced a little deeper and then retracted slightly to be positioned in the center of the lumen. This nding underscores the need to identify an underlying carotid artery. Evidence-based analysis Eleven of the 18 articles included in the 2003 NICE-sponsored metaanalysis investigated the IJV approach in adults [45]. Of these 11 studies, 7 used real-time US guidance and 4 used Doppler US guidance. Notable improvements over the landmark approach were demonstrated with realtime US guidance. The relative risk of failed rst attempts was 0.59 (95% condence interval: 0.390.88, P = 0.009) and the relative risk of failed catheter placement was 0.14 (95% condence interval: 0.060.33, P\0.0001). The relative risk of complications decreased by 57%, the mean number of attempts to successful cannulation decreased by 1.5, and the mean time to successful cannulation decreased by almost 70 seconds (all P\0.02). Subsequent to the meta-analysis, another randomized trial that compared real-time US with the landmark approach among intubated patients undergoing elective surgery was published [64]. This study investigated the use of indirect US guidance with two dierent frequency transducers: 7.5 MHz and 3.75 MHz. In measuring the number of successful rst attempts, mean number of attempts to success, and rate of complications, the investigators discovered no signicant dierences between the two frequencies. They noted that the 7.5-MHz transducer visualized structures with higher resolution but that the image quality of 3.75 MHz was acceptable
762
for the purpose of locating the IJV and CA. When the results of two US groups were pooled for comparison to the landmark group, the rate of successful rst attempts showed improvement (73% versus 86%, P\0.05); however, no statistically signicant dierences were found regarding the overall success or complication rates. The investigators noted that in a subset of 52 patients who lacked the anatomic landmark of respiratory jugular venodilation (visible bulging of the vein beneath the skin synchronized with inspiration of positive-pressure ventilation) on which they traditionally depended, US guidance was superior to the landmark approach for all outcomes. The number of successful rst attempts was almost tripled (86% versus 30%, P\0.001) and the number of successful cannulations was greatly improved (100% versus 78%, P\0.05). There were no complications among the US-guided attempts, whereas the landmark approach was associated with three carotid artery punctures. These results, although not surprising, further support the use of US guidance, especially for patients with poor percutaneous landmarks. Emergency medicine literature Hudson and Rose [39] rst reported the use of US guidance for IJV cannulation specically in the ED in 1997. In this article, they described their successful experience in 2 patients with challenging percutaneous landmarks due to severe skin graft scarring or morbid obesity. Since then, two prospective studies of ED patients have been published. Hrics et al [41] reported a small case series in which patients who needed central venous access within 1 hour of arrival to the ED underwent cannulation either with real-time or indirect US guidance. The success rates were 87.5% among the 8 patients in the real-time US group and 71% among the 24 patients in the indirect US group. Miller et al [33] have published the only trial as yet of real-time US versus a landmark approach in the ED setting. This trial was pseudorandomized. It used odd- and even-day assignment of 122 patients to real-time US guidance or to the landmark approach to the IJV, SV, FV, or peripheral vein. It did not analyze results by the type of vein cannulated; however, the largest subgroup (55%) of the US-guided cases were for IJV cannulation. The investigators noted an overall decrease in the mean time to successful cannulation and number of attempts when using US. These improvements occurred across the range of operator experience. Subclavian vein approach Anatomic considerations As with the IJV, the SV oers ideal size for central access. Its proximity to structures such as the lung, subclavian artery, and brachial plexus, however, can lead to signicant morbidity. Other challenges for accessing the SV with US guidance are its deeper location and the presence of the
763
clavicle bone. Because bone does not transmit US waves, placing the transducer over it provides no information to guide the operator. US guidance can be used to cannulate the SV in its midportion as is most commonly taught with the landmark approach. In this case, the apex of the lung can be less than 1 cm away from the SV [65]. Attempting US visualization of the SV in short axis also can be challenging in this location because it involves holding the transducer on top of the clavicle. Maintaining appropriate pressure for the transducer over the clavicle can be uncomfortable for the patient and may add to the technical diculty of this approach. Supraclavicular approach Due to its inherent anatomy, the supraclavicular approach to the SV is troublesome to achieve with US guidance. In most patients, little space is available for the transducer to be placed concurrently with needle insertion, which makes real-time US guidance dicult at best. Two alternatives for successful US-guided supraclavicular approach exist. The rst is to use the indirect method (previously described). Another alternative is to use the lowIJV approach. This approach has been found to be a safe and direct route to the superior vena cava and right atrium for US-guided central venous access. Silberzweig and Mitty [66] investigated 116 low-IJV punctures among 109 patients in the interventional radiology suite. They reported no complications and an average of 1.2 attempts (range, 13) needed for success. Axillary approach Yet another approach is to access the SV more laterally on the shoulder so that the needle cannulates the axillary vein. This more distal approach eliminates the problem of holding the transducer over the uneven surface of the clavicle and removes the potential for the placed catheter to be pinched between the subclavius muscle and the costoclavicular ligament complex associated with the standard approach [67]. It also decreases the risk of pneumothorax because the lung generally is farther away from the vascular structures in the lateral shoulder. A misplaced needle passing through the axillary vein will travel posteriorly through the axillary fat and layers of muscle and, nally, to the scapula, thereby missing the pleural space [68]. The landmark approach to the axillary has been demonstrated to be safe in adults [69] and critically ill infants and children [70]. Potential advantages for using the axillary approach specically in the ED include easier access for patients in cervical collars or with neck trauma. Using this more lateral approach in critical trauma patients may also allow for more ecient simultaneous management of the airway and acquisition of central access. In addition, the landmark approach to the axillary vein has been reported to be ecacious among severe burn victims who often present with burns of the face, neck, and proximal shoulders [71].
764
Drawbacks of the axillary approach include the decreased diameter and deeper location of the axillary vein compared with the SV. Smaller caliber and deeper location may lead to diculty visualizing it with a high-frequency transducer, especially in larger patients, and may increase the need for a longer catheter to reach the vena cava in larger patients. No studies specically comparing the axillary and SV approaches have been published. Evidence-based analysis Four studies of SV access were included in the 2003 NICE-sponsored meta-analysis [45]. Of these, three used Doppler US guidance and one used real-time US guidance. The real-time US study evaluated 53 cannulation attempts among 32 critically ill patients in a combined trauma and medical intensive care unit [55]. The investigators used the axillary vein approach for US guidance and reported an improved success rate compared with the landmark method (92% versus 44%, P = 0.003). With US guidance, there was also a decrease in the complication rate (4% versus 41%, P = 0.002), the mean number of attempts (1.4 versus 2.5, P = 0.0007), and the mean number of insertion kits used (1.0 versus 1.4, P = 0.0003). Malpositioned catheters in the landmark group also led to the need for additional chest radiographs. Fry et al [30] reported 100% success with the US-guided placement of 43 SV catheters in patients who had relative contraindications to the landmark approach. The investigators subjectively noted that awake patients who have US-guided access seem less apprehensive. The investigators further suggested that the ability to watch what is going on via the ultrasound video screen, a decrease in the number of attempts, and better local anesthesia along the intended needle path contribute to improved patient satisfaction with US guidance. Femoral vein approach Anatomic considerations Modern study of the FV has discovered variations from generally accepted anatomy. Reviewing CT scans of the pelvis in 100 patients, Baum et al [72] discovered that a portion of the FV and the femoral artery overlaps in the anteroposterior plane 65% of the time. A subsequent study that used US in 50 intensive care unit patients conrmed this nding: in most patients there was overlap of the artery over the vein far closer to the inguinal ligament than conventional anatomical textbooks would indicate [73]. Furthermore, landmarks were not predictive of the underlying anatomy that was documented on US. Evidence-based analysis Only one randomized trial of US guidance for FV access has been published. This trial was undertaken among 20 patients undergoing cardio-
765
pulmonary resuscitation in the ED [40]. Compared with the landmark approach, real-time US had a higher success rate (90% versus 65%, P = 0.058), a lower mean number of needle passes (2.3 3 versus 5.0 5, P = 0.006), and a lower rate of arterial catheterization (0% versus 20%, P = 0.025). The investigators suggested that the better performance of US was due to the ability to localize a nonpalpable FV by visualizing it instead. They also noted that chest compressions were associated with changes in FV diameter visualized on US. This nding is contrary to the expectation of arterial pulsations with chest compressions. It implies that palpating for a pulse during cardiopulmonary resuscitation as part of the landmark approach may be misleading. Furthermore, the color and pulsatility of returning blood may be unreliable for predicting arterial or venous source, especially in patients whose oxygen saturation is low [74]. This study demonstrated the ability of US to accurately identify the FV without dependence on these traditional signs. In a prospective series, Kwon et al [75] reported a 100% success rate among 28 patients who needed acute hemodialysis access. Compared with 38 patients with the landmark approach, these patients experienced a higher a rate of successful rst attempts (92.9% versus 55.3%, P\0.05) and improved mean total procedure time (45.1 18.8 versus 79.4 61.7 seconds, P\0.05). Femoral artery puncture occurred in 7.1% of US cases compared with 15.8% of landmark cases. Peripheral veins The subset of ED patients with poor peripheral access is well known to many emergency nurses and physicians. US oers a potential alternative to central venous access, surgical cutdowns, and blind, deep brachial vein catheterization for patients who need simple intravenous access but have no palpable or visible peripheral veins. Studies of peripherally inserted central venous catheter lines have shown that real-time US guidance is safe and successful in adult [76,77] and pediatric populations [78]. A case series of US-guided brachial and basilic vein cannulation among 100 ED patients with dicult intravenous access demonstrated a 91% overall success rate and a 73% rate of success on rst attempt [42]. Two cases of brachial artery puncture were reported. The mean time to successful cannulation was 77 129 seconds (range, 4600 seconds). No trials that have compared US guidance with the landmark approach have been published.
Issues in pediatric patients Procedural challenges Central venous access in infants and children is challenging under any circumstances. There are various possible reasons for the greater morbidity associated with central venous cannulation in pediatric patients: supercial
766
anatomic landmarks may be less distinct, vessel diameters are generally smaller, the proximity of important anatomic structures may be greater, there may be less patient cooperation than among adults, anatomic anomalies may be present, and nonpediatric specialists may not be as experienced with pediatric vascular access [46,79]. For these reasons, US guidance should be integrated into methods of central venous access for pediatric patients in the ED. Evidence for ultrasound guidance in pediatric patients Studies of real-time US guidance for central vein cannulation in pediatric patients currently are found mainly in the anesthesia literature. Three randomized trials of US guidance versus landmark method for IJV cannulation among infants and children have been published [4648]. These trials and at least four case series [34,7981] constitute the current body of evidence that supports the application of US guidance in pediatric patients. The 2003 NICE-sponsored meta-analysis used the three trials to determine overall relative risk reductions of 85% for failed placement and 73% for complications of IJV cannulation in pediatric patients [45]. Anatomic factors contribute to complications of the landmark approach to IJV cannulation in children. Alderson et al [46] determined that 18% of patients aged 3 days to 5.5 years had anatomic variations of the IJV. The investigators reported a superior success rate (100% versus 80%), shorter mean time to successful cannulation (27.4 versus 48.9 seconds), and lower mean number of attempts (1.37 versus 2.0) with US guidance. Two patients in the landmark group and one in the US group suered a carotid artery puncture. In comparing real-time US to the landmark approach for IJV cannulation among 95 infants aged 12 months or less who underwent elective cardiovascular surgery, Verghese et al [47] found that the US approach signicantly improved success and complication rates. US guidance was successful and had no associated carotid artery punctures. In contrast, the landmark method had a 77% success rate and a 25% rate of carotid artery puncture. Almost half of the patients with carotid artery puncture sustained additional complications, including hemothorax, pneumothorax, jugular venous hematoma, catheter kinking, and threading diculty. Among the subset of patients with unsuccessful landmark attempts, 25% were subsequently successfully catheterized under US guidance. In a subsequent study, Verghese et al [48] found statistically signicant improvements in the success rate and median number of attempts with realtime US guidance over the landmark approach. The small sample size of 16 patients in each group, however, limited the statistical evaluation of trends toward improved time to successful cannulation and complication rates. Three carotid artery punctures occurred using the landmark approach compared with one using real-time US guidance.
767
Limitations of pediatric emergency department ultrasound-guided venous cannulation The three existing trials have been criticized for their relatively small sample sizes of less than 100 patients each. Because these trials reported results only with IJV access, denitive conclusions regarding other venous sites presently are impossible. No studies have been reported on SV cannulation and only one small study reported the successful use of realtime US to facilitate FV catheterization [34]. Although the pediatric studies consistently have reported positive ndings for IJV cannulation, the importation of their success in the well-controlled, elective setting of scheduled surgery to the acute circumstances in the ED has yet to be demonstrated. To date, no studies of US-guided venous cannulation conducted in the pediatric ED setting have been published.
Limitations of emergency department ultrasound-guided venous access Transducer type Most US-guided vascular access is performed using a linear, highfrequency (610 MHz) transducer. The linear transducer provides a larger eld of view compared with a sector transducer. This larger eld of view allows visualization of the advancing needle through its entire course. In addition, because most of the target vessels are supercial, a high-frequency transducer can be used that yields superior resolution of the subcutaneous tissues, the advancing needle, and the vessels to be cannulated or avoided. As a separate purchase, the cost of a linear, high-frequency transducer can be prohibitive. Some EDs may not have budgeted for additional transducers when they originally acquired US equipment. Consequently, many emergency physicians do not have access to this transducer. One alternative is to use the large curvilinear transducer commonly used for abdominal imaging; however, this transducer has some drawbacks for real-time US guidance. Typically, the large curvilinear probe is bulkier than a linear transducer, and its lower-frequency images can make the procedure more technically challenging. The curvilinear transducers greatest obstacle is its curved visual eld. Although the center of the visual eld is relatively linear, the lateral aspects of the screen are curved to the extent that advancement of the needle under real-time US guidance is distorted. One possibility is to use the curvilinear transducer for indirect guidance, but this approach has not been formally studied. Another approach is to use an endovaginal transducer for US-guided vascular access. This transducer is a common component of many ED US systems and its use for venous access has been promoted in the gynecologic and emergency medicine literature [82,83]. It has been described only for the short-axis approach to the IJV, and ecacy was not studied. For physicians
768
with access to only the curvilinear or endovaginal transducer, the endovaginal transducer may be the superior choice. Sterile barrier Another equipment issue unique to US-guided vascular access is that a sterile barrier typically is needed. Such barriers usually are designed to cover both the transducer and its cable (see Fig. 4) and allow sterile performance by a single operator. Although there are many relatively inexpensive varieties of sterile transducer covers, occasionally, one may not be available. In this case, other alternatives can be employed. The easiest method is to use a sterile glove. A conducting agent is placed inside the glove wherever the largest uninterrupted at surface is located. An assistant can then place the transducer inside the glove. The operator then folds back the ngers of the glove and holds the transducer so that the at surface of the glove forms the scanning surface for the transducer. It is important to eliminate any air bubbles that may be interposed between the glove and the transducers scanning surface because they would compromise image quality severely. Mechanical guides Mechanical guides typically come in two forms. The rst is a built-in needle slot within the central or side portion of the transducer that directs the needle at a predetermined angle within the plane of view of the US beam. Another form is a separate guide that can be tted to a transducer (see Fig. 7). Presently, these guides are not interchangeable among dierent transducers. Most companies manufacture them for each linear transducer that they produce. Mechanical guides may not be necessary, especially for experienced operators [55]. Education and training aids A signicant issue pertaining to US-guided vascular access is the time and cost of training new operators. It is unfortunate that practical education for US-guided venous access currently is not available, and standard methods for teaching other US examinations, such as normal or dialysis models, cadavers, swine, or simulators, have signicant limitations. Because the procedure is invasive, practicing on normal or dialysis models is problematic. Although the anatomy of cadavers may demonstrate vascular structures well [84], the entry site would be revealed after the initial puncture, thereby limiting the educational benets for subsequent students. Swine or other animal models not only have unique vascular anatomy but ethical and cost issues also limit their use. Lastly, although simulation would seem to be an attractive alternative, only one vascular model currently exists. This model is limited to teaching peripheral vein cannulation using
769
landmarks, not US guidance [85]. A newer development in the area of USguided vascular access education is the use of phantoms. Phantoms are generally easy and inexpensive to produce. They simulate vessels well and, hence, the mechanics of US-guided cannulation. Time to cannulation and operator experience The time required to set up and complete the procedure commonly is considered a drawback to US-guided vascular access. As with any novel procedure, there is a learning curve; however, this curve has been shown to be short and steep [58]. Improved results have been demonstrated across the spectrum of operator experience [86]. Furthermore, due to fewer failed attempts, the average time to vessel cannulation is the same as or is decreased with US guidance versus the landmark technique. Hilty et al [40] compared US-guided FV cannulation with the landmark approach in 20 patients presenting in cardiac arrest. The average time to ash of blood under US guidance was 30.8 32 seconds versus 33.8 35 seconds for the landmark technique. Time to cannulation also was decreased with US guidance (121.0 60 versus 124.2 69 seconds). In another study that compared the short- and long-axis approaches on a US phantom, the mean time to cannulation was 2.36 minutes versus 5.02 minutes, which was statistically signicant [59]. Although it did not compare landmark and US-guided approaches, this study suggests that shorter intervals to vessel cannulation can be obtained, especially for inexperienced operators, when the vessel is approached in the short axis. Summary The evidence that supports the general application of US guidance for venous access in the ED has reached a critical mass. The increasing familiarity of emergency physicians with US and the recent focus on patient safety and clinical outcomes has intensied attention on the capacity for US to improve patient care in the ED. US guidance can increase the safety and eciency of venous access procedures and oers improved outcomes. The potential for these improvements is compelling, especially among certain types of ED patients such as those with dicult or complicated access. Varying levels of evidence support the use of US guidance over the traditional landmark approach for venous access in adult and pediatric populations and for central and peripheral veins. Many dierent techniques may be applied, depending on the clinical situation and equipment available. References
[1] Randolph AG, Cook DJ, Gonzales CA, Pribble CG. US guidance for placement of central venous catheters: a meta-analysis of the literature. Crit Care Med 1996;24:20538.
770
[2] Tripathi M, Tripathi M. Subclavian vein cannulation: an approach with denite landmarks. Ann Thorac Surg 1996;61:23840. [3] Simpson ET, Aitchison JM. Percutaneous infraclavicular subclavian vein catheterization in shocked patients: a prospective study in 172 patients. J Trauma 1982;22:7814. [4] Scalea TM, Sinert R, Duncan AO, Rice P, Austin R, Kohl L, et al. Percutaneous central venous access for resuscitation in trauma. Acad Emerg Med 1994;1:52531. [5] Williams MR, Dunn EL. Percutaneous central venous cathetersa continuum of use. J Emerg Med 1985;2:3359. [6] Westfall MD, Price KR, Lambert M, Himmelman R, Kacey D, Dorevitch S, et al. Intravenous access in the critically ill trauma patient: a multicentered, prospective, randomized trial of saphenous cutdown and percutaneous femoral access. Ann Emerg Med 1994;23:5415. [7] Parsa MH, Tabora F. Central venous access in critically ill patients in the emergency department. Emerg Med Clin N Am 1986;4:70944. [8] Swanson RS, Uhlig PN, Gross PL, McCabe CJ. Emergency intravenous access through the femoral vein. Ann Emerg Med 1984;13:2447. [9] Getzen LC, Pollak EW. Short-term femoral vein catheterization: a safe alternative venous access? Am J Surg 1979;138:8758. [10] Emerman CL, Bellon EM, Lukens TW, May TE, Eron D. A prospective study of femoral versus subclavian vein catheterization during cardiac arrest. Ann Emerg Med 1990;19: 2630. [11] Dronen S, Thompson B, Nowak R, Tomlanovich M. Subclavian vein catheterization during cardiopulmonary resuscitation: a prospective comparison of the supraclavicular and infraclavicular percutaneous approaches. JAMA 1982;247:322730. [12] Durbec O, Viviand X, Potie F, Vialet R, Albanese J, Martin C. A prospective evaluation of the use of femoral catheters in critically ill adults. Crit Care Med 1997;25:19869. [13] Chiang VW, Baskin MN. Uses and complications of central venous catheters inserted in a pediatric emergency department. Pediatr Emerg Care 2000;16:2302. [14] Brunnette DD, Fischer R. Intravascular access in pediatric cardiac arrest. Am J Emerg Med 1988;6:5779. [15] Newman BM, Jewett TC Jr, Karp MP, Cooney DR. Percutaneous central venous catheterization in children: rst line choice for venous access. J Pediatr Surg 1986;21:6858. [16] Casado-Flores J, Barja J, Martino R, Serrano A, Valdvielso A. Complications of central venous catheterization in critically ill children. Pediatr Crit Care Med 2001;2:5762. [17] Sznajder JI, Zveibil FR, Bitterman H, Weiner P, Bursztein S. Central vein catheterization: failure and complication rates by three percutaneous approaches. Arch Intern Med 1986; 146:25961. [18] Bo-Lin GW, Andersen DJ, Andersen KC, McGoon MD. Percutaneous central venous catheterization performed by medical house ocers: a prospective study. Cathet Cardiovasc Diagn 1982;8:239. [19] Jastremski MS, Matthias HD, Randell PA. Femoral venous catheterization during cardiopulmonary resuscitation: a critical reappraisal. J Emerg Med 1984;1:38791. [20] Manseld PF, Hohn DC, Fornage BD, Gregurich MA, Ota DM. Complications and failures of subclavian-vein catheterization. N Engl J Med 1994;331:17358. [21] Goldfarb G, Lebrec D. Percutaneous cannulation of the internal jugular vein in patients with coagulopathies: an experience based on 1000 attempts. Anesthesiology 1982;56:3213. [22] Merrer J, De Jonghe B, Golliot F, Lefrant J, Ray B, Barre E, et al. Complications of femoral and subclavian venous catheterization in critically ill patients. JAMA 2001;286: 7007. [23] Cook TL, Dueker CW. Tension pneumothorax following internal jugular cannulation and general anesthesia. Anesthesiology 1976;45:5545. [24] McEnany MT, Austen WG. Life-threatening hemorrhage from inadvertent cervical arteriotomy. Ann Thorac Surg 1997;24:2336.
771
[25] Mason MS, Wheeler JR, Jae AH, Gregory RT. Massive bilateral hydrothorax and hydromediastinum: an unusual complication of percutaneous internal jugular vein cannulation. Heart Lung 1980;9:8836. [26] Parikh RK. Horners syndrome: a complication of percutaneous catheterisation of internal jugular vein. Anaesthesia 1972;27:3279. [27] Digby S. Fatal respiratory obstruction following insertion of a central venous line. Anaesthesia 1994;40:10134. [28] Johnson FE. Internal jugular vein catheterization: prospective study. N Y State J Med 1978;78:216871. [29] Hateld A, Bodenham A. Portable US for dicult central venous access. Br J Anaesth 1999;82:8226. [30] Fry WR, Clagett GC, ORourke PT. US-guided central venous access. Arch Surg 1999; 134:73841. [31] Gilbert TB, Sene MG, Becker RB. Facilitation of internal jugular venous cannulation using an audio-guided Doppler US vascular access device: results from a prospective, dual center, randomized, crossover clinical study. Crit Care Med 1995;23:605. [32] Bold RJ, Winchester DJ, Madary AR, Gregurich MA, Manseld PF. Prospective, randomized trial of Doppler-assisted subclavian vein catheterization. Arch Surg 1998;133: 108993. [33] Miller AH, Roth BA, Mills TJ, Woody JR, Longmoor CE, Foster B. US guidance versus the landmark technique for the placement of central venous catheters in the emergency department. Acad Emerg Med 2002;9:8005. [34] Sheridan RL, Petras L, Lydon M. Ultrasonic imaging as an adjunct to femoral vein catheterization in children. J Burn Care Rehabil 1997;18:1568. [35] Cook T, Roepke T. Prevalence and structure of US curricula in emergency medicine residencies. J Emerg Med 1998;16:6557. [36] Ullman JI, Stoelting RK. Internal jugular vein location with the US Doppler blood ow detector. Anesth Analg 1978;57:118. [37] Legler D, Nugent M. Doppler localization of the internal jugular vein facilitates central venous cannulation. Anesthesia 1984;60:4812. [38] Yonei A, Nonoue T, Sari A. Real-time ultrasonic guidance for percutaneous puncture of the internal jugular vein. Anesthesiology 1986;64:8301. [39] Hudson PA, Rose JS. Real-time ultrasound-guided internal jugular vein catheterization in the emergency department. Am J Emerg Med 1997;15:7982. [40] Hilty WM, Hudson PA, Levitt MA, Hall JB. Real-time US-guided femoral vein catheterization during cardiopulmonary resuscitation. Ann Emerg Med 1997;29:3317. [41] Hrics P, Wilber S, Blanda MP, Gallo U. US-assisted internal jugular catheterization in the ED. Am J Emerg Med 1998;16:4013. [42] Keyes LE, Frazee BW, Snoey ER, Simon BC, Christy D. US-guided brachial and basilic vein cannulation in emergency department patients with dicult intravenous access. Ann Emerg Med 1999;34:7114. [43] American College of Emergency Physicians. Use of US imaging by emergency physicians. Ann Emerg Med 2001;38:46970. [44] Agency for Healthcare Research and Quality. Making health care safer: a critical analysis of patient safety practices [summary]. Evid Rep Technol Assess 2001;43:ix, 1668. [45] Hind D, Calvert N, McWilliams R, Davidson A, Paisley S, Beverley C, et al. Ultrasonic locating devices for central venous cannulation: meta-analysis. BMJ 2003;327:3617. [46] Alderson PJ, Burrows FA, Stemp LI, Holtby HM. Use of ultrasound to evaluate internal jugular vein anatomy and to facilitate central venous cannulation in pediatric patients. Br J Anaesth 1993;70:1458. [47] Verghese ST, McGill WA, Patel RI, Sell JE, Midgely FM, Ruttiman UE. Ultrasoundguided internal jugular venous cannulation in infants: a prospective comparison with the traditional palpation method. Anaesthesiology 1999;91:717.
772
[48] Verghese ST, McGill WA, Patel RI, Sell JE, Midgely FM, Ruttiman UE. Comparison of three techniques for internal jugular vein cannulation in infants. Paediatr Anaesth 2000;10: 50510. [49] Nemcek AA. The use of ultrasound as an adjunct to the performance of vascular procedures. J Vasc Interv Radiol 1996;7:86975. [50] Nadig C, Leidig M, Schmiedeke T, Hoken B. The use of ultrasound for placement of dialysis catheters. Nephrol Dial Transplant 1998;13:97881. [51] Williamson RA, Varner MW, Weiner CP. Use of needle guide to improve sonographically directed amniocentesis. Am J Obstet Gynecol 1984;149:1078. [52] Williamson RA, Varner MW, Grant SS. Reduction in amniocentesis risks using a real-time needle guide procedure. Obstet Gynecol 1985;65:7515. [53] Petrikovsky B, Schneider EP, Klein VR, Wyse LJ. Cordocentesis using the combined technique: needle guide-assisted and free-hand. Fetal Diagn Ther 1997;12:2524. [54] Hatada T, Ishii H, Ichii S, Okada K, Yamamura T. Ultrasound-guided ne-needle aspiration biopsy for breast tumor: needle guide versus freehand technique. Tumori 1999; 85:124. [55] Gualtieri E, Deppe SA, Sipperly ME, Thompson DR. Subclavian venous catheterization: greater success rate for less experienced operators using ultrasound guidance. Crit Care Med 1995;23:6927. [56] Parker SH, Stavros AT, Dennis MA. Needle biopsy techniques. Radiol Clin N Am 1995; 33:117186. [57] Matalon TAS, Silver B. US guidance of interventional procedures. Radiology 1990;174: 437. [58] Denys BG, Uretsky BF, Reddy PS. Ultrasound-assisted cannulation of the internal jugular vein: a prospective comparison to the external landmark-guided technique. Circulation 1993;87:155762. [59] Blaivas M, Brannam L, Fernandez E. Short axis versus long axis approaches for teaching ultrasound guided vascular access [abstract]. Acad Emerg Med 2003;10:572b. [60] Weissleder R, Elizondo G, Stark DD. Sonographic diagnosis of subclavian and internal jugular vein thrombosis. J Ultrasound Med 1987;6:57787. [61] Denys BG, Uretsky BF. Anatomical variations of internal jugular vein location: impact on central venous access. Crit Care Med 1991;19:15169. [62] Troianos CA, Kuwik RJ, Pasqual JR, Lim AJ, Odasso DP. Internal jugular vein and carotid artery anatomic relation as determined by ultrasonography. Anesthesiology 1996; 85:438. [63] Docktor B, So B, Saliken JC, Gray R. Ultrasound monitoring in cannulation of the internal jugular vein: anatomic and technical considerations. Can Assoc Radiol J 1996;47: 195201. [64] Hayashi H, Amano M. Does ultrasound imaging before puncture facilitate internal jugular vein cannulation? Prospective randomized comparison with landmark-guided puncture in ventilated patients. J Cardiothorac Vasc Anesth 2002;16:5725. [65] Skolnick ML. The role of sonography in the placement and management of jugular and subclavian central venous catheters. AJR 1994;163:2915. [66] Silberzweig JE, Mitty HA. Central venous access: low internal jugular vein approach using imaging guidance. AJR 1998;170:161720. [67] Galloway S, Bodenham A. Ultrasound imaging of the axillary veinanatomical basis for central venous access. Br J Anaesth 2003;90:58995. [68] Nickalls RWD. A new percutaneous infraclavicular approach to the axillary vein. Anaesthesia 1987;42:1514. [69] Taylor BL, Yellowlees I. Central venous cannulation using the infraclavicular axillary vein. Anesthesiology 1990;72:558. [70] Metz RI, Lucking SE, Chaten FC, Williams TM, Mickell JJ. Percutaneous catheterization of the axillary vein in infants and children. Pediatrics 1990;85:5313.
773
[71] Andel H, Rab M, Felfernig M, Andel D, Koller R, Kamolz LP, et al. The axillary vein central venous catheter in severely burned patients. Burns 1999;25:7536. [72] Baum PA, Matsumoto AH, Teitelbaum GP, Zuuribier RA, Barth KH. Anatomic relationship between the common femoral artery and vein: CT evaluation and clinical signicance. Radiology 1989;173:7757. [73] Highes P, Scott C, Bodenham A. Ultrasonography of the femoral veins in the groin: implications for vascular access. Anesthesia 2000;55:1198202. [74] Todd MR, Barone JE. Recognition of accidental arterial cannulation after attempted central venipuncture. Crit Care Med 1991;19:10813. [75] Kwon TH, Kim YL, Cho DK. Ultrasound-guided cannulation of the femoral vein for acute haemodialysis access. Nephrol Dial Transplant 1997;12:100912. [76] Sofocleous CT, Schur I, Cooper SG, Quintas JC, Brody L, Shelin R. Sonographically guided placement of peripherally inserted central venous catheters: review of 355 procedures. AJR 1998;170:16136. [77] Chrisman HB, Omary RA, Nemcek AA, Ryu RK, Saker MB, Vogelzang RL. Peripherally inserted central venous catheters: guidance with the use of US versus venography in 2650 patients. J Vasc Interv Radiol 1999;10:4735. [78] Donaldson JS, Morello FP, Junewick JJ, ODonovan JC, Lim-Dunham J. Peripherally inserted central venous catheters: US-guided vascular access in pediatric patients. Radiology 1995;197:5424. [79] Etheridge SP, Berry JM, Krabill KA, Braunlin EA. Echocardiographic-guided internal jugular venous cannulation in children with heart disease. Arch Pediatr Adolesc Med 1995; 149:7780. [80] Asheim P, Mostad U, Aadahl P. Ultrasound-guided central venous cannulation in infants and children. Acta Aneasthesiol Scand 2002;46:3902. [81] Liberman L, Hordof AJ, Hsu DT, Pass RH. Ultrasound-assisted cannulation of the right internal jugular vein during electrophysiologic studies in children. J Inter Card Electrophysiol 2001;5:1779. [82] Jelsema RD, Deppe G, Isada NB. Vaginal probe ultrasound guidance for internal jugular catheterization. J Gynecol Surg 1992;8:2435. [83] Phelan MP. A novel use of the endocavity (transvaginal) ultrasound probe: central venous access in the ED. Am J Emerg Med 2003;21:2202. [84] Nip IL, Haruno MM. A systematic approach to teaching insertion of a central venous line. Acad Med 2000;75:552. [85] Reznek MA, Rawn CL, Krummel TM. Evaluation of educational eectiveness of a virtual reality intravenous insertion simulator. Acad Emerg Med 2002;9:131925. [86] Geddes CC, Walbaum D, Fox JG, Mactier RA. Insertion of internal jugular hemodialysis cannulae by direct US guidancea prospective comparison of experienced and inexperienced operators. Clin Nephrol 1998;50:3205.
Ultrasound diagnosis of deep venous thrombosis

