Chest Pain
Chest Pain
RADIOLOGIC
CLINICS
OF NORTH AMERICA
Radiol Clin N Am 44 (2006) 165–179
Chest pain is one of the most common chief com- location, duration, radiation, quality, and exacer-
plaints in emergency medicine. During the acute bating and relieving factors. A detailed history sets
presentation of a patient who has chest pain, chest in motion further diagnostic testing and manage-
imaging is invaluable, especially in the initial sta- ment decisions.
bilization of a life-threatening cardiac or pulmo-
nary event. The initial approach to evaluating chest
pain includes excluding life-threatening causes, such Major pathologies that produce chest pain
as aortic dissection, pulmonary embolism (PE),
pneumothorax, pneumomediastinum, pericarditis, Pneumothorax
and esophageal perforation. Perfect coupling between the visceral and parietal
The evaluation of an unstable patient who has pleura is required for effective ventilation. Patients
chest pain or shortness of breath begins with a pri- who have pneumothorax have gas in the intra-
mary medical survey to evaluate airway, breathing, pleural space. This abnormality uncouples the vis-
and circulation. In tandem with this rapid assess- ceral and parietal pleura and thus elevates the
ment, the emergency physician requests radio- intrapleural pressure, which affects ventilation, gas
graphic images of the chest, which provide exchange, and perfusion.
visualization of the thoracic anatomy. The first Pneumothorax commonly is divided into two
image obtained is the anteroposterior chest radio- types: primary spontaneous pneumothorax (PSP),
graph, using portable radiography or fixed equip- which usually occurs without a precipitating event
ment, depending on the patient’s presenting in patients who have no clinical lung disease, and
clinical appearance. The initial study is invaluable secondary spontaneous pneumothorax, a compli-
in providing clinically relevant information that cation of underlying lung disease. In actuality, most
directs the patient’s care. patients who have PSP have underlying lung dis-
Although technologic advances have improved ease, most commonly rupture of a subpleural bleb
diagnostic accuracy greatly in recent years, a thor- [1]. Iatrogenic pneumothorax is difficult to iden-
ough history and physical examination remain the tify; its incidence is increasing due to the more
most important components in the evaluation pro- widespread use of mechanical ventilation and
cess. It is imperative to obtain as many details interventional procedures such as central line
about the pain as possible, including its onset, placement and lung biopsy [2]. When pneumo-
Division of Emergency Medicine, University of Maryland School of Medicine, Baltimore, MD, USA
* Corresponding author. Division of Emergency Medicine, University of Maryland School of Medicine,
110 South Paca Street, Sixth Floor, Suite 200, Baltimore, MD 21201.
E-mail address: [email protected] (K.H. Butler).
0033-8389/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.rcl.2005.11.002
radiologic.theclinics.com
166 Butler & Swencki
As little as 50 mL of pleural gas can be seen on chest tion develops when injured tissue forms a one-way
film. A lateral chest film with a 1-cm intrapleural valve and allows air to enter the pleural space but
space corresponds to a 10% pneumothorax. The prevents it from escaping naturally. Arising from
size of the pneumothorax is accounted for by the numerous causes, this condition progresses rapidly
collapsed lung and, to a lesser degree, the expand- to respiratory insufficiency, cardiovascular collapse,
ing chest cage. and, ultimately, death if it is unrecognized and un-
The value of expiratory chest radiographs in treated. Favorable patient outcomes require urgent
detecting pneumothoraces has been overstated. In clinical diagnosis and immediate management.
a study of 85 patients who had pneumothoraces
and 93 controls, inspiratory and expiratory upright Conditions that mimic pneumothorax
chest radiographs had equal sensitivity for pneu- Large subplural bullae can mimic a loculated pneu-
mothorax detection [9]. Because expiratory films mothorax. In most cases, the medial border of the
provide no added benefit, only inspiratory films bulla is concave toward the chest wall, whereas a
are recommended as the initial radiograph of visceral pleural contour is straight or convex later-
choice for pneumothorax. ally. Skin folds can be differentiated from a pneu-
In the supine patient, approximately 500 mL of mothorax by density profile: they form a negative
pleural air is needed for definitive diagnosis of black Mach band instead of the white visceral
pneumothorax [10]. The pleural gas accumulates pleural line. Skin folds increase gradually in opac-
in the subpulmonic location and outlines the ante- ity, with an abrupt drop-off at the edge, and
rior pleural reflection, the costophrenic sulcus, and usually extend beyond the ribcage or stop short of
the anterolateral border of the mediastinum. The the ribs.
overall transradiancy of the entire affected hemi- Bilateral pneumothoraces may be seen after
thorax can be increased on the side of a pneu- heart/lung transplant surgery. Replacement of the
mothorax in the recumbent patient. heart and lungs leaves an open communication
Small pneumothoraces can be visualized more between the two sides of the thorax, which may
easily in the lateral decubitus view. In this position, allow air or fluid to shift from one side to the other.
as little as 5 mL of pleural gas is visible on the Extensive mediastinal dissection can disrupt the
nondependent side [10]. anterior junction line, allowing a unilateral pneu-
mothorax to propagate to the contralateral hemi-
thorax. Placement of a single thoracotomy tube
Ultrasound detection of pneumothorax
decompresses and evacuates both pleural cavities.