Jason A. Tracy, MD, Jonathan A. Edlow, MD*
Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Harvard Aliated Emergency Medicine Residency, Harvard Medical School, West Clinical Center 2, One Deaconess Road, Boston, MA 02215, USA
Deep venous thrombosis (DVT) is common in the emergency department (ED). Concern for DVT results in over 500,000 lower extremity ultrasound (US) examinations being ordered each year [1]. The true incidence of DVT and pulmonary embolism (PE) is dicult to assess due to the fact that they frequently are misdiagnosed. Up to 60% of untreated DVTs progress to PE, which results in over 50,000 deaths each year [24], underscoring the importance of correct diagnosis of DVT. Physical examination ndings alone are not accurate in the diagnosis of DVT [5]. US traditionally is the diagnostic method of choice, but US performed by the radiologist or US technologist is not always available to the emergency physician (EP). The overall time to perform a complete US by an on-call technologist can be hours when travel time, transport, and interpretation are included. Further, when on-call services are not available, patients needlessly might be admitted and possibly anticoagulated. Even with the advent of low-molecular-weight heparin, this process is costly, potentially risky, and cumbersome. Only 30% of patients sent to the ED to be evaluated for DVT have that diagnosis [68]. Because most PE deaths occur within the rst 1 to 24 hours, rapid, accurate diagnosis in the ED is critical [7,9]. Because the evaluation of lower extremity DVT in the symptomatic outpatient incorporates more than just a US, a discussion of the dierent diagnostic modalities follows. The authors pay particular attention to the combined use of pretest probability (how likely the patient is to have a DVT), D-dimer testing, and ED US.
* Corresponding author. E-mail address: jedlow@bidmc.harvard.edu (J.A. Edlow). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.008
776
J.A. Tracy, J.A. Edlow / Emerg Med Clin N Am 22 (2004) 775796
Venous anatomy of the lower extremity The lower extremity venous system starts at the junction of the external iliac vein and the common femoral vein as it passes under the inguinal ligament (Fig. 1). US imaging starts at this easy-to-nd landmark. When this area is imaged, it becomes clear that the femoral vein does not always lie medial to the artery as traditionally taught. Commonly, the femoral vein lies deep to the femoral artery. As one moves distally in the leg, the femoral vein bifurcates into the deep femoral vein and the supercial femoral vein. Although called supercial, the supercial femoral vein is the main deep vein of the lower leg, and clot within this vein is considered a DVT. Following the supercial femoral vein distally becomes dicult at approximately two thirds of the way to the knee as it dives deep into the adductor
Fig. 1. Normal venous anatomy of the lower extremity.
777
(Hunters) canal before it emerges behind the knee as the popliteal vein. Here, the popliteal vein lies supercial to the popliteal artery. Again, following the popliteal vein distally leads to its trifurcation into the anterior tibial, posterior tibial, and peroneal veins. The venous system distal to this trifurcation is considered the calf, and imaging of these veins usually is not performed in the limited ED US.
Pretest probability Most recent literature incorporates the probability of a patient having a DVT into clinical algorithms. This process of prescreening patients into those who probably have a DVT and those who probably do not have a DVT improves the sensitivity of studies in the lower risk population signicantly. This pretest probability is derived from a clinicians intuition or by means of a validated clinical scoring system. Both are derived from use of the historical, epidemiologic, and physical examination risk factors. The best-known clinical scoring system was developed by Wells et al [10] and has been validated by other authors [1116]. This scoring system assigns a numeric score to patients based on a certain set of validated DVT risk factors (Table 1). Patients are determined to be at low, medium, or high risk. The incidence of DVT in these groups ranges from 3% to 13%, 17% to 38%, and 60% to 75%, respectively [10,1315]. When clinical intuition and this clinical scoring system are compared, intuition underestimates the risk of DVT by as much as 38% [17]. This underestimation can lead to the
Table 1 Clinical Model for Predicting the Pretest Probability of Deep-Vein Thrombosis. (From Wells PS, Anderson DR, Bormonis J, et al: Value of assessment of pretest probability of deep-vein thrombosis in clinical management. Lancet 350:1796, 1997; with permission.) Clinical Feature Score
Active cancer (treatment ongoing or within the previous six 1 months or palliative) Paralysis, paresis, or recent plaster immobilization of the lower 1 extremities Recently bedridden for more than 3 days or major surgery 1 within 4 weeks Localized tenderness along the distribution of the deep venous 1 system Entire leg swollen 1 Calf swelling by more than 3 cm when compared with the 1 asymptomatic leg (measured 10 cm below the tibial tuberosity) Pitting edema (greater in the symptomatic leg) 1 Collateral supercial veins (non-varicose) 1 Alternative diagnosis as likely or greater than that of deep-vein 2 thrombosis Low probability 0 or less, moderate probability 12, high probability 3 or more.
778
misuse of screening tests, which can result in over treatment or underdiagnosis of DVT. Methods of deep venous thrombisis diagnosis Various methods for diagnosing DVT have been studied (Table 2). Contrast venography is touted as the gold standard in the diagnosis of DVT, with a sensitivity of 99% [18,19]. Contrast venography, however, rarely is used, is costly to perform, and requires a specialized radiology suite. Complications include postinjection DVT in up to 2% of patients [20], dye extravasation, supercial phlebitis, and allergic reactions [20,21]. A main advantage of venography is its accurate, direct detection of calf, iliac, and inferior vena cava DVTsareas where US lacks sensitivity [20,21]. MRI has been investigated as a noninvasive means of diagnosing DVT. Although costly, MRI has many advantages: it is safe in pregnancy; it can distinguish between acute and chronic DVT, which is not done well by US [2226]; the sensitivities in diagnosis of calf and pelvic DVT approach 100% (also problem areas for US) [23,24,2628]; and MRI has the advantage of diagnosing other causes of symptoms due to pelvic masses or lymph node enlargement [29]. Recent studies have used helical CT pulmonary angiography to image the lower extremity veins during the CT evaluation of PE, with results similar to US [3034]. This promising technique allows a single study to diagnose both DVT and PE. In addition to the potential time and cost savings, CT venography also allows diagnosis of the elusive pelvic DVT [30]. As these new imaging modalities are starting to be used clinically, previously studied brinogen scanning, impedance plethysmography, and radionuclide venography are now used rarely in routine clinical practice.
Table 2 Methods of DVT Diagnosis Test Contrast Venography Impedance Plethysmography Magnetic Resonance Imaging Advantages Gold standard Sensitivity approaches 100% Easily interpretable Non-invasive Disadvantages Invasive Requires specialized equipment Rare, but serious side eects Sensitivity inadequate Expensive Not readily available Limited data
Highly accurate Safe during pregnancy Non-invasive Computed Tomography Non-invasive Can diagnose pelvic DVT Concurrently exclude PE Ultrasonography Highly accurate Non-invasive D-Dimer Rapid laboratory study Can aide in exclusion of DVT
Not accurate for calf or pelvic DVT Complete study is time consuming Only used to rule-out DVT
779
Many recent clinical algorithms have incorporated D-dimer use into the evaluation of DVT. During the formation of a thrombus, there is a concurrent degradation of brin by plasmin, with subsequent release of D-dimer [35,36]. D-dimer levels can be elevated in other conditions including sepsis, recent myocardial infarction or stroke, recent surgery or trauma, disseminated intravascular coagulation, pregnancy or recent delivery, active collagen vascular disease, metastatic cancer, and liver disease [36,37]. Used as a screening test due to its high negative predictive value (NPV), D-dimer evaluation can be used only to rule out the diagnosis of DVT, particularly when combined with other diagnostic modalities including pretest probability and US. There are numerous types of D-dimers available, making comparison of studies dicult and careful incorporation into clinical practice essential. The ELISA is the gold standard in D-dimer measurement. Rapid ELISA tests now are widely available and provide results in less than 1 hour; the results are similar to the gold standard ELISA [35,3843]. Second-generation latex microparticle assays now are on the market, with results similar to the ELISA, making these assays a potential alternative to the rapid ELISA. In a prospective multicenter study of 556 patients, the MDA D-dimer assay (bioMerieux, Durham, North Carolina) had a sensitivity of 98.2% and an NPV of 99.7%, regardless of the patients pretest probability. In patients with a low-to-moderate pretest probability, anticoagulation and US were withheld. Only 1 patient was found to have a DVT during a 90-day follow-up [12]. Similar results were found by Schutgens et al [15] in a group of 812 patients; 1 of 176 patients (NPV, 97.7%) with a lowto-moderate pretest probability developed thrombosis. Another widely studied D-dimer assay is the whole-blood agglutination assay (SimpliRED, Agen Biomedical LTD, Brisbane, Australia). This assay uses a bispecic antibody to D-dimer and red blood cells and is a rapid, qualitative test performed at the bedside with a drop of blood. Although this assay has a lower sensitivity (as low as 65% in one study [44]), when combined with pretest probability, impedance plethysmography, or US as part of a clinical algorithm, NPVs as high as 100% can be achieved [11,37,4552].
Ultrasound US is the most accurate and extensively studied noninvasive test for the diagnosis of DVT; it essentially has replaced venography in the initial evaluation of outpatient DVT. The concept is simple: failure of veins to compress is indirect evidence of the presence of thrombus. This simple and widely available technique is referred to as B-mode imaging. Other methods of US examination are potentially available to the EP. Doppler US examines ow within a vessel and converts this to an audible
780
signal: a low-pitched venous hum or a higher pitched arterial ow pattern. Indirect evidence of DVT includes the absence of ow, the absence of respiratory variation in ow, and decreased augmentation of ow with compression of distal vessels [53]. Color ow Doppler imaging intuitively is easy to use. Color Doppler displays frequency shifts as color superimposed on a gray-scale US image and shows directional velocity in a colored representation: ow toward the transducer is red, whereas ow away from the transducer is blue. With the probe orientated correctly, arterial ow looks red and venous ow is blue. Color Doppler allows identication of vessels in technically dicult patients such as those with obesity and edema. When Doppler US is combined with B-mode imaging, it is referred to as duplex US. Duplex US is not required for evaluation of the lower extremity because noncompressibility that is visible on B-mode imaging is the most important requirement for diagnosis of DVT [54,55]. The sensitivity and specicity of compression US for proximal DVT range from 93% to 100% and 97% to 100%, respectively [7,5457]. In a large review of US for DVT, Kearon et al [58] found that US had an overall sensitivity of 89%, a specicity of 94%, a positive predictive value of 94%, and an NPV of 90%. For a proximal DVT, these results improve dramatically, with a sensitivity of 95%, a specicity of 96%, a positive predictive value of 97%, and an NPV of 98%. Studies have found sensitivities as low as 33% to 70% in the US diagnosis of calf DVT [1,20,54,55,57,59,60], with as many as 20% of these clots extending proximally [7,56,61,62]. Some recommend a follow-up US in symptomatic patients whose rst US is negative, especially if they have a moderate-tohigh suspicion of having a DVT [63]. The recommended time for follow-up US from the rst examination varies from 3 to 14 days, with most averaging 7 days [1,7,61,6466]. This strategy reduces the risk of thromboembolism to less than 2% at 6 months [58], with some studies reporting a failure rate of less than 0.7% [1,65]. US examination in the ED has disadvantages other than its poor sensitivity in the diagnosis of calf DVT. US can be nondiagnostic in up to 16% of patients when performed by an EP [67]. False-negative results can occur in those with clot in the adductor canal or with complete occlusion of the femoral vein. This obstruction causes ow in dilated collateral veins, giving the appearance of a patent vein [7]. US is operator dependent and can be technically dicult in the obese patient or the patient with signicant lower extremity edema. Because US cannot image pelvic DVTs, the results (no matter who is performing the US), should be used with caution in the pregnant patient with lower extremity swelling because there is a higher prevalence of pelvic DVT in these patients. Evaluation of patients with a history of DVT, particularly those on chronic anticoagulation, is dicult to assess with US because acute and chronic thrombi appear similar. For this reason, most of the studies discussed in this article have excluded patients with previous DVT or those on chronic anticoagulation [55,63,68].
781
Limited ultrasound Since the initial use of US for the diagnosis of DVT, debate has ensued regarding the specic deep veins and which portions must be imaged. The deep venous system extends from the groin to the lower calf (see Fig. 1), and imaging of the entire deep venous system is time consuming and technically dicult. A complete examination requires imaging of every few centimeters of the deep vein for compression, quality of blood ow, and phasicity with breathing. A simplied (or limited) US only images the deep veins with Bmode compression in the groin and popliteal fossa, reducing the time of the examination. Lensing et al [55] found this practice to be an eective means of decreasing US time while maintaining 100% sensitivity in diagnosing proximal DVT. In all 220 symptomatic outpatients (including those with calf DVT), the sensitivity and specicity were 91% and 99%, respectively. Again, the sole criterion for diagnosis of a DVT was the lack of compression of the common femoral vein or the popliteal vein. In a larger study by Birdwell et al [1], 405 symptomatic outpatients were evaluated with the same limited approach, but these investigators withheld any DVT treatment based on the US results and had the patients return in 5 to 7 days for reimaging. Over the follow-up period, no patients died or developed symptoms of PE. Of the 342 who initially had normal limited US, 7 had evidence of DVT on repeat testing. At 30-day follow-up (100% of patients followed), there were two cases of thromboembolism (0.6%), for an NPV of 99.4%. Schutgens et al [15] recently studied the combination of limited US, quantitative D-dimer, and pretest probability. Every patient had a D-dimer level and a Wells pretest probability score calculated at presentation. Group 1 had a low-to-moderate pretest probability and a negative D-dimer, with no further interventions. In group 2, those with a high pretest probability and a negative D-dimer had a single limited US performed. Finally, in group 3, those with positive D-dimer (regardless of pretest probability) had serial limited US performed at presentation and on day 7. Of the 812 patients studied, this strategy had an NPV of 99.4% for group 1, 97.2% for group 2, and 97.9% for group 3. The criterion for diagnosis of DVT by US was simply a lack of compression at the groin or knee by standard B-mode compression US. Tick et al [16] prospectively studied the simple, limited US and applied it based on pretest probability in 811 patients. Those with a low pretest probability (by the Wells criteria) received a limited US and those with a moderate-to-high pretest probability and a normal US had a SimpliRed D-dimer evaluation performed. In the low-probability group, 5 patients (2%) developed venous thromboembolism. No patients in the moderate-tohigh group with a negative limited US and D-dimer had adverse outcomes. In those with an abnormal D-dimer (in the moderate-to-high group), two
782
adverse events occurred following a normal serial limited US at day 8. This trial again documents the utility of limited compression US for use in the ED. The time savings of the limited US also has been studied. Poppiti and colleagues [69] studied limited US in a group of 72 patients and found that it had a sensitivity of 100% and a specicity of 98% for the diagnosis of DVT. In addition, these investigators demonstrated a dramatic time reduction from 37 to 5.5 minutes for limited US compared with complete duplex US. In a retrospective review by Pezzullo et al [70], limited US decreased examination time by 54%. As EPs attempt to decrease a patients length of stay, it is clear that the limited US achieves this goal. As is true with the complete lower extremity US, there are variations in study design with respect to which veins should be imaged with the limited US. Although the previously mentioned studies looked at only the common femoral vein and popliteal vein locations [1,55], some studies incorporate an additional view of the supercial femoral vein [57,7173]. To evaluate the necessity of this additional US view, Cogo et al [18] reviewed 562 venograms and found that there were no isolated supercial femoral vein DVTs. Further, in their review of the literature, Rosen and McArdle [74] reported a collective risk of an isolated supercial femoral vein DVT to be 2.4%. Although a prospective comparison of the two-view versus the three-view technique has not been performed, there is no data to support using the three-view technique over the two-view technique. The radiology community is not in full agreement as to whether a limited US should be performed at all. Frederick et al [75] evaluated 721 patients with suspicion of lower extremity DVT referred to a vascular laboratory. These investigators performed a complete US of the lower extremity and recorded clot location. Because 21% of patients had a DVT isolated to a single vein (including the popliteal vein and supercial femoral vein), this group of radiologists believed that a limited US should not be performed. This study did not evaluate whether a limited US would have detected these isolated DVTs and was not a prospective evaluation of management of DVT based on US results; therefore, its applicability is limited. Imaging of bilateral lower extremities in the symptomatic patient with unilateral symptoms is not required because the risk of DVT in the contralateral leg is less than 1% [74]. Imaging of the contralateral leg, however, can assist when anatomic interpretation in the aected leg is dicult. In the patient with bilateral lower extremity swelling, debate exists as to whether DVT is possible: Naidich et al [76] reported up to 24% of these patients had a DVT, whereas Sheiman and McArdle [77] found no DVTs. Rosen and McArdle [74] reviewed these studies and found that the main dierence between the two groups was the high prevalence of cancer in the Naidich study (41% versus 12%). With limited US being simple to perform, it seems reasonable to image both extremities in the ED, although this practice has not been studied specically.
783
Additional indirect evidence of DVT includes echogenicity of the venous contents and venous distention of the deep veins with Valsalvas maneuver. Although it intuitively makes sense that combining these methods of diagnoses with B-mode US will increase sensitivity and specicity, the data do not support using these potentially time-consuming techniques. Lensing et al [55] found echogenicity to be 100% sensitive but only 52% specic, with turbulent ow within the vein being the main cause of a false-positive result. They also found that Valsalvas maneuver, when imaging the common femoral and popliteal veins, produced a sensitivity and specicity of 55% and 67%, respectively. Similar results were found by Appelman et al [54], with no increase in accuracy over compression alone when combined with evaluation of clot echogenicity and Valsalvas maneuver. Because of this and the fact that the data support the use of limited B-mode compression US, many investigators do not recommend evaluating for echogenicity and venous distention [54,55].
Ultrasound performed by emergency physicians The use of EP-performed US continues to expand, particularly now that emergency medicine residencies are required to train residents in its use [78]. There is evidence forming that EPs reliably can perform limited US in the ED for numerous indications, as is discussed in this issue. The evaluation of the lower extremity in the patient with a suspected DVT is an area where EP-specic data is limited but quickly expanding. Initial work by Jolly et al [79] retrospectively assessed the ability of two EPs to evaluate patients with suspected DVT who presented to the ED in o hours. These physicians were trained in a surgical vascular laboratory and successfully performed 25 to 30 examinations during the training period. They evaluated only 23 patients with color Doppler, 15 of whom had formal examination the following day. Of these 15, the EPs had a sensitivity of 100%, a specicity of 75%, a positive predictive value of 78%, and an NPV of 100%. Of the 8 patients who did not undergo a follow-up study, 1 was lost to follow-up and another was later diagnosed with a pelvic DVT. Two others had positive ED examinations and were treated. The investigators indicated that these evaluations were performed by EPs who usually did not work clinically in the ED. Some evaluations took nearly 30 minutes, minimizing one advantage in a single-coverage, busy ED. Although this study was small, retrospective, and had obvious limitations, it provided initial data specic to US use by EPs. Frazee et al [67] prospectively compared EP-performed limited B-mode compression US with full duplex US performed in a vascular laboratory. The EPs (three attendings and three senior residents) had extensive emergency US experience for other types of US but underwent only a 2-hour practical course with a vascular technician. Compared with duplex US, the
784
limited US performed in a convenience sample of 76 patients had a sensitivity of 88.9%, a specicity of 75.9%, and an NPV of 95.7%. It should be noted, however, that this study had indeterminate results in 16% of patients (67% during the rst half of the study), and these evaluations were considered positive for statistical purposes, thereby signicantly reducing the specicity. The investigators suggested that the addition of color Doppler might have reduced the number of false-positives results. Further, the reduced sensitivity can be explained by the fact that the two patients who had false-negative limited US results had complete compression by duplex US but were believed to have other indirect evidence of DVT. Their primary care physicians treated them as having a cellulitis, and a 2year follow-up showed no evidence of DVT or PE. Similar limitations exist in this study with regard to design; however, the results suggest that when combining a low-to-moderate pretest probability with a limited compression US performed by an EP, DVT reliably could be excluded or at least treatment could be withheld until a formal evaluation is undertaken. Blaivas et al [80] performed a prospective observational study using limited US but incorporated duplex imaging for visualization of ow and augmentation in the femoral and popliteal veins. Five EPs, each with extensive US experience for other ED US applications, took an identical 2hour didactic and 3-hour practical training course. They then examined 112 patients at high risk for DVT and recorded the time to perform this examination. Of these 112 patients, 34 were found to have a DVT of both vessels (21), the popliteal vein (12), or the femoral vein (1). Evaluation by a vascular laboratory agreed with 110 of 112 evaluations, with 1 falsepositive evaluation in the ED and 1 false-negative evaluation performed by the vascular laboratory in a patient later found to have a conrmed popliteal DVT by venography. The initial EP US was correct and, therefore, only 1 of 112 patients were misdiagnosed. No proximal DVTs were missed by the EPs. The vascular laboratory found two isolated calf DVTs that were not imaged by the EPs. One patient received anticoagulation and the other had self-resolution of the DVT on repeat US at 7 days. The average time to perform this examination was only 3 minutes, 28 seconds. These investigators believed that because the vascular laboratory was not 100% sensitive and specic, reporting these values would be inappropriate, but they did note a 98% agreement between the two studies (kappa 0.9). Conrming previous studies, these investigators also found the presence or absence of augmentation and spontaneous blood ow to be inconsistent with respect to the presence or absence of DVT. They believed that color Doppler helped to visualize vessels that otherwise would have been dicult to locate and noted the correct EP diagnosis of three cases of saphenous vein thrombosis and six cases of Baker cysts [80]. In a recent abstract, a group of 20 EPs underwent 2 hours of training on how to perform a 2-point limited US and then performed 80 examinations on a convenience sample of patients in which DVT was suspected. This
785
group found EP US to be 100% sensitive and 98.2% specic in the diagnosis of DVT compared with traditional radiology US; however, 13.8% of the study group had equivocal studies, with 18% of these patients ultimately being diagnosed with DVTs. The investigators believed that an EPperformed US can exclude a DVT in the ED [81]. In addition to the limitation of this being a convenience sample of patients, this study did not address the incorporation of D-dimer or pretest probability into a clinical algorithm. Clearly, there is a paucity of data about EP-performed venous US for diagnosis of DVT. Nonetheless, the preliminary data suggest that it could save time and, when positive, could be enough to initiate therapy.
Technique and ndings Imaging vascular structures requires using a high-resolution linear array probe. These probes usually have frequencies ranging from 7 to 10 MHz and allow visualization of supercial and deep vessels. In obese patients, a lower frequency may be required to allow for deeper penetration at the expense of resolution. Newer probes incorporate this frequency range into a single probe. As discussed previously, color Doppler can aid in the visualization of vessels (Fig. 2), particularly in technically dicult patients and in those who have structures (cysts or lymph nodes) that might mimic vessels.
Fig. 2. US image with Doppler (color not shown) of the left supercial femoral vein (V) and artery (A) without (left) and with (right) compression. Smaller vessels located next to each other are identied more easily with color Doppler.
786
To increase venous distention, the patient should be placed supine with the head of the patients bed raised to 45(. Probe placement, however, might be obscured in the morbidly obese patient, making head elevation impossible. Initial probe placement is in the groin, with visualization of the femoral vessels at the inguinal line (Fig. 3). The femoral artery and vein are visualized rst, with the artery having a more pulsatile appearance compared with the vein. The normal vein should be easily collapsible, with enough pressure to indent the overlying skin. At this pressure, the arterial diameter should remain intact. Again, if there is diculty in dierentiating artery from vein, color Doppler can be used to determine ow, and the intensity of ow will be apparent as a graphic representation. For those familiar with the use of spectral Doppler, the arterial and venous waveforms can be assessed with venous ow appearing as a slower undulating pattern compared with the triphasic arterial ow traditionally heard during Doppler US of the extremities. When a DVT is present, there is no ow within the vein (Fig. 4). The probe is moved distally to the bifurcation of the femoral vein into the supercial femoral vein and deep femoral vein. Here, compression should be applied with complete collapse of the normal vein (Fig. 5). An area of approximately 5 cm should be compressed at the bifurcation. Any inability to collapse the vein indicates a DVT (Fig. 6). Echogenic thrombus may be visible but is not required for diagnosis. The vessels of interest also can be viewed in the sagittal plane, thereby assisting in vein orientation and direction. A suspected DVT seen in the transverse plane can be viewed with color Doppler in the sagittal plane; both orientations show no ow when color Doppler is used (Fig. 7). Imaging of the popliteal vein is performed next with the knee slightly exed and with the hip slightly externally rotated (Fig. 8). The same
Fig. 3. Patient positioning and probe location for imaging of the deep femoral veins of the groin.
787
Fig. 4. A sagittal view of the popliteal vein and artery with spectral Doppler indicating no ow.
technique is used to compress the popliteal vein, with complete collapse required for the exclusion of DVT (Fig. 9). Again, an area of approximately 5 cm is scanned in quick, successive movements around the trifurcation. Duplication of the popliteal vessels is common [82], and both vessels should be imaged in similar fashion. Because imaging of the popliteal vein can be more dicult than the femoral vein bifurcation, variations in technique can be attempted. Having
Fig. 5. US images of the left common femoral vein (V) and artery (A) without (left) and with (right) compression. When the CFV easily collapses, there is no evidence of DVT.
788
Fig. 6. US images of the left greater saphenous vein (GSV) and common femoral vein (CFV) without (left) and with (right) compression. The GSV and the CFV do not collapse with compression, which is indicative of DVT. The surrounding supercial femoral artery (A) and deep femoral artery (A) are visible and do not collapse.
Fig. 7. Doppler US image (color not visible) of the bifurcation of the supercial femoral vein (SFV) and deep femoral vein (DFV) in the sagittal plane. These veins did not collapse with compression, and Doppler conrms no ow in these veins. Flow is visible in the surrounding supercial femoral artery (SFA) and deep femoral artery (DFA).
789
Fig. 8. Probe location for imaging of the popliteal fossa.
the patient sit with his or her knee dangling over the edge of the bed can increase venous distention and ease the visualization of the popliteal vein. Further, with the patient in the prone position, the knee can be exed 45( to 90(, with the probe placed in the popliteal fossa. This position allows for easier probe placement because the probe and cord do not come in contact with the bed.
Fig. 9. US image of the left popliteal vein (V) and artery (A) without (left) and with (right) compression. A DVT is indicated because the vein is still visible with compression.
790
Fig. 10. US image of a supercial vein without (left) and with (right) compression. In this patient with a lower extremity cellulitis and no US of DVT, a diagnosis of supercial thrombophlebitis was made.
As discussed earlier, the full extent of the required components of limited US is debated. At its minimum, the femoral bifurcation is imaged rst and the popliteal vein is imaged second. Imaging of other locations in the thigh and knee is not required, but noting the extent of clot within the vessel can be useful when follow-up evaluations are performed to assess clot extension. Because many EPs likely are to request conrmatory formal vascular evaluations, however, this notation may not be necessary.
Fig. 11. US image of a 76-year-old man with right knee pain and swelling. (A) A collection of uid is visible within the popliteal fossa consistent with a Baker cyst. (B) When Doppler is applied to this area, there is no evidence of ow.
Fig. 12. Clinical algorithm for the diagnosis of suspected lower extremity DVT.
791
792
As the diagnosis of DVT is excluded, other conditions are detectable by ED US. In the patient with a lower extremity cellulitis, a thrombophlebitis easily can be diagnosed. A supercial thrombosis appears similar to incomplete vein collapse on compression of the supercial vessel (Fig. 10). A lymph node in the groin and a Baker cyst in the popliteal fossa can be discerned from a vessel when color Doppler is available (Fig. 11). These structures may appear as round or ovoid vessels in the transverse view; however, they lack ow and usually are discrete structures in the tissue, unlike a vein that continues in either direction. The occasional popliteal artery aneurysm also may be visible.
Clinical applications Convincing data supporting the use of US by EPs outside of an articial study setting currently are lacking. Although many EPs currently incorporate limited lower extremity US into their practice, this is not yet the standard of care. There is, however, a signicant body of evidence to support the combined use of pretest probability and either a D-dimer assay or a limited US in the diagnosis of DVT (Fig. 12). Larger studies are needed to document the use of this technique performed by EPs. If such data become available, the incorporation of a validated D-dimer measurement and real-time limited US by the EP would improve the care of patients with DVT signicantly.
Acknowledgment The authors would like to thank Wende R. Buras, MD, PhD, for her illustration assistance. References
[1] Birdwell BG, Raskob GE, Whitsett TL, et al. The clinical validity of normal compression ultrasonography in outpatients suspected of having deep venous thrombosis. Ann Intern Med 1998;128:17. [2] Huisman MV, Buller HR, ten Cate JW, et al. Unexpected high prevalence of silent pulmonary embolism in patients with deep venous thrombosis. Chest 1989;95:498502. [3] Monreal M, Ruiz J, Olazabal A, et al. Deep venous thrombosis and the risk of pulmonary embolism. A systematic study. Chest 1992;102:67781. [4] Moser KM, Fedullo PF, LitteJohn JK, et al. Frequent asymptomatic pulmonary embolism in patients with deep venous thrombosis. JAMA 1994;271:2235. [5] Cranley JJ, Canos AJ, Sull WJ. The diagnosis of deep venous thrombosis. Fallibility of clinical symptoms and signs. Arch Surg 1976;111:346. [6] Cogo A, Lensing AW, Prandoni P, et al. Distribution of thrombosis in patients with symptomatic deep vein thrombosis. Implications for simplifying the diagnostic process with compression ultrasound. Arch Intern Med 1993;153:277780.
793
[7] Frazee BW, Snoey ER. Diagnostic role of ED ultrasound in deep venous thrombosis and pulmonary embolism. Am J Emerg Med 1999;17:2718. [8] Hirsh J, Hull RD, Raskob GE. Clinical features and diagnosis of venous thrombosis. J Am Coll Cardiol 1986;8:114B27B. [9] Moser KM. Venous thromboembolism. Am Rev Respir Dis 1990;141:23549. [10] Wells PS, Anderson DR, Bormanis J, et al. Value of assessment of pretest probability of deep-vein thrombosis in clinical management. Lancet 1997;350:17958. [11] Anderson DR, Kovacs MJ, Kovacs G, et al. Combined use of clinical assessment and D-dimer to improve the management of patients presenting to the emergency department with suspected deep vein thrombosis (the EDITED Study). J Thromb Haemost 2003;1: 64551. [12] Bates SM, Kearon C, Crowther M, et al. A diagnostic strategy involving a quantitative latex D-dimer assay reliably excludes deep venous thrombosis. Ann Intern Med 2003;138:78794. [13] Constans J, Boutinet C, Salmi LR, et al. Comparison of four clinical prediction scores for the diagnosis of lower limb deep venous thrombosis in outpatients. Am J Med 2003;115:43640. [14] Miron MJ, Perrier A, Bounameaux H. Clinical assessment of suspected deep vein thrombosis: comparison between a score and empirical assessment. J Intern Med 2000;247: 24954. [15] Schutgens RE, Ackermark P, Haas FJ, et al. Combination of a normal D-dimer concentration and a non-high pretest clinical probability score is a safe strategy to exclude deep venous thrombosis. Circulation 2003;107:5937. [16] Tick LW, Ton E, van Voorthuizen T, et al. Practical diagnostic management of patients with clinically suspected deep vein thrombosis by clinical probability test, compression ultrasonography, and D-dimer test. Am J Med 2002;113:6305. [17] Smithline HA, Mader TJ, Ali FM, et al. Determining pretest probability of DVT: clinical intuition vs. validated scoring systems. Am J Emerg Med 2003;21:1612. [18] Baker WF Jr. Diagnosis of deep venous thrombosis and pulmonary embolism. Med Clin N Am 1998;82:45976. [19] Hull R, Hirsh J, Sackett DL, et al. Clinical validity of a negative venogram in patients with clinically suspected venous thrombosis. Circulation 1981;64:6225. [20] Fraser JD, Anderson DR. Deep venous thrombosis: recent advances and optimal investigation with US. Radiology 1999;211:924. [21] Redman HC. Deep venous thrombosis: is contrast venography still the diagnostic gold standard? [editorial] Radiology 1988;168:2778. [22] Erdman WA, Jayson HT, Redman HC, et al. Deep venous thrombosis of extremities: role of MR imaging in the diagnosis. Radiology 1990;174:42531. [23] Evans AJ, Sostman HD, Knelson MH, et al. 1992 ARRS Executive Council Award. Detection of deep venous thrombosis: prospective comparison of MR imaging with contrast venography. AJR Am J Roentgenol 1993;161:1319. [24] Fraser DG, Moody AR, Morgan PS, et al. Diagnosis of lower-limb deep venous thrombosis: a prospective blinded study of magnetic resonance direct thrombus imaging. Ann Intern Med 2002;136:8998. [25] Spritzer CE, Evans AC, Kay HH. Magnetic resonance imaging of deep venous thrombosis in pregnant women with lower extremity edema. Obstet Gynecol 1995;85:6037. [26] Spritzer CE, Norconk JJ Jr, Sostman HD, et al. Detection of deep venous thrombosis by magnetic resonance imaging. Chest 1993;104:5460. [27] Burke B, Sostman HD, Carroll BA, et al. The diagnostic approach to deep venous thrombosis. Which technique? Clin Chest Med 1995;16:25368. [28] Moody AR. Direct imaging of deep-vein thrombosis with magnetic resonance imaging [letter]. Lancet 1997;350:1073. [29] Laissy JP, Cinqualbre A, Loshkajian A, et al. Assessment of deep venous thrombosis in the lower limbs and pelvis: MR venography versus duplex Doppler sonography. AJR Am J Roentgenol 1996;167:9715.
794
[30] Cham MD, Yankelevitz DF, Shaham D, et al. Deep venous thrombosis: detection by using indirect CT venography. The Pulmonary Angiography-Indirect CT Venography Cooperative Group. Radiology 2000;216:74451. [31] Coche EE, Hamoir XL, Hammer FD, et al. Using dual-detector helical CT angiography to detect deep venous thrombosis in patients with suspicion of pulmonary embolism: diagnostic value and additional ndings. AJR Am J Roentgenol 2001;176:10359. [32] Duwe KM, Shiau M, Budorick NE, et al. Evaluation of the lower extremity veins in patients with suspected pulmonary embolism: a retrospective comparison of helical CT venography and sonography. 2000 ARRS Executive Council Award I. American Roentgen Ray Society. AJR Am J Roentgenol 2000;175:152531. [33] Garg K, Kemp JL, Wojcik D, et al. Thromboembolic disease: comparison of combined CT pulmonary angiography and venography with bilateral leg sonography in 70 patients. AJR Am J Roentgenol 2000;175:9971001. [34] Loud PA, Katz DS, Bruce DA, et al. Deep venous thrombosis with suspected pulmonary embolism: detection with combined CT venography and pulmonary angiography. Radiology 2001;219:498502. [35] Brill-Edwards P, Lee A. D-dimer testing in the diagnosis of acute venous thromboembolism. Thromb Haemost 1999;82:68894. [36] Colucciello S. Protocols for deep venous thrombosis: a state-of-the-art review (part II). Emerg Med Rep 2000;20:25. [37] Wells PS, Brill-Edwards P, Stevens P, et al. A novel and rapid whole-blood assay for Ddimer in patients with clinically suspected deep vein thrombosis. Circulation 1995;91: 21847. [38] Bernardi E, Prandoni P, Lensing AW, et al. D-dimer testing as an adjunct to ultrasonography in patients with clinically suspected deep vein thrombosis: prospective cohort study. The Multicentre Italian D-dimer Ultrasound Study Investigators Group. BMJ 1998;317:103740. [39] DAngelo A, DAlessandro G, Tomassini L, et al. Evaluation of a new rapid quantitative D-dimer assay in patients with clinically suspected deep vein thrombosis. Thromb Haemost 1996;75:4126. [40] Freyburger G, Trillaud H, Labrouche S, et al. D-dimer strategy in thrombosis exclusiona gold standard study in 100 patients suspected of deep venous thrombosis or pulmonary embolism: 8 DD methods compared. Thromb Haemost 1998;79:327. [41] Lee AY, Julian JA, Levine MN, et al. Clinical utility of a rapid whole-blood D-dimer assay in patients with cancer who present with suspected acute deep venous thrombosis. Ann Intern Med 1999;131:41723. [42] Leroyer C, Escore M, Le Moigne E, et al. Diagnostic value of a new sensitive membrane based technique for instantaneous D-dimer evaluation in patients with clinically suspected deep venous thrombosis. Thromb Haemost 1997;77:63740. [43] Perrier A, Desmarais S, Miron MJ, et al. Non-invasive diagnosis of venous thromboembolism in outpatients. Lancet 1999;353:1905. [44] Farrell S, Hayes T, Shaw M. A negative SimpliRED D-dimer assay result does not exclude the diagnosis of deep vein thrombosis or pulmonary embolus in emergency department patients. Ann Emerg Med 2000;35:1215. [45] Aschwanden M, Labs KH, Jeanneret C, et al. The value of rapid D-dimer testing combined with structured clinical evaluation for the diagnosis of deep vein thrombosis. J Vasc Surg 1999;30:92935. [46] Ginsberg JS, Kearon C, Douketis J, et al. The use of D-dimer testing and impedance plethysmographic examination in patients with clinical indications of deep vein thrombosis. Arch Intern Med 1997;157:107781. [47] Kearon C, Ginsberg JS, Douketis J, et al. Management of suspected deep venous thrombosis in outpatients by using clinical assessment and D-dimer testing. Ann Intern Med 2001;135:10811.
795
[48] Kraaijenhagen RA, Piovella F, Bernardi E, et al. Simplication of the diagnostic management of suspected deep vein thrombosis. Arch Intern Med 2002;162:90711. [49] Lennox AF, Delis KT, Serunkuma S, et al. Combination of a clinical risk assessment score and rapid whole blood D-dimer testing in the diagnosis of deep vein thrombosis in symptomatic patients. J Vasc Surg 1999;30:794803. [50] Turkstra F, van Beek EJ, ten Cate JW, et al. Reliable rapid blood test for the exclusion of venous thromboembolism in symptomatic outpatients. Thromb Haemost 1996;76:911. [51] Wells PS, Anderson DR, Bormanis J, et al. SimpliRED D-dimer can reduce the diagnostic tests in suspected deep vein thrombosis [letter]. Lancet 1998;351:14056. [52] Wells PS, Anderson DR, Rodger M, et al. Evaluation of D-dimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003;349:122735. [53] Richlie DG. Noninvasive imaging of the lower extremity for deep venous thrombosis. J Gen Intern Med 1993;8:2717. [54] Appelman PT, De Jong TE, Lampmann LE. Deep venous thrombosis of the leg: US ndings. Radiology 1987;163:7436. [55] Lensing AW, Prandoni P, Brandjes D, et al. Detection of deep-vein thrombosis by realtime B-mode ultrasonography. N Engl J Med 1989;320:3425. [56] Cronan JJ. Venous thromboembolic disease: the role of US. Radiology 1993;186:619. [57] Vogel P, Laing FC, Jerey RB Jr, et al. Deep venous thrombosis of the lower extremity: US evaluation. Radiology 1987;163:74751. [58] Kearon C, Julian JA, Newman TE, et al. Noninvasive diagnosis of deep venous thrombosis. McMaster Diagnostic Imaging Practice Guidelines Initiative. Ann Intern Med 1998;128:66377. [59] Cronan JJ, Dorfman GS, Scola FH, et al. Deep venous thrombosis: US assessment using vein compression. Radiology 1987;162:1914. [60] Kearon C, Ginsberg JS, Hirsh J. The role of venous ultrasonography in the diagnosis of suspected deep venous thrombosis and pulmonary embolism. Ann Intern Med 1998;129: 10449. [61] Philbrick JT, Becker DM. Calf deep venous thrombosis. A wolf in sheeps clothing? Arch Intern Med 1988;148:21318. [62] Vaccaro JP, Cronan JJ, Dorfman GS. Outcome analysis of patients with normal compression US examinations. Radiology 1990;175:6459. [63] Fesmire FM, Kline JA, Wolf SJ, et al. American College of Emergency Physicians Clinical policy: critical issues in the evaluation and management of adult patients presenting with suspected lower-extremity deep venous thrombosis. Ann Emerg Med 2003; 42:12435. [64] Birdwell BG, Raskob GE, Whitsett TL, et al. Predictive value of compression ultrasonography for deep vein thrombosis in symptomatic outpatients: clinical implications of the site of vein noncompressibility. Arch Intern Med 2000;160:30913. [65] Cogo A, Lensing AW, Koopman MM, et al. Compression ultrasonography for diagnostic management of patients with clinically suspected deep vein thrombosis: prospective cohort study. BMJ 1998;316:1720. [66] Heijboer H, Buller HR, Lensing AW, et al. A comparison of real-time compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic outpatients. N Engl J Med 1993;329:13659. [67] Frazee BW, Snoey ER, Levitt A. Emergency department compression ultrasound to diagnose proximal deep vein thrombosis. J Emerg Med 2001;20:10712. [68] Killewich LA, Bedford GR, Beach KW, et al. Diagnosis of deep venous thrombosis. A prospective study comparing duplex scanning to contrast venography. Circulation 1989; 79:8104. [69] Poppiti R, Papanicolaou G, Perese S, et al. Limited B-mode venous imaging versus complete color-ow duplex venous scanning for detection of proximal deep venous thrombosis. J Vasc Surg 1995;22:5537.
796
[70] Pezzullo JA, Perkins AB, Cronan JJ. Symptomatic deep vein thrombosis: diagnosis with limited compression US. Radiology 1996;198:6770. [71] Aronen HJ, Svedstrom E, Yrjana J, et al. Compression sonography in the diagnosis of deep venous thrombosis of the leg. Ann Med 1994;26:37780. [72] Quintavalla R, Larini P, Miselli A, et al. Duplex ultrasound diagnosis of symptomatic proximal deep vein thrombosis of lower limbs. Eur J Radiol 1992;15:326. [73] Schindler JM, Kaiser M, Gerber A, et al. Colour coded duplex sonography in suspected deep vein thrombosis of the leg. BMJ 1990;301:136970. [74] Rosen MP, McArdle C. Controversies in the use of lower extremity sonography in the diagnosis of acute deep vein thrombosis and a proposal for a unied approach. Semin Ultrasound CT MR 1997;18:3628. [75] Frederick MG, Hertzberg BS, Kliewer MA, et al. Can the US examination for lower extremity deep venous thrombosis be abbreviated? A prospective study of 755 examinations. Radiology 1996;199:457. [76] Naidich JB, Torre JR, Pellerito JS, et al. Suspected deep venous thrombosis: is US of both legs necessary? [comment] Radiology 1996;200:42931. [77] Sheiman RG, McArdle CR. Bilateral lower extremity US in the patient with unilateral symptoms of deep venous thrombosis: assessment of need. Radiology 1995;194:1713. [78] American College of Emergency Physicians. Use of ultrasound imaging by emergency physicians. Ann Emerg Med 2001;38:46970. [79] Jolly BT, Massarin E, Pigman EC. Color Doppler ultrasonography by emergency physicians for the diagnosis of acute deep venous thrombosis. Acad Emerg Med 1997;4: 12932. [80] Blaivas M, Lambert MJ, Harwood RA, et al. Lower-extremity Doppler for deep venous thrombosiscan emergency physicians be accurate and fast? Acad Emerg Med 2000;7: 1206. [81] Wong HE, Moore C, Skrupky R, et al. Accuracy of emergency physician compression ultrasonography and D-dimer in the bedside diagnosis of deep venous thrombosis [abstract]. Ann Emerg Med 2003;42:S89S90. [82] Quinlan DJ, Alikhan R, Gishen P, et al. Variations in lower limb venous anatomy: implications for US diagnosis of deep vein thrombosis. Radiology 2003;228:4438.
Procedural applications of ultrasound