Bedside ultrasound has become standard in most
emergency departments. Focused abdominal so-
Treatment of pneumothorax
nography for trauma has been integrated into the
The treatment of pneumothorax is based on its clas-
assessment of the unstable patient. A key element
sification. A tension pneumothorax usually results in
in ultrasound assessment of the chest for pneu-
cardiopulmonary compromise (shock, bradycardia,
mothorax is the presence or absence of the ‘‘sliding
hypoxia) and requires immediate needle decom-
lung sign.’’ On ultrasound of the normal chest, the
pression (thoracentesis), which can be accomplished
lung surface can be seen sliding along the chest wall
by inserting a large-bore (16- or 18-gauge) needle
during inspiration and expiration. In a patient who
(smaller needles are satisfactory for premature in-
has pneumothorax, this sign is absent, which sug-
fants, newborns, and infants) through the second
gests that the air adjacent to the chest wall is not
or third interspace (near the apex of the lung) in
contained within the lung.
the midclavicular line. Immediate decompression cannot
Ultrasound has proven to be more sensitive
wait for radiographic confirmation. Tube thoracostomy
than flat anteroposterior chest radiography in the
may be required after the initial decompression if the
diagnosis of trauma-induced pneumothorax. Ultra-
pneumothorax reaccumulates.
sound provides added benefit by allowing sonolo-
Management of a simple pneumothorax de-
gists to differentiate between small, medium, and
pends on its size and cause. A clinically stable pa-
large pneumothoraces, with good agreement with
tient who has a small PSP (occupying <15% of the
CT results [11].
hemithorax) should be observed in the emergency
department for 3 to 6 hours and discharged home
Tension pneumothorax if a repeat chest film demonstrates no progression
Tension pneumothorax shows a distinct shift of the of the pneumothorax. If the patient is to be ad-
mediastinum to the contralateral side and flatten- mitted to the hospital, oxygen therapy may be ini-
ing or inversion of the ipsilateral hemidiaphragm. tiated to hasten absorption of the pneumothorax.
This is the result of accumulation of air under Clinically stable patients who have a large PSP should
pressure in the pleural space. This emergent condi- be admitted to the hospital for tube thoracostomy.
168 Butler & Swencki
Viral pneumonias Viral pneumonias are located of pneumonia. Pleural effusion is evident in one
predominantly in spaces along and around the third of cases.
alveoli. Therefore, these pneumonias appear reticu- In immunosuppressed patients, distinctive, rounded,
lar on plain radiograph and often are bilateral and nodular opacities may be seen; these lesions may
diffuse in distribution. Associated with thickening expand or cavitate. Pulmonary abscesses may occur
of the interlobular septa, viral pneumonias can be in the immunosuppressed host. Infiltration that
associated with Kerly B lines on chest radiograph. progresses on chest radiography, despite appropriate
Rarely associated with complications or even pleu- antibiotic therapy, is common, and radiographic
ral effusion, viral pneumonia can lead to secondary improvement lags behind clinical improvement
bacterial pneumonias. In viral pneumonia, four by several days. Complete clearing of infiltrates re-
radiographic findings are common: parahilar peri- quires 1 to 4 months.
bronchial infiltrates, hyperexpansion, segmental or
lobar atelectasis, and hilar adenopathy. Acute coronary syndrome
Acute coronary syndrome (ACS) is a spectrum of
Pneumocystis carinii pneumonia Although the acute myocardial ischemia that spans acute myo-
radiographic findings in patients who have Pneu- cardial infarction (AMI) and unstable angina [36].
mocystis carinii pneumonia (PCP) vary, most chest Less than 25% of patients who are admitted with
radiographs reveal bilateral, symmetric, fine to me- suspicion of ACS still have this diagnosis at dis-
dium reticular heterogeneous opacities [23–25]. As charge [37].
the disease worsens, the opacities coalesce and
eventually appear as a bilateral homogeneous con- History and physical examination
solidation. Uncommonly, a more coarse reticular Chest pain or discomfort is the most common
pattern or a miliary pattern may be noted [25,26]. presenting complaint in patients who have ACS
Unilateral or unilobar involvement may occur, [37]. The character and radiation of the pain are
but the radiographic pattern remains fine reticular important for the diagnosis [38]. The pain usually
opacities. Predominant upper lobe involvement is described as a deep visceral discomfort and may
occurs with increased frequency in patients who be difficult to localize to one region of the chest
have used aerosolized pentamidine for prophylaxis [38]. The character of the pain often is described as
[21,27,28]; however, because the use of this form pressure, a weight on the chest, tightness, constric-
of prophylaxis has waned, the incidence of this tion about the throat, or an aching feeling. The
appearance has decreased. The presence of hilar pain is not affected by respiration or movement.
or mediastinal adenopathy as well as pleural fluid Beginning gradually and reaching maximum sever-
is rare and suggests another disease process. Usu- ity after 2 or 3 minutes, the pain lasts for minutes
ally, these findings are seen in patients who have or longer [38]. Physical exertion or emotional stress
been taking aerosolized pentamidine and have may be associated with the onset of pain, and
developed disseminated pneumocystosis [29,30]. the pain may subside with rest [36]. Radiation of
Cases of pneumocystosis following the use of dap- the pain to the arm or neck increases the likeli-
sone prophylaxis have been reported [31]. Calcified hood of AMI [38]. The patient may have associated
hilar and mediastinal lymph nodes have been symptoms of shortness of breath, nausea, vomiting,
reported but are rare [32]. Approximately 10% of profound weakness, dizziness, palpitations, and
patients who have HIV disease and subsequently diaphoresis [36].
proven PCP have had normal chest radiographs. Chest pain is absent in up to 6.2% of patients
In some circumstances, gallium scanning or high- who have ACS and 9.8% of patients who have AMI
resolution CT may demonstrate lung abnormal- [39]. Atypical presentations are more likely in
ities, particularly ground-glass opacities [33]. In elderly patients and diabetics, who have an altered
many institutions, however, treatment is recom- ability to localize symptoms [38], and in women
mended empirically, without a request for further and younger people [36]. Atypical symptoms
imaging [34,35]. include epigastric pain, indigestion, stabbing chest
pain, pleuritic chest pain, chest pain that is repro-
Legionella Virtually all patients who have Legion- ducible on palpation, and isolated dyspnea [36].
naire’s disease have abnormal chest radiography Risk factors for cardiac disease are elicited during
that shows pulmonary infiltrates at the time of the history. Traditional risk factors for coronary
clinical presentation. In a few cases of nosocomial artery disease (CAD) include hypertension, hyper-
disease, fever and respiratory tract symptoms have cholesterolemia, cigarette smoking, diabetes, pe-
preceded the appearance of the infiltrate on chest ripheral vascular disease, family history of CAD,
radiography. Findings on chest films are nonspe- personal history of CAD, male gender, and increas-
cific and do not distinguish Legionella from causes ing age [36–38]. These are long-term risk factors for
170 Butler & Swencki
CAD; the absence of risk factors for CAD should 87% and a specificity of 96% for AMI [38]. These
not be used to exclude the diagnosis of ACS serial tests should be performed over 4 to 9 hours.