Carrie D. Tibbles, MD*, William Porcaro, MD
Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Harvard Aliated Emergency Medicine Residency, Harvard Medical School, West Campus Clinical Center 2, One Deaconess Road, Boston, MA 02215, USA
Invasive procedures are essential to the daily practice of emergency medicine. Many of these procedures, ranging from a simple incision and drainage to an emergent pericardiocentesis, can be performed under ultrasound guidance. A growing body of evidence suggests that sonographically guided procedures are more ecient and safer than conventional techniques. This article reviews the use of ultrasound to facilitate common procedures in the emergency department.
Incision and drainage of supercial abscesses In the assessment of a soft tissue infection, it can be dicult to detect an underlying abscess using clinical criteria alone [1]. Ultrasound of the suspected area can be used to localize a uid collection accurately and guide aspiration or incision and drainage [2]. The initial reports of ultrasound-guided incision and drainage focused on intravenous drug users [3,4]. In these studies, ultrasound was able to differentiate between cellulitis and abscess and identify septic thrombophlebitis, adenitis, and pseudoaneurysm [4]. Blaivas [5] published a case report of a breast abscess successfully drained under ultrasound guidance in the emergency department after 15 unsuccessful blind attempts at needle aspiration. Ultrasound also is useful for the characterization of odontogenic facial infections. In a recent study published in the Journal of Oral and Maxillofacial Surgery, Peleg et al [6] performed ultrasound on 50 patients with acute odontogenic infections. Infected uid was aspirated in the 22 patients found to have uid
* Corresponding author. E-mail address: ctibbles@bidmc.harvard.edu (C.D. Tibbles). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.010
798
C.D. Tibbles, W. Porcaro / Emerg Med Clin N Am 22 (2004) 797815
on ultrasound. Of the 28 patients diagnosed with cellulitis, 6 had repeat scans for persistent symptoms and 4 had evidence of abscess formation. Page-Wills and colleagues [7] attempted to determine the impact of bedside ultrasound on the management of soft tissue infections in the emergency department. Fifty-six patients with soft tissue infections were classied as having a low (12), indeterminate (32), or high probability (12) of abscess, based on history and physical examination. As expected, ultrasound detected uid in all of the high-probability cases. More important, ultrasound found uid in 7 of the 12 low-probability patients, with pus drained in 6 patients (50%). Fluid also was identied in 25 of the 32 indeterminate patients, with pus subsequently drained in 23 patients. The physicians in the study believed that ultrasound inuenced the management of the case more than half the time. The investigators concluded that an abscess often is not reliably identied by physical examination alone and recommended the routine use of ultrasound to assess soft tissue infections. The appearance of a subcutaneous abscess on ultrasound can be varied. Typically, an abscess appears as a hypoechoic mass. The margin may be well dened or blend into the surrounding cellulitis (Fig. 1). There may be a hyperechoic rim or debris within the abscess cavity [2]. Posterior acoustic enhancement is common [1]. An abscess also can be isoechoic, particularly in the early stages, and this enhancement may be the only reliable sign [2]. In more occult cases, Loyer et al [8] suggested gentle repetitive pressure with the transducer to induce motion of any purulent uid to improve sensitivity. Occasionally, infected uid appears as a hyperechoic mass. Ultrasound also identies any adjacent structures such as vessels, nerves, or tendons that
Fig. 1. Ultrasound image of a subcutaneous abscess. (Courtesy of R. Reardon, MD, Hennepin County Medical Center, Minneapolis, MN.)
799
should be avoided during the incision [9]. The region of interest is scanned using a high-frequency probe (eg, 5.0 or 7.5 MHz). For a very supercial abscess, a 10-MHz transducer or gel-lled standoff pad can be used to optimize resolution [9]. The abscess is scanned in two perpendicular planes, and the depth of the abscess should be noted. In certain clinical scenarios such as an infected bursa or breast abscess, needle aspiration may be preferred over a surgical incision. In these cases, sonography can be used to visualize the needle entering the abscess cavity, although this nal step is not always necessary [1].
Management of peritonsillar abscesses Peritonsillar abscesses are the most common deep-space infections of the head and neck [10]. Physical examination alone cannot always reliably differentiate between peritonsillar cellulitis and an abscess [11]. Traditionally, a diagnostic needle aspiration is performed to assess for a uid collection; however, this procedure is not without complications, and a falsenegative rate of up to 12% has been reported [12]. In 1993, Haeggstrom and colleagues [13] published one of the early reports on the use of intraoral ultrasound in the management of peritonsillar abscesses in 12 patients. The investigators concluded that ultrasound is advantageous for the management of peritonsillar abscess because the precise location and extent of the abscess can be dened. If a uid collection is found, then the ultrasound is used to guide the aspiration or the incision and drainage. In addition, the relationship of the abscess to the carotid artery can be determined. Using a CT scan as the gold standard, Scott and coinvestigators [12] found intraoral ultrasound to be 89% sensitive and 100% specic for the diagnosis of peritonsillar abscess. Similar results have been found by other investigators, and intraoral ultrasonography now is routinely used by otolaryngologists for the management of a peritonsillar abscess [11,14]. OBrien [15] found that intraoral sonography was well tolerated by emergency department patients, even those with trismus. The anatomy of this region is complex [16]. A high-frequency curvilinear intracavity probe covered with a sterile sheath is used for the examination. Application of a topical anesthestic is recommended [9]. When scanning, the most important structure to identify is the carotid artery. It courses posterolateral to the tonsil and is typically located within 5 to 25 mm of the abscess cavity. The carotid artery is pulsatile and noncompressible compared with the internal jugular vein located posterior to the artery in the carotid sheath [16].
Detection and removal of foreign bodies Many extremity wounds harbor an occult foreign body. A retained foreign body may lead to numerous complications including infection,
800
inammation, and loss of function [17]. A retrospective review found that 38% of foreign bodies in the hand were missed on initial physical examination [18]. In general, missed foreign bodies are the second leading cause of lawsuits in emergency medicine (14%) and account for 5% of all settlements [17]. The standard screening modality for suspected foreign bodies in extremity wounds is plain radiography; however, only radiopaque objects such as metal, glass, and stone can be detected. Radiolucent objects such as thorns, wood, and plastic are missed [19]. Numerous strategies and imaging modalities have been employed in the attempt to dene a method for accurate identication of foreign bodies in soft tissue. Ginsberg and colleagues [20] performed an in vitro study comparing detection of wood, glass, and plastic by various imaging modalities. They found that ultrasound was capable of detecting wood and plastic fragments placed between strips of steak submerged in water, but plain radiography, CT, and xerography (radiographs on charged selenium plates) were not. MRI, although very accurate for foreign body detection, is not practical for routine use in the emergency department [17]. Various tissue models have been used to examine the sensitivity and specicity of ultrasound for the detection of foreign bodies. The accuracy of ultrasound varies, depending on the physical model used, the experience of the sonographers, and the size of the foreign body [9]. Schlager and colleagues [21] examined the performance of emergency physicians using a 7.5-MHz ultrasound transducer to locate gravel, cactus, glass, metal, wood, and plastic fragments inserted in beef cubes. Fifty-nine of 60 foreign bodies were detected, with one false-positive identication (98% sensitivity, 98% specicity). The investigators noted that their model did not account for bone artifact, air in the wound, and other factors that complicate examination in an actual patient. In a similar study, credentialed sonographers examined chicken breasts impregnated with various foreign bodies. Wood was found to exhibit the strongest acoustic shadow, followed by plastic. The investigators concluded that ultrasound is an excellent technique for the localization and removal of nonradiopaque foreign bodies [22]. Bray and colleagues [23] used a fresh frozen cadaver model to characterize the ability of ultrasound to detect wood, glass, and metal. Scans performed by a radiologist using a 10-MHz probe resulted in a sensitivity of 94% and a specicity of 99%. Jacobson et al [24] also used a cadaveric model to detect implanted wooden foreign bodies. They reported a sensitivity of 86.7% for 2.5-mm fragments and a sensitivity of 93.3% for 5.0-mm fragments, with an overall specicity of 96.7%. In contrast, Manthey et al [25] performed a descriptive study in which staff radiologists used a 7.5-MHz transducer to examine chicken thighs with implanted gravel, metal, glass, cactus, wood, and plastic fragments. Results were compared with plain radiographs. In this study, ultrasound performed poorly, with an overall sensitivity of 43%, specicity of 70%, false negative rate of 50%, and false positive rate of 30%.
801
Ultrasound has been found to be reliable for the detection of foreign bodies in several clinical studies. Fornage and Schernberg [26] compared ultrasound and radiography in several case reports of patients with suspected foreign bodies in the hand or foot. Ultrasound identied the foreign bodies as hyperechogenic foci in all of the patients who had foreign bodies found at surgery. In 7 of these patients, a surrounding area of hypoechogenicity that corresponded to inammatory changes also was identied. Banerjee and Das [27] prospectively studied 45 patients with suspected foreign body inclusions. Each patient had plain lms and an ultrasound examination. Radiologists detected 19 of 20 foreign bodies that were radiopaque. Seven foreign bodies were identied in 25 patients with negative radiography. Crawford and Matheson [28] studied 39 patients with persistent pain or infection after negative plain radiographs. Ultrasound found foreign bodies in 19 of 20 patients who subsequently had wood or thorns removed in the operating room. There was one false-positive identication in the 19 patients who had no material discovered after open exploration. Shiels and colleagues [29] studied 26 patients with suspected retained foreign bodies. Wood, stone, glass, metal, or pencil lead was detected by ultrasound in 19 of the 21 patients with a conrmed foreign body at surgery. The investigators also reported that ultrasound-guided removal worked best when a single physician performed the scanning and the instrument control. A linear probe with a frequency between 7.5 and 10 MHz is appropriate when searching for a foreign body [17]. An acoustic standoff pad can be used to improve visualization of the supercial soft tissue structures that otherwise might be lost in the transducers dead zone [9]. In general, foreign bodies are hyperechogenic. Some materials have unique sonographic patterns such as a comet-tail effect [17]. Metals produce a comet-tail shadow due to reverberation, whereas glass causes a more diffuse, scattered image. Sand or gravel fragments create an acoustic shadow reminiscent of gallstones. Ultrasound images of wood and plastic produce more distant, fainter shadows [17]. Many foreign bodies are surrounded by areas of hypogenicity, consistent with inammatory changes in the surrounding soft tissue (Fig. 2). Air introduced into the wound can complicate examinations. Slight compression of the tissue with the transducer may help to eliminate this artifact [9].
Arthrocentesis Joint pain and swelling are common complaints in the emergency department. Joint eusions are the result of multiple processes, including infection, inammation, crystal arthopathies, and connective tissue diseases [30]. Synovial uid analysis often is necessary to determine a denitive diagnosis [31]. Most joints can be tapped using standard anatomic landmarks;
802
Fig. 2. Ultrasound of the hand with a wooden toothpick in the supercial soft tissue. Note the surrounding hypoechogenicity secondary to tissue inammation and mild shadowing. (Courtesy of A. Dewitz, MD, Boston Medical Center, Boston, MA.)
however, arthrocentesis of certain joints such as the hip and ankle technically can be difcult. Ultrasound localizes the uid collection and can be used to guide the needle during the procedure [9]. Using a high-frequency probe, as little as 1 mL of uid can be visualized [30]. Ultrasound also is helpful to conrm the absence of a joint effusion after a dry tap. Bedside ultrasonography is faster and less expensive than CT and MRI for the emergency department evaluation of joint effusions. Ultrasound also can be used efciently to compare the symptomatic joint with the opposite side. In addition, no sedation is required for the examination of young children [30]. In 2001, Valley and Stahmer [30] published guidelines for the use of emergency bedside ultrasonography to detect joint effusions. This comprehensive review discusses the specic sonographic windows used for imaging each joint and the sonographic characteristics of joint effusions. In general, hip and ankle arthrocenteses are considered more technically difcult to perform, and the specic technique for these joints is reviewed in the following paragraphs. In 1999, Smith [32] published a case report of an emergency department patient with hip pain and fever. After several unsuccessful blind attempts by the consulting orthopedist to obtain joint uid, bedside ultrasound was used to localize the effusion and synovial uid was easily obtained. The patient ultimately was diagnosed with pseudogout and avoided an operation. Several dierent ultrasound transducers can be used to image the hip joint, with the optimum frequency depending on the body habitus of the patient. Valley and Stahmer [30] suggested using a 5.0- to 7.5-MHz probe in the average-sized adult. The hip should be slightly exed and internally rotated. The transducer is oriented along the long axis of the femoral neck,
803
and the transducer marker is positioned medially and superiorly [9]. The cortex of femoral neck and head appears as bright echoes a few centimeters below the skin (Fig. 3). The symptomatic hip should be compared with the contralateral side [31]. Effusions typically are visualized as an anechoic convex stripe in the anterior synovial recess, an area adjacent to the femoral neck [32]. A joint effusion is diagnosed when uid extending along the joint capsule measures greater than 5 mm or when a difference of 2 mm is appreciated between the two sides [30]. In a review of 19 ultrasound-guided hip joint aspirations, joint uid was aspirated more easily when the ultrasound demonstrated capsular distention [31]. Roy et al [33] published a recommended technique for using ultrasonography to assist with ankle arthrocentesis. The tibiotalar joint is imaged anteriorly with a 5.0- to 7.5-MHz transducer placed laterally to the extensor hallicus longus tendon. The transducer is oriented longitudinally, with the orientation marker pointing cephalad. An effusion is seen as an anechoic collection in the V-shaped anterior recess of the tibiotalar joint (Fig. 4). Of note, it is common to have uid in the tibiotalar joint, and the clinical circumstances will determine the need for arthrocentesis. Ultrasound also can be used to guide joint aspirations of the elbow, shoulder, and knee. Ultrasound particularly is helpful after a failed blind approach, when the eusion is small, or when patient characteristics such as obesity make the procedure more technically dicult.
Suprapubic bladder aspiration Suprapubic aspiration of urine is used to obtain a sterile urine specimen in the evaluation of a potentially septic infant. Although this procedure generally is considered safe, potential complications range from transient hematuria to bowel injury [34]. Success rates of suprapubic bladder
Fig. 3. Sonographic appearance of the hip joint demonstrating the anterior synovial recess. (Courtesy of R. Reardon, MD, Hennepin County Medical Center.)
804
Fig. 4. Normal sonographic appearance of the tibiotalar joint with physiologic joint uid. (Courtesy of A. Dewitz, MD, Boston Medical Center.)
aspiration vary among studies and range from 25% to 100% [35,36]. Pollack et al [35] compared suprapubic aspiration with catheterization in 100 emergency department patients and found a success rate of only 46% in the suprapubic aspiration group using the standard blind technique. Gochman and coinvestigators [36] conducted a prospective study to determine whether portable ultrasound improved the success rate of suprapubic aspiration in febrile infants. In this study of 66 patients, suprapubic aspiration was performed in the ultrasound group when the ultrasound revealed urine in the bladder and performed in the noultrasound group using the standard blind technique. The patients underwent urethral catheterization when the bladder was empty on ultrasound and the volume was recorded. The success rate was 79% in the ultrasound group and 52% in the no-ultrasound group. It is logical that one of the main determinants of success with suprapubic aspiration is the presence of an adequate volume of urine in the bladder. The investigators concluded that using ultrasound to visualize urine in the bladder reduced the number of nonproductive attempts. They also found that ultrasound had a sensitivity of 90% for the detection of urine compared with urethral catheterization. Similarly, OCallaghan and McDougall [37] were able to obtain urine 100% of the time using suprapubic aspiration when the ultrasound demonstrated a full bladder and only 36% of the time when ultrasound was not used. Not only is ultrasound useful to identify urine in the bladder but Kiernan and colleagues [38] proposed that using real-time scanning to visualize the needle throughout the procedure also would further improve the success
805
rate. Ultrasound guidance reduced the number of attempts, with an average of 1.7 needle passes in the ultrasound-guidance group compared with 4.4 needle passes in the no-ultrasound group. The overall success rate in the ultrasound group was 96%, a signicant improvement over 60% in the noultrasound group. Ultrasonography did not increase the time needed to perform the procedure. The bladder is visualized easily, even for a novice sonographer (Fig. 5). In neonates and small infants, a high-frequency transducer such as 5- or 7.5MHz probe commonly is used [37]. The bladder is imaged in the transverse and longitudinal planes. The area is prepped aseptically and a 22-gauge spinal needle is advanced into the bladder (Fig. 6). The needle rst is seen indenting the top of the bladder. The trajectory of the needle can be visualized with continuous scanning throughout the procedure [38]. A similar scanning technique also can be used to facilitate the placement of a suprapubic catheter in a trauma patient after a urethral or major pelvic injury. A 3.5-MHz transducer is typically used for scanning in this circumstance. After the introduction of the needle, a guide wire is inserted and the catheter is placed using the standard Seldinger technique.
Paracentesis In the United States, nontraumatic ascites most commonly is caused by cirrhosis and alcoholic liver disease. Other less common precipitants include congestive heart failure, myxedema, bowel obstruction or infraction,
Fig. 5. Normal sonographic appearance of the bladder. (Courtesy of R. Reardon, MD, Hennepin County Medical Center.)
806
Fig. 6. Technique for ultrasound-assisted suprapubic aspiration.
nephrotic syndrome, and infection [39]. The main indications for paracentesis in the emergency department are to rule-out spontaneous bacterial peritonitis, relieve symptoms associated with a tense ascites, and begin to formulate a diagnosis in new-onset ascites [40]. Large-volume paracentesis may be performed in the outpatient or emergency department setting for patients with chronic ascites who require frequent drainage. Albumin often is used concomitantly to avoid hyponatremia and renal insufciency [39]. Williams and Simel [41] studied the sensitivity and specicity of physician history and physical examination in determining the diagnosis of ascites. They found that ascites is unlikely when patients report no increase in abdominal girth and very unlikely in male patients with no complaint of leg swelling. There was no single sign that was specic and sensitive for ascites. Flank dullness (>80%), bulging anks (>72%), and shifting dullness to percussion had a high sensitivity (>83%). When performed in a conventional blind manner, paracentesis is safe. In a prospective study of 229 paracenteses performed on 125 patients, only two major complications (transfusion-requiring abdominal wall hematomas) and two minor complications (nontransfusion-requiring hematomas) occurred, yielding complication rates of 0.9% each. None of the procedures resulted in spontaneous bacterial peritonitis or death [40]. Some studies, however, suggest that the safety and efcacy of paracentesis may be improved through the use of ultrasound guidance. Bard and colleagues [42] studied 27 consecutive patients with ascites detected by ultrasound. It was determined that most free abdominal uid accumulated rst around the bladder, followed by the right paracolic gutter, and nally, the left ank region. In 6 of the 8 patients with uid in the ank region, air-lled loops of bowel were interposed between the abdominal wall and the uid collection,
807
suggesting that a blind approach may have injured the bowel. The investigators concluded that ultrasound is helpful for selecting the puncture site so as to avoid intraperitoneal structures during the paracentesis [42]. The use of ultrasound may help to avoid complications such as abdominal wall hematoma, intraperitoneal hemorrhage, and bowel perforation [9]. Abdominal ultrasound for ascites should be performed using a 3.5-MHz transducer. In noncomplicated cases, the patients left lower quadrant should be examined rst. On ultrasound, ascitic uid appears anechoic with occasional echogenic strands (Fig. 7) [42]. After a puncture site is identied and marked, the patient should remain in the same position for the paracentesis [9]. Using real-time scanning, the trajectory of the needle can be visualized throughout the procedure. Ultrasound also can be used after a large-volume paracentesis to assess the adequacy of drainage.
Thoracentesis Bedside ultrasound allows for more rapid detection and safer percutaneous drainage of pleural eusions than conventional methods. Focused thoracic ultrasound reliably can detect pleural eusions. Rozycki et al [43] studied the use of bedside thoracic ultrasound as an extension of the physical examination in the surgical critical care unit. A surgeon and a medical student performed examinations on 70 patients. The investigators reported a sensitivity of 83.6%, a specicity of 100%, and an accuracy of
Fig. 7. Ultrasound of the peritoneal cavity in a patient with ascites. (Courtesy of R. Reardon, MD, Hennepin County Medical Center.)
808
94% for the detection of pleural effusions with focused thoracic ultrasound (Fig. 8). Several emergency department studies conrm these results [4446]. Blind thoracentesis has been associated with complication rates as high as 20% to 50% [9]. Weingardt et al [47] systematically examined the puncture sites after 26 failed blind thoracentesis attempts in patients who then were referred for ultrasound-guided drainage. Thoracentesis was successfully performed using ultrasound guidance in 88% of the cases. Examination of the puncture sites revealed many errors in needle placement. Errors included directing the needle below the hemidiaphragm, above the uid collection, into consolidated lung, or over solid organs including the spleen, liver, and left kidney. Jones and colleagues [48] performed a prospective study of ultrasound-guided thoracentesis performed by interventional radiologists on patients in a seated position; 941 thoracenteses were performed on 605 patients. Complications that were recorded included pain in 25 patients, shortness of breath in 9 patients, cough in 8 patients, a vasovagal reaction in 6 patients, bleeding in 2 patients, hematoma in 2 patients and re-expansion pulmonary edema in 2 patients. A pneumothorax occurred in 24 patients (2.5%), with 8 requiring thoracostomies (0.8%). In contrast, the reported rates of pneumothorax in blind thoracenteses range from 4% to 30.3%. The investigators concluded that the complication rate for thoracentesis performed under ultrasound guidance is signicantly lower than that reported for blind thoracentesis. Diacon et al [49] had physicians of varying levels of experience examine patients with known pleural effusions and propose potential puncture sites. Ultrasonography then was used to evaluate these sites. If ultrasound showed !10 mm of uid perpendicular to the skin, then the site was determined to be accurate. Ultrasound decreased the possibility of
Fig. 8. Thoracic ultrasound in a patient with pleural eusion. Hyperechogenic diaphragm bound by liver on the right and hypoechogenic pleural uid on the left. (Courtesy of R. Reardon, MD, Hennepin County Medical Center.)
809
accidental organ puncture by 10% compared with clinically guided site selection. Kohan et al [50] compared ultrasonography to decubitus radiography as a guide for thoracentesis in a prospective study of 205 patients with pleural effusions. Decubitus radiographs were performed on all of the subjects, whereas ultrasonography was performed on a randomized subset. In cases of loculated (P\0.02) and free-owing (P\0.05) small effusions, ultrasound-guided thoracentesis was superior for obtaining adequate uid removal. There was no statistical difference between the two imaging techniques for large effusion. OMoore and colleagues [51] used ultrasonography to perform 187 diagnostic and therapeutic interventional procedures in the pleural space. In their subset of diagnostic aspirations, uid was obtained in 97% of cases, which included patients with very small effusions in whom unguided thoracentesis was believed to be impossible or unsafe. Only 3% of patients developed pneumothoraces when the procedure was performed under ultrasound guidance. Grogan et al [52] retrospectively studied clinical guidance used by medical residents versus ultrasound guidance used by radiology residents in 52 patients. There were statistically fewer serious complications in the ultrasound-guided group (0 of 19) compared with blind methods. Ultrasound of the pleural space may be performed with a 3.5-MHz or a 5.0MHz transducer. The patient is positioned in a seated-upright position when possible. Scanning is performed from the anterior axillary line to the paravertebral space horizontally and from the superior and inferior aspects of the uid collection longitudinally. The location of the hemidiaphragm and liver or spleen should be noted [9]. The width of the uid collection and its variance during a respiratory cycle also is important [53]. The depth of the uid collection is determined to gage optimal depth for needle insertion. Depth may be determined by referencing the scale markers on the ultrasound screen or by using M-mode [9]. Care should be taken to angle the transducer in all directions during the entire respiratory cycle to be assured that the lung, liver, spleen, or heart will not be in the trajectory of the needle. Mechanically ventilated patients must be approached in the supine position or oblique to the lateral decubitus position. Lichtenstein et al [53] developed a protocol for ultrasound-guided thoracentesis in ventilated patients. They recommended that an interpleural distance of !15 mm and visibility of the effusion over three intercostal spaces should be present before attempting the thoracentesis. Gervais and colleagues [54] performed a retrospective review to determine whether postprocedure chest radiography was necessary after ultrasoundguided thoracentesis. They reviewed the outcome after thoracentesis in 434 patients, including 92 who were intubated. Ten pneumothoraces occurred, 6 in intubated patients and 4 in spontaneous breathers (P\0.01). Two of the patients in the intubated group and none of the nonintubated patients required tube thoracostomy. The investigators suggested that a postprocedure chest radiograph may not be necessary after ultrasound-guided thoracentesis in spontaneously breathing patients.
810
Pericardiocentesis Echocardiography is very sensitive for the diagnosis of pericardial eusion and tamponade. On ultrasound, a pericardial eusion is seen as a dark or anechoic space between the heart and the pericardium (Fig. 9) [55]. Pericardial effusions generally are classied as acute or chronic. Tamponade occurs when the pericardial effusion raises the interpericardial pressure to a level that impairs ventricular lling, resulting in hemodynamic instability and rapid clinical deterioration of the patient [56]. Timely detection of a pericardial effusion or cardiac tamponade in victims of chest trauma is one of the most important applications of emergency department ultrasound [57]. In acute traumatic hemopericardium, the relatively inelastic pericardium cannot accommodate the accumulation of uid, and small pericardial effusions can result in cardiac tamponade. Hemopericardium from ventricular rupture or aortic dissection presents in a similar fashion [55]. Pericardiocentesis is not denitive therapy for acute hemopericaridium but can be an important temporizing measure. Conversely, nonhemorrhagic pericardial effusions usually accumulate slowly. Common etiologies of chronic effusions include malignancy, infections, and connective tissue disorders [56]. The pericardium becomes more compliant over time and can accommodate large effusions before tamponade develops. In each of these scenarios, the emergency physician rst must recognize the clinical signs of pericardial tamponade. The classic physical examination ndings of cardiac tamponadespecically, mufed heart sounds, hypotension, and distended neck veinsare present in less than 40% of conrmed cases of
Fig. 9. Cardiac ultrasound demonstrating a large pericardial eusion. (Courtesy of R. Reardon, MD, Hennepin County Medical Center.)
811
cardiac tamponade and may be difcult to assess in a noisy emergency department [58]. Emergency echocardiography is a rapid, noninvasive, and accurate method to conrm the presence or absence of a pericardial effusion in these patients. Several studies have demonstrated that emergency physicianperformed ultrasound is very accurate for the detection of pericardial effusions, with sensitivities approaching 100% [5759]. Ultrasound also is used to estimate the volume of the effusion and assess the hemodynamic effects by looking for right ventricular collapse, abnormal septal movement, and decreased respiratory variation of the inferior vena cava [60,61]. After overt signs of cardiac tamponade have developed, emergent pericardiocentesis can be a life-saving intervention. ECG-guided or blind pericardiocentesis using the subxiphoid approach generally is taught as the standard technique for emergent pericardiocentesis. This technique, however, has a complication rate as high as 50%, and deaths have been reported in many series [6264]. Complications include ventricular or atrial penetration, coronary artery lacerations, pneumothorax, dysrhythmias, or inadvertent puncture of the diaphragm and intrabdominal organs [65]. Studies from the cardiology literature strongly support the use of ultrasound guidance during pericardiocentesis [60,61,6668]. Although many of these studies describe elective procedures in the interventional cardiology laboratory, many of the conclusions are relevant to pericardiocentesis performed during emergency department resuscitations. Tsang et al [66] recently summarized the experience of echocardiographically guided pericardiocentesis performed at the Mayo Clinic over the last 21 years, whereby 1127 therapeutic pericardiocenteses were performed in 997 patients. The overall success rate was 97%, with an overall complication rate of 4.7%signicantly lower than that reported for blind pericardiocentesis. Importantly, the complication rate remained constant throughout the duration of the study despite the increased number of procedures performed for urgent indications in the later years of the study. Several other investigators have published similar results [60,65,69,70]. Echocardiographically guided pericardiocentesis also has been found to be safe in pediatric patients [71]. Pericardiocentesis is best performed with the patient in the semierect or left lateral position, but this not always feasible in the emergency setting. The heart is scanned quickly using a 3.5- or 5.0-MHz probe using the standard cardiac views to determine the extent of the eusion [56]. The relationship of the uid collection to the chest wall also can be claried. Tsang and colleagues [66,68] recommended that the ideal entry site is the point where the effusion is the closest to the transducer and the uid collection is maximal. In the review by Tsang et al [68], the chest wall was selected 79% of the time and was preferred over the subxiphoid approach. Taavitsainen et al [70] and Callahan et al [60] also found that the subxiphoid approach did not offer the most direct access to the effusion in most cases. Traditionally, pericardiocentesis has been performed in the emergency
812
department using a subxiphoid approach. Further study is needed to clarify the optimal approach for emergent pericardiocentesis under ultrasound guidance. Some studies recommend using a two-operator technique, with one person performing the scan and the other performing the pericardiocentesis. A sterile cover can be placed over the transducer to facilitate scanning throughout the procedure [65]. The single-operator technique using a probe-mounted needle also has been demonstrated to be safe and effective [69,72]. These devices are commercially available and can be used for any ultrasound-guided drainage procedure. The angle of the transducer is used to determine the trajectory of the needle, thus avoiding injury to vital structures such as the lung or intercostal arteries. The transducer is held 1 to 2 cm from the puncture site. Using real-time scanning, the needle can be visualized entering the pericardium, further reducing the risk of inadvertent puncture of adjacent structures [68,73]. Chiang and Lin [74] described injecting a small volume of agitated saline through the needle and using the resultant echoes to conrm needle placement. Injuries to the heart and adjacent structures were decreased further in the Mayo series by the use of a polytef-sheathed needle. After the needle enters the pericardium, the needle is withdrawn, leaving a sheath or cannula in place. A syringe attached to the cannula is used to aspirate the effusion [6668]. In conclusion, echcardiographically guided pericardiocentesis has become standard in cardiology, and many of the advantages found in the interventional laboratory may be applicable to practice in the emergency department.
Summary Many invasive procedures are now safer and more ecient with the use of ultrasound guidance. As emergency physicians continue to develop skills in sonography, new applications of this technology will continue to impact the practice of emergency medicine.
Acknowledgment The authors would like to acknowledge Dr. Robert Reardon and Dr. Andreas Dewitz for providing the ultrasound images accompanying this article.
References
[1] Craig JG. Infection: ultrasound-guided procedures. Radiol Clin N Am 1999;37(4):66978. [2] Loyer EM, DuBrow RA, David CL, et al. Imaging of supercial soft-tissue infections: sonographic ndings in cases of cellulitis and abscess. AJR Am J Roentgenol 1996;166(1): 14952.
813
[3] Yiengpruksawan A, Ganepola GA, Freeman HP. Acute soft tissue infection in intravenous drug abusers: its differential diagnosis by ultrasonography. J Natl Med Assoc 1986;78(12): 11936. [4] Sandler MA, Alpern MB, Madrazo BL, et al. Inammatory lesions of the groin: ultrasonic evaluation. Radiology 1984;151(3):74750. [5] Blaivas M. Ultrasound-guided breast abscess aspiration in a difcult case. Acad Emerg Med 2001;8(4):398401. [6] Peleg M, Heyman Z, Ardekian L, et al. The use of ultrasonography as a diagnostic tool for supercial fascial space infections. J Oral Maxillofac Surg 1998;56(10):112931 [discussion: 1132]. [7] Page-Wills C, Simon B, Christy D, et al. Utility of ultrasound on emergency department management of suspected cutaneous abscess. Acad Emerg Med 2000;7:493. [8] Loyer EM, Kaur H, David CL, et al. Importance of dynamic assessment of the soft tissues in the sonographic diagnosis of echogenic supercial abscesses. J Ultrasound Med 1995; 14(9):66971. [9] Dewitz A, Frazee B. Soft tissue applications. In: Ma O, Mateer J, editors. Emergency ultrasound. New York: Mcgraw-Hill; 2003;36191. [10] Wolf M, Even-Chen I, Kronenberg J. Peritonsillar abscess: repeated needle aspiration versus incision and drainage. Ann Otol Rhinol Laryngol 1994;103(7):5547. [11] Buckley AR, Moss EH, Blokmanis A. Diagnosis of peritonsillar abscess: value of intraoral sonography. AJR Am J Roentgenol 1994;162(4):9614. [12] Scott PM, Loftus WK, Kew J, et al. Diagnosis of peritonsillar infections: a prospective study of ultrasound, computerized tomography and clinical diagnosis. J Laryngol Otol 1999;113(3):22932. [13] Haeggstrom A, Gustafsson O, Engquist S, et al. Intraoral ultrasonography in the diagnosis of peritonsillar abscess. Otolaryngol Head Neck Surg 1993;108(3):2437. [14] Strong EB, Woodward PJ, Johnson LP. Intraoral ultrasound evaluation of peritonsillar abscess. Laryngoscope 1995;105(8 Pt 1):77982. [15] OBrien E. Intraoral sonography of peritonsillar abscess: feasibility and sonnographic appearance. Ann Emerg Med 1999;34:26. [16] Kew J, Ahuja A, Loftus WK, et al. Peritonsillar abscess appearance on intra-oral ultrasonography. Clin Radiol 1998;53(2):1436. [17] Schlager D. Ultrasound detection of foreign bodies and procedure guidance. Emerg Med Clin N Am 1997;15(4):895912. [18] Anderson MA, Newmeyer WL III, Kilgore ES Jr. Diagnosis and treatment of retained foreign bodies in the hand. Am J Surg 1982;144(1):637. [19] Flom LL, Ellis GL. Radiologic evaluation of foreign bodies. Emerg Med Clin N Am 1992; 10(1):16377. [20] Ginsburg MJ, Ellis GL, Flom LL. Detection of soft-tissue foreign bodies by plain radiography, xerography, computed tomography, and ultrasonography. Ann Emerg Med 1990;19(6):7013. [21] Schlager D, Sanders AB, Wiggins D, et al. Ultrasound for the detection of foreign bodies. Ann Emerg Med 1991;20(2):18991. [22] Turner J, Wilde CH, Hughes KC, et al. Ultrasound-guided retrieval of small foreign objects in subcutaneous tissue. Ann Emerg Med 1997;29(6):7314. [23] Bray PW, Mahoney JL, Campbell JP. Sensitivity and specicity of ultrasound in the diagnosis of foreign bodies in the hand. J Hand Surg (Am) 1995;20(4):6616. [24] Jacobson JA, Powell A, Craig JG, et al. Wooden foreign bodies in soft tissue: detection at US. Radiology 1998;206(1):458. [25] Manthey DE, Storrow AB, Milbourn JM, et al. Ultrasound versus radiography in the detection of soft-tissue foreign bodies. Ann Emerg Med 1996;28(1):79. [26] Fornage BD, Schernberg FL. Sonographic diagnosis of foreign bodies of the distal extremities. AJR Am J Roentgenol 1986;147(3):5679.
814
[27] Banerjee B, Das RK. Sonographic detection of foreign bodies of the extremities. Br J Radiol 1991;64(758):10712. [28] Crawford R, Matheson AB. Clinical value of ultrasonography in the detection and removal of radiolucent foreign bodies. Injury 1989;20(6):3413. [29] Shiels WE II, Babcock DS, Wilson JL, et al. Localization and guided removal of soft-tissue foreign bodies with sonography. AJR Am J Roentgenol 1990;155(6):127781. [30] Valley VT, Stahmer SA. Targeted musculoarticular sonography in the detection of joint effusions. Acad Emerg Med 2001;8(4):3617. [31] Mayekawa DS, Ralls PW, Kerr RM, et al. Sonographically guided arthrocentesis of the hip. J Ultrasound Med 1989;8(12):6657. [32] Smith SW. Emergency physician-performed ultrasonography-guided hip arthrocentesis. Acad Emerg Med 1999;6(1):846. [33] Roy S, Dewitz A, Paul I. Ultrasound-assisted ankle arthrocentesis. Am J Emerg Med 1999; 17(3):3001. [34] Ozkan B, Kaya O, Akdag R, et al. Suprapubic bladder aspiration with or without ultrasound guidance. Clin Pediatr (Phila) 2000;39(10):6256. [35] Pollack CV Jr, Pollack ES, Andrew ME. Suprapubic bladder aspiration versus urethral catheterization in ill infants: success, efciency and complication rates. Ann Emerg Med 1994;23(2):22530. [36] Gochman RF, Karasic RB, Heller MB. Use of portable ultrasound to assist urine collection by suprapubic aspiration. Ann Emerg Med 1991;20(6):6315. [37] OCallaghan C, McDougall PN. Successful suprapubic aspiration of urine. Arch Dis Child 1987;62:10723. [38] Kiernan SC, Pinckert TL, Keszler M. Ultrasound guidance of suprapubic bladder aspiration in neonates. J Pediatr 1993;123(5):78991. [39] Runyon BA. Care of patients with ascites. N Engl J Med 1994;330(5):33742. [40] Runyon BA. Paracentesis of ascitic uid. A safe procedure. Arch Intern Med 1986;146(11): 225961. [41] Williams JW Jr, Simel DL. The rational clinical examination. Does this patient have ascites? How to divine uid in the abdomen. JAMA 1992;267(19):26458. [42] Bard C, Lafortune M, Breton G. Ascites: ultrasound guidance or blind paracentesis? Can Med Assoc J 1986;135(3):20910. [43] Rozycki GS, Pennington SD, Feliciano DV. Surgeon-performed ultrasound in the critical care setting: its use as an extension of the physical examination to detect pleural effusion. J Trauma 2001;50(4):63642. [44] Ma OJ, Mateer JR. Trauma ultrasound examination versus chest radiography in the detection of hemothorax. Ann Emerg Med 1997;29(3):3125 [discussion: 3156]. [45] Rothlin MA, Naf R, Amgwerd M, et al. Ultrasound in blunt abdominal and thoracic trauma. J Trauma 1993;34(4):48895. [46] Sisley AC, Rozycki GS, Ballard RB, et al. Rapid detection of traumatic effusion using surgeon-performed ultrasonography. J Trauma 1998;44(2):2916 [discussion: 2967]. [47] Weingardt JP, Guico RR, Nemcek AA Jr, et al. Ultrasound ndings following failed, clinically directed thoracenteses. J Clin Ultrasound 1994;22(7):41926. [48] Jones PW, Moyers JP, Rogers JT, et al. Ultrasound-guided thoracentesis: is it a safer method? Chest 2003;123(2):41823. [49] Diacon AH, Brutsche MH, Soler M. Accuracy of pleural puncture sites: a prospective comparison of clinical examination with ultrasound. Chest 2003;123(2):43641. [50] Kohan JM, Poe RH, Israel RH, et al. Value of chest ultrasonography versus decubitus roentgenography for thoracentesis. Am Rev Respir Dis 1986;133(6):11246. [51] OMoore PV, Mueller PR, Simeone JF, et al. Sonographic guidance in diagnostic and therapeutic interventions in the pleural space. AJR Am J Roentgenol 1987;149(1):15.
815
[52] Grogan DR, Irwin RS, Channick R, et al. Complications associated with thoracentesis. A prospective, randomized study comparing three different methods. Arch Intern Med 1990;150(4):8737. [53] Lichtenstein D, Hulot JS, Rabiller A, et al. Feasibility and safety of ultrasound-aided thoracentesis in mechanically ventilated patients. Intensive Care Med 1999;25(9):9558. [54] Gervais DA, Petersein A, Lee MJ, et al. US-guided thoracentesis: requirement for postprocedure chest radiography in patients who receive mechanical ventilation versus patients who breathe spontaneously. Radiology 1997;204(2):5036. [55] Chan D. Echocardiography in thoracic trauma. Emerg Med Clin N Am 1998;16(1): 191207. [56] Tayal V, Moore C, Rose G. Cardiac. In: Ma OJ, Mateer J, editors. Emergency ultrasound. New York: McGraw-Hill; 2003. p. 89127. [57] Mandavia DP, Hoffner RJ, Mahaney K, et al. Bedside echocardiography by emergency physicians. Ann Emerg Med 2001;38(4):37782. [58] Plummer D, Brunette D, Asinger R, et al. Emergency department echocardiography improves outcome in penetrating cardiac injury. Ann Emerg Med 1992;21(6):70912. [59] Rozycki GS, Feliciano DV, Ochsner MG, et al. The role of ultrasound in patients with possible penetrating cardiac wounds: a prospective multicenter study. J Trauma 1999; 46(4):54351 [discussion 55142]. [60] Callahan JA, Seward JB, Tajik AJ. Cardiac tamponade: pericardiocentesis directed by two-dimensional echocardiography. Mayo Clin Proc 1985;60(5):3447. [61] Fagan SM, Chan KL. Pericardiocentesis: blind no more!. Chest 1999;116(2):2756. [62] Krikorian JG, Hancock EW. Pericardiocentesis. Am J Med 1978;65(5):80814. [63] Wong B, Murphy J, Chang CJ, et al. The risk of pericardiocentesis. Am J Cardiol 1979; 44(6):11104. [64] Kilpatrick ZM, Chapman CB. On pericardiocentesis. Am J Cardiol 1965;16(5):7228. [65] Clarke DP, Cosgrove DO. Real-time ultrasound scanning in the planning and guidance of pericardiocentesis. Clin Radiol 1987;38(2):11922. [66] Tsang TS, Enriquez-Sarano M, Freeman WK, et al. Consecutive 1127 therapeutic echocardiographically guided pericardiocenteses: clinical prole, practice patterns, and outcomes spanning 21 years. Mayo Clin Proc 2002;77(5):42936. [67] Tsang TS, Freeman WK, Barnes ME, et al. Rescue echocardiographically guided pericardiocentesis for cardiac perforation complicating catheter-based procedures. The Mayo Clinic experience. J Am Coll Cardiol 1998;32(5):134550. [68] Tsang TS, Freeman WK, Sinak LJ, et al. Echocardiographically guided pericardiocentesis: evolution and state-of-the-art technique. Mayo Clin Proc 1998;73(7):64752. [69] Suehiro S, Hattori K, Shibata T, et al. Echocardiography-guided pericardiocentesis with a needle attached to a probe. Ann Thorac Surg 1996;61(2):7412. [70] Taavitsainen M, Bondestam S, Mankinen P, et al. Ultrasound guidance for pericardiocentesis. Acta Radiol 1991;32(1):911. [71] Tsang TS, El-Najdawi EK, Seward JB, et al. Percutaneous echocardiographically guided pericardiocentesis in pediatric patients: evaluation of safety and efcacy. J Am Soc Echocardiogr 1998;11(11):10727. [72] Hanaki Y, Kamiya H, Todoroki H, et al. New two-dimensional, echocardiographically directed pericardiocentesis in cardiac tamponade. Crit Care Med 1990;18(7):7503. [73] Tsang TS, Seward JB. Pericardiocentesis under echocardiographic guidance. Eur J Echocardiogr 2001;2(1):689. [74] Chiang HT, Lin M. Pericardiocentesis guided by two-dimensional contrast echocardiography. Echocardiography 1993;10(5):4659.
Future applications for emergency ultrasound