[37,38].
Myoglobin Serum myoglobin is another non-
The physical examination of a patient in whom
specific biomarker that appears in the peripheral
ACS is suspected generally is not helpful unless
circulation as early as 1 to 2 hours after muscle
it reveals an alternate diagnosis [37]. Thus, the
damage [36]. Again, the sensitivity of myoglobin
physical examination should focus on excluding
measurements in the diagnosis of AMI increases
other diagnoses; identifying causes of myocardial
with serial measurements [38,41]. Serum myo-
ischemia, such as uncontrolled hypertension or
globin levels should not be used in isolation
thyroid disease; and searching for signs of hemo-
for the diagnosis of ACS [36]; however, there is
dynamic instability [36]. Caution should be taken
some evidence that a normal myoglobin concentra-
in automatically attributing chest pain that is repro-
tion 2 hours after presentation can exclude AMI
ducible on examination to musculoskeletal causes,
[38,42].
because 11% of cases of partially or fully reproduc-
ible chest pain may be attributable to ACS [37]. Troponin cTnI and cTnT are specific for myocar-
Any degree of pulmonary rales on examination dial damage and have supplanted CK-MB as the
is associated with ACS; however, an S3 gallop on preferred biomarker for myocardial ischemia [36].
cardiac auscultation is nonspecific [37]. These biomarkers are not found in the blood of
Pope and colleagues [37] found that patients who healthy individuals [36]. As with the other cardiac
had a final diagnosis of ACS were more likely to biomarkers, the sensitivity of cTnI and cTnT in-
have a lower pulse rate and higher blood pressure creases with serial measurements and with duration
than patients who had other diagnoses; this prob- of symptoms [38]. Elevated levels of cTnI and cTnT
ably is associated with adrenergic excess or lower are associated with increased mortality, even when
compliance of an ischemic left ventricle. The clini- the ECG is inconclusive for ACS and CK-MB con-
cian would need to know the patient’s baseline centrations are normal [36,42].
vital signs; usually, that information is not availa-
ble in the emergency department setting, which Electrocardiography
limits the usefulness of this observation [38]. The Electrocardiography is a safe, inexpensive, and
probability of AMI is increased if the patient is readily available bedside test that represents the
diaphoretic and is decreased if the respiratory rate standard of care for patients who have expected
is normal [37]. ACS. When possible, the ECG should be obtained
while the patient is symptomatic [36]. Although the
Laboratory studies ECG is highly sensitive for AMI, it is neither highly
Because myocytes lose their membrane integrity sensitive nor specific for ACS in general [38]. Pope
in response to ischemia, they release molecules and colleagues [37] found that up to 20% of pa-
into the peripheral circulation [36]. These mole- tients who had AMI and 37% of patients who had
cules, known as cardiac biomarkers, are useful in the diagnosis of unstable angina had a normal ECG
the diagnosis of AMI. The biomarkers that can be at presentation. The ECG should be interpreted
detected do not aid in the diagnosis of unstable with consideration of the patient’s presentation.
angina, which accounts for roughly half of all cases Thus, in a patient with a clinical picture that is
of ACS [38,40,41]. The cardiac biomarkers that are consistent with ACS and a normal ECG, the proba-
in widespread use are creatinine kinase (CK), cre- bility of ischemia is not reduced substantially [38].
atinine kinase MB fraction (CK-MB), myoglobin, ST-segment and T-wave abnormalities are the
cardiac troponin I (cTnI), and cardiac troponin quintessential electrocardiographic abnormalities
T (cTnT). in the diagnosis of ACS [37,38]. ST-segment eleva-
tion indicates transmural ischemia [36], whereas
Creatinine kinase and creatinine kinase MB ST-segment depression indicates subendocardial
fraction CK and CK-MB are nonspecific biomark- ischemia [38]. Inverted T waves indicate acute is-
ers that can be found in any case of muscle damage chemia [38]. Q waves are diagnostic of infarction
[36]. Until recently, CK-MB had been the princi- but could represent previous infarction [38]. Ob-
pal serum marker of cardiac myocyte damage [36]. taining an old ECG can aid in determining if any
The sensitivity of serum CK and CK-MB concen- abnormalities have developed acutely.
trations for detection of ischemia increases with the
duration of the patient’s symptoms [38]. Serial Radiology
measurements of both biomarkers increase sensi- Chest radiography A chest film usually is ob-
tivity and specificity when performed over 4 to tained during the initial assessment of the patient
9 hours [38]. Serial CK-MB has a sensitivity of who has ACS. This imaging study is used to search
Chest Pain: A Clinical Assessment 171
for other causes of the patient’s symptoms and to patients [48]. Patients who have significant PE may
assess for contraindications to heparin therapy remain asymptomatic if the obstruction of pul-
(eg, aortic dissection). The presence of pulmonary monary circulation is less than 50% [50].
edema, which could indicate acute heart failure, The Prospective Investigation of Pulmonary
also is evaluated using plain chest radiography. Embolism Diagnosis (PIOPED) study found that
in patients who were diagnosed with PE, one or
Echocardiography For the patient who has a low
more risk factors for PE were likely to be present
risk for ACS, resting echocardiography has a high
[47]. Risk factors for venous thromboembolism are
sensitivity (93%), although only a moderate speci-
listed in Box 2.
ficity (66%), for diagnosing AMI [43]. Echocardio-
The most common symptom of acute PE is un-
graphy does not distinguish between acute and
explained dyspnea of acute onset [48,51]. Dyspnea
chronic abnormalities and requires skilled tech-
is present in more than 70% of patients who are
nicians and interpreters, which often limits its use
diagnosed with PE [44]. Palpitations, cough, anxi-
in the acute setting [40]. Echocardiography is useful
ety, lightheadedness, abdominal pain, back pain,
in providing information about the patient’s hemo-
atrial fibrillation, and hiccoughs are nonspecific
dynamic status and may help to identify other
symptoms [48,51]. Syncope occurs in 8% to 13%
causes of disease, such as PE and pericarditis [40].
of patients who have PE [52].
Nuclear imaging Thallium-201 (201Tl) and tech- The presentation of the patient who has PE
netium-99m sestamibi (99mTc-sestamibi) are radio- depends on the degree of obstruction of the pul-
nucleotides used commonly in nuclear cardiac monary circulation, the speed of accumulation of
imaging. Noninvasive tests based on those isotopes the clot burden, and the patient’s underlying health
detect ischemic or infarcted myocardium. Both im- [50]. Three clinical syndromes have been described
aging modalities can detect perfusion abnormal- in the patient who has PE: pulmonary infarction,
ities for several hours after the last symptomatic isolated dyspnea, and circulatory collapse [48,53].
episode of chest pain [36]. Abnormal results of Signs and symptoms of PE vary according to the
myocardial perfusion imaging studies done with clinical syndrome that is present. For patients who
the patient at rest indicate risk for AMI and death have pulmonary infarction, pleuritic chest pain and
and the need for revascularization, whereas normal hemoptysis may predominate [53]. Patients
images at rest indicate the patient has a low risk for who have underlying cardiovascular disease, such
cardiac complications [36]. 201Tl images must be as elderly patients, are more likely to have pulmo-
taken within 15 to 20 minutes of injection, which nary infarction [48].
limits the usefulness of this modality in the acute
setting [40]. Imaging with 99mTc sestamibi is advan-
tageous because serial imaging can be done, and
left ventricular wall motion abnormalities can be Box 2: Risk factors for pulmonary embolism
evaluated using gated single photon emission CT
Inherited hematologic risk factors
(SPECT) imaging [36]. Nuclear cardiac imaging is
most useful in patients who have low to moderate • Antithrombin III deficiency
risk for ACS and no acute ECG changes [43]. • Factor V Leiden mutation
• Proteins C and S deficiency
Pulmonary embolism • Lupus anticoagulant
PE must be considered in every patient who has • Abnormalities in fibrinolysis
chest pain and dyspnea. PE is the third most com- Acquired risk factors
mon cause of cardiovascular death among Ameri-
cans, accounting for 50,000 to 100,000 deaths per
• Advanced age
• Smoking
year [44,45]. Only 30% of PE are diagnosed before • Immobilization
death [46]. Alternatively, less than 35% of pa- • Surgery
tients who are suspected of having a PE actually • Malignancy
have one [45,47–49]. PE is a challenging diagnosis • Trauma
to reach, it often is missed, and it often is sought • Oral contraceptives/hormone replacement
but not found. • Pregnancy
• Central venous catheters
History and physical examination • Obesity
The history and physical examination are notori- • Myocardial infarction
ously insensitive for PE. The classic presentation of • Congestive heart failure
PE is chest pain, dyspnea, and hemoptysis; how- Data from Refs. [44,48,53].
ever, this triad is present in less than 20% of
172 Butler & Swencki
For patients who have isolated dyspnea, the de- mal PaCO2, and normal P(Aa)O2 gradients in the
gree of dyspnea varies with the degree of pulmo- face of angiographically proven PE [55].
nary vascular infarction [48]. For patients who have D-dimer, a breakdown product of fibrin, is found
no underlying cardiovascular disease, the extent of in the blood when plasmin acts on a fibrin clot. As
embolism correlates with the degree of arterial a marker of clot lysis, D-dimer is found in any
hypoxemia [50]. Patients who have circulatory col- condition in which there is formation or dissolu-
lapse may present following syncope, be hemody- tion of clot. Thus, D-dimer can be found in ele-
namically unstable, or present in cardiac arrest [53]. vated levels in association with PE, trauma, cancer,
No specific or sensitive physical examination disseminated intravascular coagulation, myocardial
finding is indicative of PE [53]. The most common infarction, sepsis, and preeclampsia and following
signs of PE on examination are tachypnea and surgery. Therefore, D-dimer is more useful in
tachycardia [51]; however, normal vital signs excluding PE than in diagnosing it [53,55,56,58].
should not discourage the physician from searching Wells and colleagues [56] concluded that in a
for a PE [53]. Fever, wheezing, rales, pleural rub, patient with a low clinical probability of PE using
a loud pulmonic component of the second heart the Well’s clinical scoring system and a negative
sound, right ventricular lift, right-sided fourth heart D-dimer assay, PE can be ruled out safely without
sound, cyanosis, and evidence of phlebitis may be any imaging study [55].
present [45,51,53].
ECG
Clinical scoring systems Most patients who have PE have some abnormality
Clinical scoring systems attempt to help the clini- on ECG, but ECG abnormalities in patients who
cian estimate the probability of PE. The best known have PE are nonspecific [51]. The ECG is most
of the clinical scoring systems is Well’s criteria for helpful to exclude other causes of the patient’s
prediction of PE [Table 1]. This scoring system symptoms, such as myocardial ischemia or pericar-
combines the assessment of risk factors, presenting ditis. The characteristic ECG abnormality of PE is
signs and symptoms, as well as the clinician’s sus- the S1Q3T3 pattern; however, this is found in less
picion of an alternate diagnosis [54]. This scoring than 20% of ECGs from patients who have proven
system is vulnerable because it relies heavily on the PE [53]. T-wave inversion in the precordial leads
subjective judgment of the clinician as to the pres- is the most common electrocardiographic finding
ence of alternate diagnoses [53]. and is present in 68% of patients who have PE [48].