Eric Legome, MD*, Diana Pancu, MD
Department of Emergency Medicine, NYU School of Medicine, Bellevue Hospital Center, Emergency Department, 340 A, New York, NY 10016, USA
Focused ultrasound (US) imaging performed by emergency physicians has gained signicant clinical importance during the past decade. A growing body of scientic literature documents the ability of emergency physicians to perform and interpret bedside US evaluations accurately, with a direct impact on the quality of care [16]. The role of US in the emergency department (ED) is to aid in diagnosis, inuence management decisions, facilitate rapid dispositions [79], and evaluate critical patients who are too unstable to be assessed by other imaging modalities. The initial scope of bedside US by emergency physicians was limited to six primary areas of focus. The widespread adaptation of US in emergency practice has led to expanded testing and use of new applications. The following is a discussion of newer indications and uses for emergency sonography that encompass the diagnoses of eye conditions, pneumothorax, and musculoskeletal disorders and prehospital applications. Ocular ultrasound Sonography may aid in the evaluation of multiple emergency ocular conditions. Ocular pathology such as retinal detachments, vitreous and retrobulbar hemorrhages, lens dislocation, globe rupture, foreign body retention, and central retinal artery and vein occlusion can be identied with US. Bedside sonography is benecial particularly in patients in whom soft tissue swelling from facial trauma impedes eyelid retraction. It allows for adequate initial evaluation of the eye when there is a delay in obtaining CT, MRI, or angiography [10,11]. Evaluation of the eye is performed with a high-resolution linear array probe in the range of 9 to 11 MHz. The power output and gain are
* Corresponding author. E-mail address: legome01@med.nyu.edu (E. Legome). 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.011
818
E. Legome, D. Pancu / Emerg Med Clin N Am 22 (2004) 817827
minimized to prevent image distortion by eyelid echoes and to maintain acceptable image quality. A closed eye technique is employed and the eye is scanned in the sagittal and transverse planes. Color ow Doppler imaging is necessary for evaluation of central retinal artery and venous ow. Although high-frequency US has not been shown to have specic deleterious eects, the duration of the examination and the power level setting should be limited to minimize the mechanical energy delivered to the eye. The initial description of ocular US in the emergency medicine literature was in the form of case reports. Globe rupture can be identied by a dierence in globe size between the normal and injured eye and by noting scleral folds. Vitreous detachment is visualized as a collection of echogenic material deposited between the vitreous body and the retina. Retrobulbar hematoma appears as a lucency deep to the retina, and retinal detachment appears as a band of echogenic material within the vitreous body of the eye. Lens dislocation is evident by localizing the position of the lens within the globe [11]. In a descriptive study of 61 patients who presented with acute vision change or with eye trauma, US-trained emergency medicine residents and faculty were able to accurately diagnose all ocular pathology identied with conrmatory thin-section CT of the orbits or ophthalmologic consultation [10]. Ocular pathology included lens dislocation, vitreous hemorrhage, retinal detachment, papilledema, intraocular foreign body, central retinal artery and vein occlusion, and globe rupture. In this study, emergency bedside US agreed with conrmatory studies in 60 of 61 patients, yielding a sensitivity of 100%, a specicity of 97%, a positive predictive value of 96%, and a negative predictive value of 100%. When globe penetration is suspected, a large quantity of US gel should be applied to the closed eye. The transducer does not make direct contact with the eye, thus preventing further damage to the injured globe. With correct technique, US results in less manipulation of the eye than with standard physical examination, is more accurate and specic for the diagnosis of globe penetration, and has an advantage over CT because it does not require the patient to lie at during the evaluation [10]. Bedside ocular sonography may be of use in the evaluation of patients with suspected elevated intracranial pressure [1215]. A prospective observational study of 35 patients who presented with altered mental status after blunt head trauma and loss of consciousness or after a possible spontaneous intracranial hemorrhage demonstrated that an optic sheath diameter greater than 5 mm correlated with head CT evidence of increased intracranial pressure. The optic nerve sheath diameter was measured with a highresolution 10-MHz linear array transducer 3 mm posterior to the globe. Measurements were made for both eyes and then averaged [12]. An average optic nerve sheath diameter of 5 mm or greater had a sensitivity of 100% and a specicity of 95% for predicting CT ndings indicative of elevated intracranial pressure (midline shift from mass eect of 3 mm or greater, a collapsed third ventricle, hydrocephalus, eacement of sulci with evidence of signicant edema, and abnormal mesencephalic cisterns) [12]. Although
819
correlation between optic nerve diameter and intracranial pressure is known [13,14,16], this is the rst study of bedside US optic sheath evaluation and demonstrates that emergency physicians can identify patients with increased intracranial pressure with the use of US. This skill is most benecial for the evaluation of unstable patients, when there is a delay in obtaining the head CT, or when the physical examination is unreliable due to an altered mental status. In addition, papilledema may not occur for several hours after injury. Current literature suggests that ocular sonography by emergency physicians can be performed accurately after 1 hour of lecture and 1 hour of hands-on instruction dedicated to ocular scanning [10,11]. The specics of optimal ocular US training and the ability of emergency ocular sonography to inuence the decision-making of the treating emergency physician beyond assisting with the clinical evaluation, however, remains to be established.
Pneumothorax US is considered to have limited application in respiratory disease because air reects sound waves. US evaluation of the lung depends on the presence of artifacts created by the lungchest wall interface. When a US transducer is placed on a normal chest wall, respiratory movement parallel to the chest is observed at the lung surface. The ribs are identiable by their hyperechoic surface and distal acoustic shadow. Between and below two ribs, a hyperechoic line (ie, the pleural line) readily is identied, deep to which only artifacts are present [1720]. This pleural line appears as lung sliding, a displacement of the visceral pleura with respect to the parietal pleura during respirations. The hyperechoic reverberation artifacts visible deep to the pleural line and spreading toward the lower edge of the screen are called comet-tail artifacts and are caused by the acoustic impedance mismatch of lung, pleura, and air [19,2123]. When a pneumothorax is present, air collection beyond the parietal pleura prevents visualization of the visceral pleura, precluding the observation of lung sliding. In addition, air within the pleural space hinders propagation of sound waves and prevents formation of artifacts. Therefore, the presence of pneumothorax is characterized by the absence of lung sliding and comet-tail artifacts [18,24]. The area where the normal lung ndings on US disappear is known as the lung point, which is a dynamic sign that corresponds to the edge of the pneumothorax. Its eeting appearance is visible with respiratory excursion as the transducer is held in a xed position on the patients chest [25]. The scanning technique requires longitudinal evaluation of the anterior chest wall while the patient is supine. A high-frequency (5.010.0 MHz) linear probe is ideal, although standard abdominal or cardiac probes produce diagnostic images. Probe placement is in the lateral clavicular line in the third to fourth interspace. Conrmation of lung sliding and comet-tail
820
artifacts should be undertaken for four to ve respiratory cycles on both sides of the chest for comparison. The evaluation of lung sliding with power color Doppler results in enhancement of movement at the pleural interface and may improve sensitivity [21,26]. The rst description of the role of US in the diagnosis of pneumothorax was in 1986 by a veterinarian in a study of horses [27]. Subsequent reports have documented the sonographic signs associated with pneumothorax (the absence of normal lung sliding and the lack of comet-tail artifacts) [18,19,24]. Since 1995, four prospective, operator-blinded studies with over 600 patients in total have demonstrated an overall sensitivity of US for pneumothorax of greater than 90% [18,22,24,25,28,29]. Although falsepositive US evaluations occur in patients with underlying lung disease, such as adult respiratory distress syndrome, lung brosis, and bullous disease, the negative predictive value was reported at 100% by all the studies. Table 1 summarizes these ndings. Several authors have suggested that the evaluation of the chest with echocardiography should be used to supplement the Focused Assessment with Sonography for Trauma (FAST) examination [20]. The technique is learned readily by physicians familiar with US use for trauma and adds between 2 and 4 minutes to the FAST evaluation [17,29]. The learning curve appears to be very steep. Surgery residents have been reported to have excellent accuracy with the techniques after ve examinations on healthy subjects [29]. The utility of US for the evaluation of pneumothorax needs to be validated in the ED setting. Although radiography is readily available in most centers, US evaluation can be completed faster than chest radiography and is more sensitive than supine lms [30].
Musculoskeletal applications US is a relatively new modality for the evaluation of musculoskeletal pathology; it has found increasing importance in diagnostic and procedural
Table 1 Summary of studies of ultrasound detection of pneumothorax Study and reference No. Lichtenstein and Menu 1995 [23] Lichtenstein et al, 1999 [22] Dulchavsky et al, 2001 [29] Rowan et al, 2002 [28] No. of patients 111 114 382 27 Ultrasound nding Lung sliding Comet tail Lung sliding and comet tail Lung Sliding and comet tail Gold standard CT CT CXR CT Sensitivity (%) 95.3 100 95 100 Specicity (%) 91.1 60 100 94 NPV (%) 100 100 100 100
Abbreviations: CXR, chest radiography; NPV, negative predictive values.
821
applications. Although unable to provide the full visual assessment aorded by MRI for soft tissues or by plain radiographs and CT for bone, sonography can produce eective results faster than other diagnostic modalities or when they are not available. US has the added advantage of being able to specically examine the aected area in conjunction with the physical examination [31,32]. Research and clinical studies are focusing on uses of US that may be applicable in the future to the ED and the prehospital arena to diagnose injuries such as sports medicine injuries, acute fractures and dislocations, joint and soft tissue infections, and to aid in performing procedures involving the joints and bursae. Fractures Traditional radiography is the standard modality for the initial assessment of orthopedic pathology. US, however, may allow the emergency ultrasonographer to identify long-bone fractures when traditional radiography is delayed or not available. With US, the physician can obtain immediate clinical correlation with an imaging examination focused over the aected area. Although emergency physicians do not yet have sucient experience with US in the assessment of fractures, this application may be of benet in the future for early diagnosis of long-bone injury in unstable multitrauma patients and when urgent operative repair of organs would be delayed by standard radiography. US evaluation of fractures was thought to be of limited utility in orthopedics due to the highly reective surface of bone and the inability to visualize the medullary tissue. It is this characteristic of bone, however, that makes interruptions of the linear bony cortex easy to visualize [33]. US can, in fact, demonstrate occult fractures not identiable by traditional radiography [3336]. Several studies demonstrate that US is a feasible imaging modality in skeletal trauma. In cadaveric specimens, under experimental conditions, investigators detected cortical discontinuities of 1 mm or more, visualized as an interruption of the cortical reection or as a dorsal band of echoes conned to the region of the fracture [33]. Ultrasound is used clinically in the United States primarily for intrauterine skeletal assessment and for the evaluation of pediatric trauma [37,38]. Several prospective studies indicate that US has excellent and potentially clinically applicable accuracy in diagnosing fractures of the orbital oor [39,40] and occult fractures of the foot and ankle [36,41], rib [34], femur [38,42], and humerus [4345]. It also allows visualization of the surrounding soft tissue to assess interosseus membrane integrity and the process of fracture healing [46,47]. Surrounded by large muscle groups that can contain over a liter of blood, femoral shaft fractures can be a signicant cause of hemodynamic instability. Initial treatment and triage decisions are based on clinical suspicion, although a swollen thigh may suggest a femur fracture or a quadriceps hematoma without a fracture. Application of a traction splint
822
to a fractured thigh can provide hemostasis and pain relief to the patient; however, complications include sciatic and peroneal nerves compression, distraction of an unrecognized knee or ankle dislocation, or neurovascular compromise from excessive traction. US detection of femoral shaft fractures has been reported in the literature [38,42,45]. Rapid detection of a femoral shaft fracture assists in expedient and judicious splinting and might aid prioritization in the search for a bleeding source in the hemodynamically unstable patient. This rapid fracture diagnosis may have applications in remote prehospital settings where radiography is not available. US also may be applied in the initial assessment of humeral shaft fractures. Distal humerus intra-articular fractures can be visualized with US, allowing more accurate diagnosis when the radiograph is indeterminate. Proximal humerus shaft fractures are not always identied easily on initial physical examination. Rapid diagnosis and triage in a busy ED may be possible after brief assessment with US, and multiple patients or patients with multiple injuries may be better prioritized for immediate or delayed radiography based on the initial US [48,49]. Soft tissue applications US for evaluation of soft tissue injuries has been used with success in orthopedics and sports medicine. Sonography can be used for focused assessment of the soft tissues including bursae and tendons and may be useful especially when performing a dynamic examination such as assessment for subtle tendon tears [32]. In the sports medicine and orthopedic literature, US has been employed for evaluation of several conditions. Within the shoulder, rotator cu injuries are the most common injury evaluated by US [50]. The radiology literature cites sensitivities and specicities of 93% and 94%, respectively, for focal injuries of the rotator cu tendons [51]. Although the need for acute diagnosis of specic rotator cu injuries in the ED has not been established, there may be sports medicine applications for this technique in the future. Other common indications in the radiology literature include adhesive capsulitis of the shoulder, biceps tendon dislocations, epicondylitis, and tendon injuries of the wrist and hand. Of most relevance to emergency physicians is the evaluation of tendon injuries. The normal appearance of a tendon by US is hyperechoic with a brillar echotexture [31,32]. Specic injuries including full and partial tears, tendonitis, and sprains have unique ndings. High-frequency transducers allow assessment of the tendon architecture. A focal tendon tear may be well dened and appear anechoic or hyperechoic, with cortical irregularity. Tendonitis may present with enhanced echogenicity [31]. Muscle injury may have ndings of hemorrhage, edema, or hematoma. Furthermore, the integrity of the muscle belly can be visualized in a dynamic fashion [52].
823
The Achilles tendon is evaluated easily by US due to its supercial location and prominent size. Clear radiographic criteria exist for tears, tendonitis, and for evaluation during range of motion [5356]. In addition, the gastrocnemius and plantaris tendons have been evaluated for tears by US, which potentially is useful in the ED for the rapid assessment of etiologies for a swollen calf. Ligamentous injuries of the ankle, including tendonitis, tears, and subluxations also have been evaluated by US. This application, however, may be less valuable in the ED setting where clinical ndings such as pain, swelling, and instability guide treatment. In addition to assessment of soft tissue injury, US has been used with success to guide intra-articular and bursal injections, injections of tendon sheaths, and associated aspirations [57,58]. This use of US is most appropriate for the hip. In other joints such as the shoulder and knee, it remains to be determined whether US oers an advantage over traditional methods that use anatomic landmarks. Within the emergency medicine literature, there have been no prospective, operator-blinded studies on soft tissue applications of US. A singular case report was published on pes anserinus tendonitis and on a description of joint eusions [49,59]. As with other US applications beyond the traditional primary indications, however, increased use of US for the evaluation of musculoskeletal injuries will better dene its role in the diagnosis and management of ED patients in the future.
The prehospital setting US has the potential to identify intraperitoneal bleeding in the eld and assist the in-hospital team to prepare for the trauma victim. The FAST examination has been used with apparent success in a prehospital triage setting by physicians who screened victims of an Armenian earthquake in 1988 [60]. US also has been described in the setting of aeromedical transport. A group of emergency physicians, residents, ight nurses, and technologists evaluated real and simulated patients with the FAST examination to test the feasibility of performing US studies during ight conditions. In 21 test conditions, they encountered no signicant obstacles in accomplishing the studies. The pilots did not report any endangerment to the safety of the ights [61]. Literature documenting the utility of US in prehospital aeromedical practice has been less impressive, with relatively small studies revealing obstacles in completing US evaluations. In these studies, technologic, environmental, or operator problems occurred in 16% to 34% of cases [6264]. Signicant investment in training and technology is necessary to make US ecacious in the prehospital aeromedical transport setting. The research to date has shown it to be feasible (although signicant limitations exist),
824
and it is somewhat doubtful whether it eventually will be found to be cost eective and clinically important. It is even less clear whether it can be ecacious in the setting of ground transport. Although the data is limited, a least one report has shown that US may prove essential for the eld management of disasters or in military situations for the rapid assessment and triage of patients [60]. Summary US has become an integral part of patient evaluation in emergency medicine. As discussed in this article, other potential uses for US are relatively early in their development. With the increased interest and study, however, there exists a real potential for US to be employed in an expanded clinical role in the evaluation of traumatically injured patients outside the standard FAST examination, for acute and chronic musculoskeletal injuries, and for triage of patients in disasters. References
[1] Braman BH, Coleman BG, Ramchandani P, Arger PH, Nodine CF, Dinsmore BJ, et al. Emergency department screening for ectopic pregnancy: a prospective US study. Radiology 1994;190(3):797802. [2] Jehle D, Guarino J, Karamanoukian H. Emergency department ultrasound in the evaluation of blunt abdominal trauma. Am J Emerg Med 1993;11(4):3426. [3] Schlager D, Lazzareschi G, Whitten D, Sanders AB. A Prospective study of ultrasonography in the ED by emergency Physicians. Am J. Emerg Med 1994;12(2):1859. [4] Porter RS, Nester BA, Dalsey WC, OMara M, Gleeson T, Pennell R, Beyer FC. Use of ultrasound to determine need for laparotomy in trauma patients. Ann Emerg Med 1997; 29(3):32330. [5] Kuhn M, Bonnin RL, Davey MJ, Rowland JL, Langlois SL. Emergency department ultrasound scanning for abdominal aortic aneurysm: accessible accurate, and advantageous. Ann Emerg Med 2003;36(3):21923. [6] Rosen CL, Brown DF, Chang Y, Moore C, Averill NJ, Arko LJ, et al. Ultrasonography by emergency physicians in patients with suspected cholecystitis. Am J Emerg Med 2001; 19(1):326. [7] Balivas M, Harwood RA, Lambert MJ. Decreasing length of stay with emergency ultrasound examination of the gallbladder. Acad Emerg Med 1999;6(10):10203. [8] Blaivas M, Sierzenski P, Plecque D, Lambert M. Do emergency physicians save time when locating a live intrauterine pregancy with bedside ultrasonography?. Acad Emerg Med 2000;7(9):98893. [9] Shih CH. Eect of emergency physician-performed pelvic sonography on length of stay in the emergency department. Ann Emerg Med 1997;29(3):34851[discussion: 352]. [10] Blaivas M, Theodoro D, Sierzenski PR. A study of bedside ocular ultrasonography in the emergency department. Acad Emerg Med 2002;9(8):7919. [11] Balivas M. Beside emergency department ultrasonography in the evaluation of ocular pathology. Acad Emerg med 2000;7(8):94750. [12] Blaivas M, Theodoro D, Sierzenski PR. Elevated intracranial pressure detected by bedside emergency ultrasonography of the optic nerve sheath. Acad Emerg Med 2003;10(4): 37681.
825
[13] Galetta S, Byrne SF, Smith JL. Echographic correlation of optic nerve sheath size and cerebrospinal uid pressure. J. Clin Neuroophthamlmol 1989;9(2):7982. [14] Hansen HC, Helmke K. Validation of the optic nerve sheath response to changing cerebrospinal uid pressure: ultrasound ndings during intrathecal infusion tests. J Neurosurg 1997;87(1):3440. [15] Gangemi M, Cennamo G, Maiuri F, DAndrea F. Echographic measurement of the optic nerve in patients with intracranial hypertension. Neurochirurgia (Stuttg) 1987;30(2):535. [16] Hayreh SS. Optic disc edema in raised intracranial pressure. V. Pathogenesis. Arch Ophthalmol 1977;95(9):155365. [17] Soldati G, Iacconi P. The validity of the use of ultrasonography in the diagnosis of spontaneous and traumatic pneumothorax. J. Trauma 2001;51(2):423. [18] Wernecke K, Galanski M, Peters PE, Hansen J. Pneumothorax: evaluation by ultrasoundpreliminary results. J Thorac Imaging 1987;2(2):768. [19] Dulckavsky SA, Hamilton DR, Diebel LN, Sargsyan AE, Billica RD, Williamd DR. Thoracic ultrasound diagnosis of pneumothorax. J Trauma 1999;47(5):9701. [20] Chan SS. Emergency bedside ultrasound to detect pneumothorax. Acad Emerg Med 2003; 10(1):914. [21] Cunninghan J, Kirkpatrick AW, Nicolaou S, Liu D, Hamilton DR, Lawless B, et al. Enhanced recognition of lung sliding with power color Doppler imaging in the diagnosis of pneumothorax. J Trauma 2002;52(4):76971. [22] Lichtenstein D, Meziere G, Biderman P, Gepner A. The comet-tail artifact: an ultrasound sign ruling out pneumothorax. Intensive Care Med 1999;25(4):3838. [23] Lichtenstein DA, Menu Y. A bedside ultrasound sign ruling out pneumothorax in the critically ill. Lung sliding. Chest 1995;108(5):13458. [24] Targhetta R, Bourgeois JM, Chavagneux R, Balmes P. Diagnosis of pneumothorax by ultrasound immediately after ultrasonically guided aspiration biopsy. Chest 1992;101(3): 8556. [25] Lichtenstein D, Meziere G, Biderman P, Gepner A. The lung point: an ultrasound sign specic to pneumothorax. Intensive Care Med 2000;26(10):143440. [26] Islam NB, Levy PD. Emergency bedside ultrasound to detect pneumothorax. Acad Emerg Med 2003;10(7):81920[author reply: 820-1]. [27] Ranaten N. Diagnostic ultrasound: diseases of the thorax. Vet Clin N Am 1986;2:4966. [28] Rowan KR, Kirkpatrick AW, Liu D, Forkheim KE, Mayo JR, Nicolaou S. Traumatic pneumothorax detection with thoracic US: correlation with chest radiography and CTinitial experience. Radiology 2002;225(1):2104. [29] Dulchavsky SA, Schwarz KL, Kirkpatrick AW, Billica RD, Williams DR, Diebel WV, et al. Prospective evaluation of thoracic ultrasound in the detection of pneumothorax. J Trauma 2001;50(2):2015. [30] Kirkpatrick AW, Ng AK, Dulchavsky SA, Lybrun I, Harris A, Torregianni W, Simons RK, et al. Sonographic diagnosis of a pneumothorax inapparent on plain radiography: conrmation by computed tomography. J Trauma 2001;50(4):7502. [31] Adler RS. Future and new developments in musculoskeletal ultrasould. Radiol Clin N Am 1999;37(4):62331. [32] Jacobson JA. Ultrasound in sports medicine. Radiol Clin N Am 2002;40(2):36386. [33] Grechenig W, Clement HG, Fellinger M, Seggl W. Scope and limitations of ultrasonography in the documentation of fractures-an experimental study. Arch Orthop Trauma Surg 1998;117(67):36871. [34] Mariacher-Gehler S, Michel BA. Sonography: a simple way to visualize rib fractures. AJR Am J Roentgenol 1994;163(5):1268. [35] Steiner GM, Sprigg A. The value of ultrasound in the assessment of bone. Br J Radiol 1992;65(775):58993. [36] Wang CL, Shieh JY, Wang TG, Hsieh FJ. Sonographic detection of occult fractures in the foot and ankle. J Clin Ultrascoun 1999;27(8):4215.
826
[37] Glesson AP, Stuart MJ, Wilson B, Phillips B. Ultrasound assessment and conservative management of inversion injuries of the ankle in children: plaster of Paris versus Tubigrip. J Bone Joint Surg Br 1996;78(3):4847. [38] Graif M, Stahl-Kent V, Ben-Ami T, Strauss S, Amit Y, Itzchak Y. Sonographic detection of occult bone fractures. Pediatr Radiol 1988;18(5):3835. [39] Hirai T, Manders EK, Nagamoto K, Saggers GS. Ultrasonic observationof facial bone fractures: report of cases. J. Oral maxillofac Surg 1996;54(6):7769[discussion: 779-80]. [40] Jenkins CN, Thuau H. Ultrasound imaging in assessment of fractures of the orbital oor. Clin Radiol 1997;52(9):70811. [41] Clement H, Grechenig W, Peicha G, Grechenig S. [Ultrasound diagnosis in knee and foot trauma]. Orthopade 2002;31(3):3325. [42] Watson NA, Ferrier GM. Diagnosis of femoral shaft fracture in pregnancy by ultrasound. J Accid Emerg Med 1999;16(5):3801. [43] Patten RM, Mack LA, Wang KY, Lingel J. Nondisplaced fractures of the greater tuberosity of the humerus: sonographic detection. Radiology 1992;182(1):2014. [44] Pancione L, Gatti G, Mecozzi B. Diagnosis of Hill-Sachs lesion of the shoulder. Comparison between ultrasonography and arthro-CT. Acta Radiol 1997;38(4 Pt 1):5236. [45] Hollister MS, Mack LA, Patten RM, Winter TC, Matsen FA, Veith RR. Association of sonographically detected subacromial/subdeltoid bursal eusion and intraarticular uid with rotator cu tear. AJR Am J Roentgenon 1995;165(3):6058. [46] Moed BR, Subramanian S, Van Holsbeeck M, Watson JT, Cramer KE, Karges DE, et al. Ultrasound for the early diagnosis of tibial fracture healing after static interlocked nailing without reaming: histologic correlation using a canine model. J Orthop Trauma 1998;12(3): 2005. [47] Failla JM, Jacobson J, van Holsbeeck M. Ultrasould diagnosis and surgical pathology of the torn interosseous membrane in forearm/dislocations. J Hand Surg (Am) 1999;24(2): 25766. [48] Noble VE, Legome E, Marshburn T. Long bone ultrasound: making the diagnosis in remote locations. J Trauma 2003;54(4):800. [49] Valley VT, Stahmer SA. Targeted musculoarticular sonography in detection of joint eusions. Acad Emerg Med 2001;8(4):3617. [50] Martinoli C, Bianchi S, Prato N, Pugliese F, Zamorani MP, Valle M, et al. US of the shoulder: non-rotator cu disorders. Radiographics 2003;23(2):381401[quiz: 534]. [51] van Holsbeek MT, Kolowich PA, Eyler WR, Craig JG, Shirazi KK, Habra GK, et al. US depiction of partial-thickness tear of the rotator cu. Radiology 1995;197(2):4436. [52] Takebayashi S, Takasawa H, Banzai Y, Miki H, Sasaki R, Itoh Y, et al. Sonographic ndings in muscle strain injury: clinical and MR imaging correlation. J Ultrasound Med 1995;14(12):899905. [53] Grechenig W, Clement HG, Fellinger M, Seggl W. [Value of ultrasound imaging of the Acilles rendon in traumatology]. Radiologe 1997;37(4):3229. [54] Grechenig W, Clement H, Bratschitsch G, Frankhauser F, Peicha G. [Ultrasound diagnosis of the Acilles tendon]. Orthopade 2002;31(3):31925. [55] Fessel DP, Vanderschueren GM, Jacobson JA, Cuelemans RY, Prasad A, Craig JE, et al. US of the ankle: technique, anatomy, and diagnosis of pathologic conditions. Radiographics 1998;18(2):32540. [56] OReilly MA, Massouth H. Pictorial review: the sonographic diagnosis of pathology in the Achilles tenson. Clin Radiol 1993;48(3):2026. [57] Cardinal E, Bureau NJ, Aubin B, Chhem RK. Role of ultrasound in musuloskeletal infections. Radiol Clin N Am 2001;39(2):191201. [58] Cardinal E, Chhem RK, Beauregard CG. Ultrasound-guided interventional procedures in the musculoskeletal system. Radiol Clin N Am 1998;36(3):597604. [59] Valley VT, Shermer CD. Use of musculoskeletal ultrasonography in the diagnosis of pes anserine tendinities: a case report. J Emerg Med 2001;20(1):435.
827
[60] Sarkisian AE, Khondkarian RA, Amirbekian NM, Bagdasarian NB, Khojayan RL, Oganesian YT. Sonographic screening of mass casualties for abdominal and renal injuries following the 1988 Armenian earthquake. J Trauma 1991;31(2):24750. [61] Price DD, Wilson SR, Murphy TG. Trauma ultrasound feasibility during helicopter transport. Air Med J 2000;19(4):1446. [62] Polk JD, Fallon WF Jr., Kovach B, Mancuso C, Stephens M, Malangoni MA. The Airmedical F.A.S.T. for trauma patients-the initial report of a novel application for sonography. Aviat Spce Environ Med 2001;72(5):4326. [63] Polk JD, Fallon WF Jr. The use of focused assessment with sonography for trauma (FAST) by a prehospital air medical team in the trauma arrest patient. Prehosp Emerg Care 2000;4(1):824. [64] Melanson SW, McCarthy J, Stromski CJ, Kostenbader J, Heller M. Aeromedical trauma sonography by ight crews with a miniature ultrasound unit. Preshosp Emerg Care 2001; 5(4):399402.
Emergency medicine ultrasound policies and reimbursement guidelines