Tachycardia and incomplete right bundle branch
Laboratory studies block also have been found more often in patients
Although arterial blood gas analysis (ABG) is a who have PE than in patients who have other
widely available and rapid laboratory study, it diagnoses [45].
lacks the sensitivity to diagnose or exclude PE
[53,55]. Patients who do not have underlying car- Imaging
diopulmonary disease may have normal PaO2, nor- Chest radiography Like electrocardiography, chest
radiography often is abnormal, but nonspecific,
and may elucidate other diagnoses. The PIOPED
Table 1: Wells' criteria for assigning pretest study found that the most sensitive radiographic
probability for pulmonary embolism finding for PE is atelectasis or parenchymal ab-
Criteria Points normality, with a sensitivity of 68% [47]. Other
common abnormalities that are found on chest
Clinical symptoms or signs of DVT 3
PE more likely than other diagnosis 3 radiography include pleural effusion, pulmonary
Heart rate >100 bpm 1.5 infiltrates, mild elevation of the hemidiaphragm,
Immobilization or surgery within 1.5 enlargement of the pulmonary artery, and cardio-
last 4 wk megaly [51,53]. It is important not to exclude the
History of DVT or PE 1.5 diagnosis of PE based on radiographic evidence of
Hemoptysis 1 pneumonia or congestive heart failure, because
Malignancy 1 these entities may coexist with PE [48]. The classic
signs of relative oligemia (Westermark’s sign) and
Clinical probability of PE Points wedge-shaped pulmonary opacity (Hampton’s
hump) are rare [53].
Low <2
Moderate 2–5 Ventilation–perfusion scintigraphy Historically,
High >6
ventilation-perfusion (V/Q) lung scanning has
Data from Refs. [56–58]. been the initial imaging modality of choice in
Chest Pain: A Clinical Assessment 173
patients suspected of having PE. The results of the and abnormal septal wall motion [53]. Transesopha-
V/Q scan are interpreted in association with the geal echocardiography (TEE) is more invasive—
patient’s assigned pretest probability [48,53]. A usually requiring sedation—but is more sensitive
high-probability V/Q scan in a patient who has a than TTE for detection of these hemodynamic ab-
high pretest probability has an 85% to 90% posi- normalities [51,53].
tive predictive value of PE; a normal V/Q scan in a
Ultrasound Lower extremity ultrasound imaging
patient who has a low pretest probability essen-
for the detection of DVT has the greatest usefulness
tially excludes the diagnosis of PE [47,59]. Most
in the patient who has signs and symptoms of DVT
V/Q scans fall into the category of nondiagnostic,
and PE [45]. This test should not be used as an
however, which severely limits the usefulness of
initial imaging modality for the patient who has
this imaging modality [45]. Patients who have
suspected acute PE [45], but it may be useful as an
underlying lung disease also have abnormal base-
adjunct test to detect the source of PE.
line studies [48].
Multidetector CT angiography Multidetector CT Pericarditis
angiography (MDCT-A) is becoming the initial Pericarditis is inflammation of the pericardium, the
study of choice in the acute setting for the diagnosis fibrous sac surrounding the heart and great vessels
of PE, primarily because of its widespread availabil- [61,62]. The many causes of pericarditis include
ity, speed, and noninvasive nature. In comparison collagen vascular disease, renal insufficiency, neo-
with V/Q scanning, CT is more accurate [45] and plasm, viral infections, tuberculosis, and bacterial
is more likely to show another cause of the patient’s infections [63]. In many cases, the exact etiology
symptoms if PE is not present. There has been some remains unknown [63,64]. The diagnosis of peri-
question as to the sensitivity of CT for PE. Pooled carditis is suspected in the patient who has chest
data show a wide range of sensitivities (53–100%) pain, pericardial rub on physical examination, and
and specificities (81–100%) [60]; however, for cen- characteristic ECG changes [65].
tral PE, the sensitivity of CT increases to 94% [45].
Subsegmental emboli and horizontal vessels are
History and physical examination
not well visualized on CT [48]. Other drawbacks
By history alone, pericarditis may be difficult to
of CT imaging include the use of nephrotoxic con-
differentiate from myocardial ischemia, because
trast and radiation exposure; in addition, the study
the patient may complain of retrosternal chest
requires a cooperative patient, because motion arti-
pain with a radiation pattern similar to that of
fact limits the quality of the images [45].
myocardial ischemia [65,66]. Classically, pericardi-
Magnetic resonance angiography Magnetic reso- tis presents with a retrosternal location of pain, but
nance angiography (MRA) may be used to visualize the patient may complain of pain anywhere in the
PE and lower extremity deep vein thrombosis chest [66]. The pain often is described as sharp or
(DVT) and offers the advantages of safer contrast stabbing [61]. A pleuritic component of the pain,
material, noninvasive nature, and no ionizing ra- including increasing pain with inspiration, an in-
diation [48,53]. MRA is limited in its use by ex- crease in pain with supine position, and some relief
pense and availability. In addition, MR imaging is with upright posture or sitting forward, is described
time consuming and allows only limited access to often [61,65]. The pain may radiate to the neck,
patients who become unstable [48]. arms, or left shoulder [61]. Pain that radiates to
either trapezius muscle ridge is likely to be pericar-
Pulmonary angiography Pulmonary angiography
ditis secondary to phrenic nerve innervation of the
is considered the gold standard for the diagnosis
anterior pericardium and both trapezius ridges
of PE [47]. Often, this procedure is not readily
[61,65,66]. Onset of pain is sudden and progressive
available; requires nephrotoxic contrast; and is in-
over hours to days [61,65].
vasive, time-consuming, and expensive. In addi-
Fever or other features of a nonspecific prodrome
tion, the patient must be transported away from
may precede pericarditis of infectious etiology
the emergency department, and the images rarely
[65,67]. A medical history of renal failure, known
elucidate an alternate diagnosis [53].
neoplasm, collagen vascular disease, or thyroid dis-
Echocardiography Transthoracic echocardiography order may aid in the diagnosis, because these are
(TTE) is noninvasive and can be performed at the common causes of pericarditis [68]. Temperature
bedside. Findings on echocardiography that suggest higher than 38°C is rare, but when present, may
PE include right-sided thrombus; dilation of the indicate purulent pericarditis [61].
right ventricle, pulmonary artery, or inferior vena A pericardial friction rub is pathognomic for
cava; decreased right ventricular function; loss of pericarditis and is 100% specific for the disease
right ventricular contractility; tricuspid regurgitation; [65,66]. The pericardial friction rub may come
174 Butler & Swencki
and go with time; thus, the patient should be Stage II is normalization of the ST and PR seg-
examined repeatedly [61,66]. The rub is best ments, whereas stage III is characterized by wide-
heard at the left lower sternal border with the spread T-wave inversions. The ECG normalizes
patient leaning forward at the end of expiration again in stage IV. With the exception of purulent
[5,6,61]. Generally, the rub is described as rasping, pericarditis, if the patient is treated promptly,
creaking, harsh, or high pitched. Classically, it is stage I may be the only electrocardiographic abnor-
triphasic, but it may be biphasic or monophasic mality seen [66]. The diffuse ST segment elevation
[62,65]. The stereotypic triphasic rub that corre- of pericarditis can be differentiated from myocar-
sponds to the motion of the heart during ventricu- dial ischemia by the absence of reciprocal ST de-
lar systole, diastolic ventricular filling, and atrial pression [7] and by the concave shape of the
contraction is present in only half of patients [69]. ST-segment elevations [61]. The presence of cardiac
The presence of muffled heart sounds, tachycar- tamponade is characterized by low-voltage ECG
dia, distended neck veins, hypotension, and pulsus with electrical alternans [77].
paradoxus suggests cardiac tamponade. The patient
may be in acute respiratory distress, but the lungs Radiography
generally will be clear [67]. Radiologic studies may exclude other causes of
chest pain. Chest radiography for pericarditis is
Laboratory studies aimed primarily at evaluation of the mediastinum
Laboratory studies are obtained to exclude other and lungs for possible causes of the inflammation
causes of chest pain and to elucidate the possible [61]. Cardiomegaly may be seen when an effusion
cause of the pericarditis. Markers of inflammation, of more than 250 mL has accumulated [61,65,
such as leukocytosis, elevated C-reactive protein, 67,69].
and elevated erythrocyte sedimentation rate, usu-
ally are found in patients who have acute pericar- CT and MR imaging
ditis [61]. Plasma electrolytes should be measured, CT and MR imaging may be used to image the
and renal function should be evaluated [65]. pericardium and pericardial space but are obtained
The patient’s clinical picture should guide addi- most commonly to exclude other causes of chest
tional testing, which might include blood cultures, pain or shortness of breath. CT and MR imaging
tuberculin skin test, antinuclear antibodies, rheu- evidence of thickened pericardium, enhancement
matoid factor, thyroid function tests, viral throat of the pericardium that indicates inflammation,
swabs, and specific viral and bacterial serologies and visualization of pericardial effusion support
[63–65,68]. the diagnosis of pericarditis [79,80].
Pericardiocentesis should be considered in pa-
tients who have tamponade or suspected neoplastic
Echocardiography
or purulent pericarditis [61,65]. Routine pericar-
Often, TTE is performed in patients who have sus-
diocentesis for purely diagnostic purposes is not
pected pericarditis. The presence of an effusion
recommended [70].
will help to confirm the diagnosis [61]. Evidence
Cardiac biomarker levels may be abnormal in
of tamponade on echocardiogram indicates the
patients who have pericarditis. Specifically, cTnI is
need for pericardiocentesis.
elevated in more than 30% of patients who have
acute pericarditis [71–73]. Men and younger pa-
tients are more likely to have elevated cTnI levels Thoracic aortic dissection
[71]. Elevation of cTnI is seen only in patients who Thoracic aortic dissection (TAD) is the most com-
have elevated ST segment on ECG and indicates mon aortic emergency that requires immediate sur-
myocardial cell damage [72]; however, cTnI levels gery [81]. A dissection occurs when there is a tear
do not indicate poor prognosis [71,72]. Serum CK in the intimal layer of the vessel wall. Blood passes
and CK-MB levels also may be elevated [61]. through the tear, separates the intima from the
vessel media or adventitia, and results in a false
Electrocardiography channel. Shear forces lead to dissection propaga-
Diffuse elevation of the ST segments in the pre- tion as blood continues to flow through this false
cordial and limb leads that is associated with PR channel [82].
segment depression is a classic electrocardiographic TAD can be difficult to diagnose. In the patient
indication of acute pericarditis [74]. Historically, who presents to the emergency department and
electrocardiographic abnormalities of acute pericar- has acute chest or back pain, ACSs are 80 times
ditis have been said to evolve over time, with more common than are aortic dissections [83].
four distinct stages described [75–78]. In stage I, Given that TADs occur most commonly in men
ST elevation is diffuse, with PR segment depression. who are aged 50 to 70 years and have a history of
Chest Pain: A Clinical Assessment 175
hypertension, it is not surprising that myocardial but the means of assessing these markers is not
ischemia is the most common misdiagnosis [84]. widely available in most clinical settings [91,92].