Evelyn Cardenas, MD, FACEPa,b
Department of Emergency Medicine, Los Robles Regional Medical Center, Thousand Oaks, CA 91360, USA b Assistant Clinical Professor, Department of Medicine (Emergency Medicine), UCSF/Fresno Medical Education Program, University Medical Center, Fresno, CA 93703, USA
a
Few technologies are capable of generating as much impassioned debate and discussion as the subject of emergency ultrasound imaging. Bedside ultrasound imaging by emergency medicine has attracted more attention than virtually any other bedside procedure in recent memory. Regardless of ones point of view, every emergency medicine physician should be knowledgeable of current emergency medicine expert policy and recommendations regarding the use of ultrasound imaging technology in the emergency department (ED). In the last 3 years, a number of publications have provided critical information on practice management and reimbursement issues relating to bedside ultrasound performed by emergency physicians. The following is a summary of the salient points of these publications. Because of their potential impact and relevance to emergency medicine, specic policies issued from the general house of medicine that relate to ultrasound technology also are included in this article.
Model of the Clinical Practice of Emergency Medicine The model of the Clinical Practice of Emergency Medicine (EM model) [1] replaces the Core Content of Emergency Medicine as the document that denes the body of knowledge and the scope of practice of the specialty. This document is the result of a collaborative eort of the American Board of Emergency Medicine, the American College of Emergency Physicians (ACEP), the Society for Academic Emergency Medicine,
E-mail address: emjogger@aol.com 0733-8627/04/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.emc.2004.04.006
830
E. Cardenas / Emerg Med Clin N Am 22 (2004) 829838
the Council of Emergency Medicine Residency Directors, the Emergency Medicine Residents Association, and the Residency Review Committee for Emergency Medicine. The EM model describes disease presentations, tasks, and skills central to the practice of emergency medicine by board-certied emergency physicians. The EM model specically lists bedside ultrasound skills among the necessary skills that an emergency physician should possess and states that bedside ultrasound is within the scope of practice of an emergency physician. Given that this document will serve as a foundation for future certication and recertication examinations, it is advisable that practicing emergency physicians acquire the knowledge and skills to be able to use ultrasound in the routine practice of emergency medicine. The EM model subsequently has led the Residency Review Committee for Emergency Medicine to include bedside ultrasound skills training in the accreditation review process of training programs accredited by the Accreditation Council for Graduate Medical Education [2].
American College of Emergency Physicians policy statement For over 10 years, the ACEP has had a policy regarding ultrasound imaging by emergency physicians. In 2001, ACEP published its updated policy (Box 1) [3]. This document recognizes that the use of ultrasound imaging by emergency physicians is appropriate in certain clinical situations and that emergency physicians should be reimbursed appropriately for providing emergency ultrasound procedures in the ED.
American College of Emergency Physicians Emergency Ultrasound Guidelines 2001 The ACEP Emergency Ultrasound Guidelines 2001 (Guidelines 2001) [4] is a landmark publication that denes the state of ultrasonography in emergency medicine clinical practice. It is the rst comprehensive publication of its kind since the rst emergency medicine ultrasound curriculum was published in 1994 [5]. This document is a collaborative eort of several emergency medicine physicians with expertise in emergency ultrasound. Guidelines 2001 reviews current consensus indications for use and training, prociency, and credentialing guidelines for emergency physicians who perform ultrasound imaging procedures in the ED. Guidelines 2001 is mandatory reading for ED managers, emergency physician group directors, and every emergency physician who seeks to integrate ultrasound imaging into his or her bedside practice. Guidelines 2001 denes the primary indications for the application of emergency physicianperformed bedside ultrasound imaging. The following primary ultrasound applications are discussed in detail, including
831
appropriate clinical indications and recommended ultrasound techniques/ views: 1. 2. 3. 4. 5. 6. Trauma ultrasound Emergency ultrasound in pregnancy Emergency imaging of the abdominal aorta Biliary ultrasound Renal ultrasound Procedural ultrasound
Training and prociency issues are discussed at great length in this document, including consensus recommendations for required didactic instruction and hands-on training. A minimum of 25 practice scans performed per anatomic application (minimum of 150 scans) is dened as the current standard training and prociency numeric goal for the emergency physician. These are recommendations developed by emergency physicians for emergency physicians, given the considerations of how ultrasound imaging technology is applied in emergency medicine practice. It should be noted that these are minimum guidelines only and that individuals may have dierent learning curves and require more or less experiential training. Other training guidelines have been published from nonemergency medicine organizations in the past, but the emergency medicine physician is advised to use the Guidelines 2001 document as the template to develop pathways to credentialing in a hospital-based practice.
Residency training and emergency ultrasound The status of ultrasound education in American emergency medicine residency programs is dynamic and changing in favor of continued expansion. The listing of ultrasound imaging skills in the EM model served as an impetus to expand education at the residency level. Deemed as an integral part of emergency medicine practice, residents should gain prociency in this skill. Heller et al [6] discussed consensus opinion regarding the unique issues of integrating ultrasound education into the residency curriculum. Experiential, didactic, and equipment curriculum requirements are discussed in detail in their article that summarized the opinion of the Scope of Training Task Force of the American Board of Emergency Medicine. Although the mandate is clear, the degree of ultrasound training is not yet equal across all programs. In a 2001 survey of 122 emergency medicine residency programs accredited by the Accreditation Council for Graduate Medical Education, Counselman et al [2] found that most residencies oer some form of training. The survey also found that time devoted to ultrasound training and the didactic component of training was variable and not uniform. The Focused Assessment with Sonography for Trauma examination was the
832
Box 1. Use of ultrasound imaging by emergency physicians Ultrasound imaging enhances the physicians ability to evaluate, diagnose, and treat emergency department (ED) patients. Because ultrasound imaging is often time-dependent in the acutely ill or injured patient, the emergency physician is in an ideal position to use this technology. Focused ultrasound examinations provide immediate information and can answer specic questions about the patients physical condition. Such bedside ultrasound imaging is within the scope of practice of emergency physicians. Therefore, the American College of Emergency Physicians (ACEP) endorses the following principles: Bedside ultrasound evaluation, including examination, interpretation, and equipment, should be immediately available 24 hours a day for ED patients. Emergency physicians providing emergency ultrasound services should possess appropriate training and hands-on experience to perform and interpret limited bedside ultrasound imaging. The use of ultrasound imaging by emergency physicians is appropriate in clinical situations that include, but are not limited to: thoracoabdominal trauma, ectopic pregnancy, abdominal aortic aneurysm, pericardial effusion, determining cardiac activity, biliary disease, renal tract disease, and procedures that would benet from assistance of ultrasound. Emergency ultrasound procedures and interpretations are standard emergency physician skills that should be delineated in emergency physician privileges. Dedicated ultrasound equipment within the ED should be considered optimal for patient care. Each hospital medical staff should review and approve criteria for granting ultrasound privileges based on background and training for the use of ultrasound technology and ensure that these criteria are in accordance with recommended training and education standards developed by each physicians respective specialty. Training in performing and interpreting ultrasound imaging studies should be included in emergency medicine residency curricula.
833
Continued research in the area of ultrasound should be encouraged. Emergency physicians should be appropriately reimbursed for providing emergency ultrasound procedures in the ED. This policy statement was prepared by the Emergency Medicine Practice Committee. It was approved by the ACEP Board of Directors June 2001. It replaces one by the same title approve by the ACEP Board of Directors June 1997. The original policy statement titled Use of Ultrasound for Emergency Department Patients was approved by the ACEP Board of Directors January 1991. most common procedure being taught, and most ultrasound instruction was oered by emergency medicine faculty. Emergency ultrasound and the American Medical Association In 2000, general acceptance from the house of medicine for bedside ultrasound by emergency physicians occurred through resolutions by the American Medical Association (AMA) House of Delegates. In Substitute HOD-AMA Resolution 108reimbursement for oce-based or outpatient ultrasound imagingthey rearmed Policy H-230.960. This policy states the following: 1. AMA afrms that ultrasound imaging is within the scope of practice of appropriately trained physicians. 2. AMA policy on ultrasound acknowledges that broad and diverse use and application of ultrasound imaging technologies exist in practice. 3. AMA policy on ultrasound imaging afrms that privileging of the physician to perform ultrasound imaging procedures in a hospital setting should be a function of hospital medical staffs and should be specically delineated on the departments Delineation of Privileges form. 4. AMA policy on ultrasound imaging states that each hospital medical staff should review and approve criteria for granting ultrasound privileges based on background and training for the use of ultrasound technology and strongly recommends that these criteria are in accordance with the recommended training and education standards developed by each physicians respective specialty. In addition, Resolution 108 (1) calls for AMA policy to support reimbursement for ultrasound imaging performed by appropriately trained physicians in the oce or other outpatient setting, including outpatient hospital settings such as the ED and obstetric units; (2) recognizes that
834
a physician is appropriately trained to perform ultrasound imaging in the oce or other outpatient settings if that physician meets the training and education criteria for ultrasound imaging as determined by his specialty; and (3) asks that AMA noties health insurers, managed care organizations, the Health Care Financing Administration, and other third-party payers of these policies and encourages them to modify their own coverage and payment policies to reimburse physicians for ultrasound imaging performed in the oce or other outpatient setting.
American College of Emergency Physicians emergency ultrasound coding and reimbursement document The ACEP Ultrasound Section [7] has compiled a detailed reference document that explains the issues and facts related to appropriate emergency ultrasound procedure coding. Again, this document is required reading for ED managers and directors and those emergency physicians who perform ultrasound procedures in their daily practice. Tables 1 and 2 outline the primary clinical indications for emergency ultrasound, along with the appropriate corresponding current procedural terminology (CPT) codes. Other nonprimary uses of ultrasound in the ED (testicular torsion, deep venous thrombosis, musculoskeletal) and appropriate corresponding codes also are discussed in this document. The entire document can be downloaded for viewing from the ACEP Web site (www.acep.org). The following is a list of general facts taken from the document: Almost all ultrasound procedures performed by emergency physicians are described by CPT. CPT coding is not specialty specicthat is, all physicians, regardless of specialty, must use the same CPT codes. CPT codes for ultrasound procedures typically are dened as complete; however, code denitions for limited studies exist. Emergency physician ultrasound procedures typically are limited procedures and focused to one anatomic region or a single problem. A limited code denition for transvaginal imaging does not exist. CPT code 76830 is considered a complete procedure. If used by an emergency physician when evaluating a focused problem (Fetal Heart Tones, gestation location, and so forth), 76830 should be adjusted by the addition of the appropriate CPT modier (-52). CPT modiers add more information to the code denition. A detailed list of modiers can be found in the Current Procedural Terminology Manual [8]. Two common CPT modiers are -26, the professional component modier, and -52, the reduced service modier. The professional component modier (-26) denotes the physicians professional service of interpretation of the ultrasound study in a separate and signed written document/report. Hospital-based emergency physicians should
E. Cardenas / Emerg Med Clin N Am 22 (2004) 829838 Table 1 Summary of ultrasound procedures and related current procedural terminology codes Clinical indication/application FAST examination to evaluate for hemoperitoneum CPT code 76705-26 Code description
835
93308-26
Evaluation of abdominal aorta (r/o Abdominal Aortic Aneurysm) Evaluation of pericardial space (r/o eusion) Pregnancy state is known prior to study and ultrasound examination is utilized to determine status of the pregnancy or to evaluate for a pregnancy-related condition Pregnancy state not known prior to study or ultrasound is utilized to assess for a nonpregnancy-related pelvic condition
76775-26
93308-26
76815-26
Echography, abdominal, B-scan and/or real-time with image documentation limited (eg, single organ, quadrant, follow-up) Echography, transthoracic, real-time with image documentation (2-D) or without m-Mode recording, follow-up or limited Echography, retroperitoneal (eg, renal, aorta, nodes), B-scan and/or real-time with image documentation; limited Echography, transthoracic, real-time with image documentation (2-D) or without M-mode recording, follow-up or limited Echography, pregnant uterus, B-scan and/or real-time with image documentation; limited (fetal size, heart beat, placental location, fetal position), or emergency in delivery room
76857-26
Echography, pelvic (nonobstetric), B-scan, and/or real-time with image documentation; limited or follow-up
Evaluation of renal disease (ie, hydronephrosis, hydroureter, stone and so forth) Evaluation of biliary tract disease (ie, ductal dilatation, stones, signs of cholycystitis, and so forth)
and/or 76830-52, -26 76775-26
and/or Echography, transvaginal Echography, retroperitoneal (eg, renal, aorta, nodes), B-scan and/or real-time with documentation; limited Echography, abdominal, B-scan and/or real-time with image documentation; limited (eg, single organ, quadrant, or follow-up)
76705-26
Abbreviations: AAA, Abdominal Aortic Aneurysm; CPT, Current Procedural Terminology; 2-D, two-dimensional; FAST, Focused Assessment with Sonography for Trauma.
use the professional component modier when using ultrasound CPT codes. It would be inappropriate to use global coding (reporting an unmodied code) because the emergency physician does not own and maintain the hospital ultrasound equipment. The reduced service modier (-52) is used to denote a reduction in the service dened by
836
Table 2 Ultrasound procedure guidance and related current procedural terminology codes Clinical indication/application Pericardiocentesis Vascular access CPT code Code description 76930-26 76937a Ultrasonic guidance for pericardiocentesis, imaging supervision and interpretation Ultrasonic guidance for vascular access requiring ultrasound evaluation of potential access sites, documentation of selected vessel patency, and concurrent real-time ultrasound visualization of vascular needle entry, with permanent recording and reporting (list separately in addition to code for the primary procedure) Ultrasonic guidance for needle placement (eg, biopsy, aspiration, injection localization, device imaging and supervision)
Ultrasound guidance for needle 76942-26 placement when performing paracentesis, thoracentesis, suprapubic aspiration, foreign body localization, location of abscess for drainage (eg, peritonsillar abscess)
a
Abbreviation: CPT, current procedural terminology. New code in 2004.
the CPT code denition (as in the previous example with transvaginal echography). Medical necessity should be documented in the physicians procedure note, along with the procedure description and study ndings. Appropriate International Classication of Diseases9 coding of the medical record will support the medical necessity of ultrasound procedures. Emergency physicianperformed ultrasound procedure notes should include the following: A written interpretation and report for each procedure performed should be maintained in the patients medical record Documentation/indication of the medical necessity of the ultrasound procedure A description of organs and structures studied and an interpretation of the ultrasound ndings Who specically performed the procedure Documentation of the scope of the study (complete versus limited, repeat by same physician or different physician, reduced level of service) Image retention is necessary when the hospital bills the technical component of the ultrasound procedure. Under CPT, image archiving by the physician is not required if the professional component modier (-26) is coded alone.
837
Medicare policy regarding documentation requirements and image retention is variable between carrier jurisdictions. Emergency physicians should consult the Centers for Medicare and Medicaid Servicessponsored Local Medical Review Policies Web site (www.lmrp.net) to determine the policies specic to the Medicare carrier in their practice location. Emergency physicians are advised to review this document carefully with their billing manager to ensure that their work and the procedures they perform are documented and coded appropriately for claims submission.
2004 Current procedural terminology coding updates A new CPT code76937has been developed to report ultrasound guidance for vascular access [9]. It is assumed that the same physician who performs the guidance procedure and the line placement procedure will report both of the codes describing the services performed. Use of code 76937 includes ultrasound imaging for preaccess assessment of venous site patency and the real-time guidance of the needle into the vein lumen. Reporting of code 76937 requires 1. Preaccess vein assessment with ultrasound imaging 2. Permanent image recording of sites examined/site chosen for use in line placement 3. Guidance procedure: real-time visualization with ultrasound as the needle is advanced into selected vein 4. Permanent recording showing the needle entering the vein lumen 5. Written/typed description of the preaccess assessment and the guidance procedure included in the patient medical record, along with the previously designated images Code 76937 should not be reported in the situation in which ultrasound is used only to locate a vein and mark a skin entry site. Such nonguided procedures do not meet the denition of 76937.
References
[1] Hockberger RS, Binder LS, Graber MA, et al. Model of the clinical practice of emergency medicine. Ann Emerg Med 2001;37(6):74570. [2] Counselman FL, Sanders A, Slovis CM, et al. The status of bedside ultrasonography training in emergency medicine residency programs. Acad Emerg Med 2003;10(1):3742. [3] American College of Emergency Physicians. Use of ultrasound imaging by emergency physicians. Policy 400121. Available at: www.acep.org. Accessed May 8th 2004. [4] American College of Emergency Physicians. Emergency Ultrasound Guidelines 2001. Available at: www.acep.org. Accessed May 8th 2004. [5] Mateer J, Plummer D, Heller M, et al. Model curriculum for physician training in emergency ultrasonography. Ann Emerg Med 1994;23:95102.
838
[6] Heller MB, Mandavia D, Tayal VS, et al. Residency training in emergency ultrasound: fullling the mandate. Acad Emerg Med 2002;9(8):8359. [7] American College of Emergency Physicians. Emergency ultrasound coding and reimbursement. Available at: www.acep.org. Accessed May 8th 2004. [8] American Medical Association. Current procedural terminology 2001. Available at: www.ama-assn.org. Accessed May 8th 2004. [9] American Medical Association. Current procedural terminology changes 2004an insiders view. Available at: www.ama-assn.org. Accessed May 8th 2004.