Electrocardiography
History and physical examination In the patient who has TAD, the ECG may be
The acute onset of severe pain, which is maximal normal or show left ventricular hypertrophy from
at symptom onset, is the most common initial long-standing hypertension [87]. Changes sugges-
symptom [85]. Usually, the pain is in the midline, tive of myocardial ischemia related to coronary ar-
may be present in the back, and rarely radiates [85]. tery involvement in the dissection or occlusion of
A tearing or ripping quality of pain is classic and the artery may be present in up to one third of
highly specific for TAD [86]; however, the Interna- patients [87,93]. Two thirds of patients who have
tional Registry of Acute Aortic Dissection (IRAD) TAD have nonspecific ECG abnormalities of non-
found that most often, the pain was described as specific ST-segment or T-wave changes [85].
‘‘sharp’’ [81]. Migratory pain has been considered
classic for TAD, with pain corresponding to the Imaging
propagation of the dissection, but was found in Chest radiography Chest radiography is not spe-
only 14% of patients in the IRAD [81]. Hyper- cific for aortic dissection but is useful in combina-
tension is the most common predisposing risk tion with the history and physical examination. In
factor for TAD [81]. Inherited disorders, such as approximately 50% of cases of dissection, the classic
Marfan’s syndrome and Ehler-Danlos’ syndrome, radiographic sign of widened mediastinal shadow
associated with abnormal connective tissue struc- is seen [85]. Some type of chest film abnormality
ture have high rates of TAD [87,88]. Among [Box 3] is present in 90% of patients [86].
women who are younger than 40 years who expe-
CT In the IRAD study, CT often was the initial
rience TAD, half are pregnant [85,87]. Cocaine use
imaging test for patients who have suspected aortic
has been associated with TAD [89]. A history of
dissection, likely secondary to its widespread avail-
syncope, with or without chest pain, was docu-
ability and noninvasive nature [81]. CT has a sen-
mented in 12% of patients who had TAD [81].
sitivity of 93.8% and a specificity of 87% for TAD
Pulse deficits or blood pressure differentials are
[96]. A positive contrast CT for TAD shows the
independent predictive variables for TAD [83];
raised intimal flap of the dissection between the
however, pulse deficits were documented in only
true and false lumens of the aorta [96]. Throm-
20% of patients in the IRAD [81]. Shear injury of
bus also may be visualized within the false lumen
the left carotid artery or compression of the aortic
[96]. If a dissection is not present, CT images may
branches that supply the spinal cord may pro-
identify another cause of the patient’s symptoms.
duce focal neurologic deficits [87,88]. When these
Drawbacks of CT imaging include relative difficulty
deficits are present on examination, there is an
in identifying the origin of the intimal tear, inabil-
increased likelihood of TAD [86]. A pulsatile ster-
ity to assess involvement of the aortic branch ves-
noclavicular joint is rare but may indicate dissec-
tion [87].
When the dissection is proximal, an aortic regur- Box 3: Common abnormalities found on chest
gitation murmur may be heard. The IRAD study radiography for aortic dissection
found this murmur in 44% of patients who had
proximal TAD [81]. Cardiac tamponade with re- • Widened mediastinal shadow
sultant physical examination findings of muffled • Altered configuration of the aorta
heart sounds, elevated jugular venous distention, • Localized hump on the aortic arch
and narrow pulse pressure may be found when • Widening of the distal aortic knob past the
origin of the left subclavian artery
blood fills the pericardium [82].
• Aortic wall thickness indicated by the
width of the aortic shadow beyond
Laboratory studies intimal calcification
Generally, routine laboratory studies are not help- • Displacement of the calcification in the
ful in the diagnosis of TAD [87]. Often, laboratory aortic knob
studies are obtained to assess other causes of chest • Double aortic shadow
pain. Recent studies suggest that D-dimer con- • Disparity in the sizes of the ascending and
centration may be useful as a diagnostic tool for descending aorta
the diagnosis of TAD [90]. Serum biomarkers of • Presence of a pleural effusion, most
commonly on the left
smooth muscle myosin heavy chain and soluble
elastin fragments have been found in higher con- Data from Refs. [85,94,95].
centrations in patients who had aortic dissection,
176 Butler & Swencki
sels, and lack of information about aortic valve Despite advances in medical care, cardiopulmonary
regurgitation [85,96]. emergencies remain a major cause of morbidity and
mortality in the United States. Rapid bedside radio-
MR imaging MR imaging has a reported sensitivity
graphic detection of intrathoracic disorders is critical
and specificity of 98% for TAD [96]. This imaging
in clinical decision making related to these poten-
modality provides quality images of the entire
tially life-threatening emergencies.
aorta, showing extent of the dissection, site of the
tear, involvement of branch vessels, and involve-
ment of the aortic valve [85,97]. However, MR References
imaging is expensive, time consuming, and not
[1] Tamura M, Ohta Y, Sato H. Thoracoscopic ap-
widely available and limits access to the potentially pearance of bilateral spontaneous pneumotho-
unstable patient [87]. Because of this, MR imaging rax. Chest 2003;124(6):2368–71.
rarely is obtained in the emergent setting. [2] Steele R, Irvin CB. Central line mechanical com-
Echocardiography The sensitivity and specificity plication rate in emergency medicine patients.
Acad Emerg Med 2001;8(2):204–7.
of TTE for the detection of TAD vary widely, de-
[3] National Healthcare Quality Report. Agency for
pending on the location of the dissection [85]. TTE Healthcare Research and Quality, Department of
does not visualize the aortic arch well and is vir- Health and Human Services, 2003.
tually useless for the descending aorta [97]; how- [4] Chang AK, Barton ED. Pneumothorax, iatrogenic,
ever, TTE is useful for detection of tamponade or spontaneous and pneumomediastinum. Avail-
aortic insufficiency, which are complications of able at http://www.emedicine.com/EMERG/
proximal dissections [87]. topic469.htm. Accessed November 4, 2005.