2004 Vol.22 Issues 3 Ultrasound in Emergency Medicine

Încărcat de

Informații document

Descriere originală:

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

2004 Vol.22 Issues 3 Ultrasound in Emergency Medicine

Încărcat de

Drepturi de autor:

Formate disponibile

Emerg Med Clin N Am 22 (2004) xvxvi

Ultrasound in Emergency Medicine

Carlo L. Rosen, MD Richard E. Wolfe, MD Guest Editors

C.L. Rosen, R.E. Wolfe / Emerg Med Clin N Am 22 (2004) xvxvi

Emerg Med Clin N Am 22 (2004) 581599

Ultrasound in abdominal trauma

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

Fig. 1. Normal view of hepatorenal interface (Morisons pouch).

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

Fig. 2. Free uid in hepatorenal interface.

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

Fig. 4. FAST examination views.

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

Fig. 6. Fluid in pericolic gutter. Note oating bowel loops.

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

Other extra-abdominal source/DPL

CT and/or Serial exams (+/FAST)

Other extra-abdominal source/DPL

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

J.S. Rose / Emerg Med Clin N Am 22 (2004) 581599

Emerg Med Clin N Am 22 (2004) 601619

Bedside echocardiography in chest trauma

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

D.P. Mandavia, A. Joseph / Emerg Med Clin N Am 22 (2004) 601619

Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640

T.J. Ciccone, S.A. Grossman / Emerg Med Clin N Am 22 (2004) 621640