TEE is much better than TTE in the detection [5] Ferraro P, Beauchamp G, Lord F, et al. Sponta-
of TAD. TEE has a reported sensitivity of up to neous primary and secondary pneumothorax:
98%, with a specificity of 77% [96]. A positive a 10-year study of management alternatives.
Can J Surg 1994;37(3):197–202.
TEE may show a double-lumen aorta separated by
[6] Cottin V, Streichenberger N, Gamondes JP, et al.
the dissection membrane, which moves with the Respiratory bronchiolitis in smokers with spon-
differential flow through the lumens [87]. TEE also taneous pneumothorax. Eur Respir J 1998;12(3):
can identify the site of dissection, sense abnormal 702–4.
flow, visualize a thrombus, assess for involvement [7] Yellin A, Shiner RJ, Lieberman Y. Familial multi-
of aortic valve and aortic branch vessels, and detect ple bilateral pneumothorax associated with
pericardial effusion [85]. The advantages of TEE are Marfan syndrome. Chest 1991;100(2):577–8.
that it can be performed at the patient’s bedside in a [8] Alifano M, Roth T, Broet SC, et al. Catamenial
swift manner [85]. In many centers, TEE represents pneumothorax: a prospective study. Chest 2003;
the noninvasive study of choice for the patient 124(3):781–2.
[9] Seow A, Kazerooni EA, Pernicano PG, et al.
suspected of having TAD [85]. TEE requires esopha-
Comparison of upright inspiratory and expira-
geal intubation, which often necessitates sedation, tory chest radiographs for detecting pneumo-
and may not be available in some centers, espe- thoraces. AJR Am J Roentgenol 1996;166(2):
cially in the evening and on weekends. 313–6.
[10] Carr JJ, Reed JC, Choplin RH, et al. Plain and
Aortography Traditionally, aortography has been
computed radiography for detecting experimen-
considered the gold standard for the diagnosis of tally induced pneumothorax in cadavers: impli-
TAD, but it has been replaced by less invasive and cations for detection in patients. Radiology 1992;
more readily available radiologic studies. Findings 183(1):193–9.
on aortography that are indicative of TAD include [11] Blaivas M, Lyon M, Duggal S. A prospective
distortion of contrast flow, flow reversal, flow sta- comparison of supine chest radiography and
sis, failure of major vessels to fill with contrast, and bedside ultrasound for the diagnosis of traumatic
aortic valve insufficiency [85]. Aortography may pneumothorax. Acad Emerg Med 2005;12(9):
underestimate the size of the dissection if a throm- 844–9.
bus is present [97]. The procedure is invasive and [12] Mandell LA. Epidemiology and etiology of
community-acquired pneumonia. Infect Dis Clin
requires mobilization of an angiographic team.
North Am 2004;18:761–76.
Aortography rarely is obtained for the acute diag- [13] Niederman MS, McCombs JS, Unger AN, et al.
nosis of TAD [81]. The cost of treating community-acquired pneu-
monia. Clin Ther 1998;20(4):820–37.
[14] Hoyert DL, Heron M, Murphy SL, et al. Deaths:
Summary final data for 2003. Hyattsville (MD): US
Department of Health and Human Services,
The use of radiographic imaging remains vital in the Centers for Disease Control and Prevention,
assessment of patients who present with chest pain. National Center for Health Statistics; January
Chest Pain: A Clinical Assessment 177
19, 2006. Available at: http://www.cdc.gov/nchs. lung disease in the immunocompromised pa-
Accessed on January 24, 2006. tient. Radiol Clin North Am 1994;32:731–44.
[15] Vincent JL, Bihari DJ, Suter PM, et al. The [31] Boiselle PM, Tocino I, Hooley RJ, et al. Chest
prevalence of nosocomial infection in intensive radiograph interpretation pf Pneumocystis carinii
care units in Europe. Results of the European pneumonia, bacterial pneumonia, and pulmo-
Prevalence of Infection in Intensive Care (EPIC) nary tuberculosis in HIV-positive patients: ac-
Study. EPIC International Advisory Committee. curacy, distinguishing features, and mimics. J
JAMA 1995;274(8):639–44. Thorac Imaging 1997;12:47–53.
[16] Franquet T. Imaging of pneumonia: trends and [32] Janzen DL, Padley SPG, Adler BD, et al. Acute
algorithms. Eur Resp J 2001;18:196–208. pulmonary complications in immunocompro-
[17] Courtoy I, Lande AE, Turner RB. Accuracy of mised non-AIDS patients: comparison of diag-
radiographic differentiation of bacterial from nostic accuracy of CT and chest radiography.
nonbacterial pneumonia. Clin Pediatr (Phila) Clin Radiol 1993;47:159–65.
1989;28(6):261–4. [33] Chastre J, Trouillet JL, Vuagnat A, et al. Noso-
[18] Ponka A, Sarna S. Differential diagnosis of vi- comial pneumonia in patients with acute respi-
ral, mycoplasmal and bacteraemic pneumococcal ratory distress syndrome. Am J Respir Crit Care
pneumonias on admission to hospital. Eur J Re- Med 1998;157:1165–72.
spir Dis 1983;64(5):360–8. [34] Seidenfeld JJ, Pohl DF, Bell RD, et al. Incidence,
[19] Van Mieghem IM, De Wever WF, Verschakelen site and outcome of infections in patients with
JA. Lung infection in radiology: a summary of adult respiratory distress syndrome. Am Rev
frequently depicted signs. JBR-BTR 2005;88(2): Respir Dis 1986;134:12–6.
66–71. [35] Niederman MS, Fein AM. Sepsis syndrome, the
[20] Murray JF, Mills J. Pulmonary infectious compli- adult respiratory distress syndrome and noso-
cations of human immunodeficiency virus infec- comial pneumonia: a common clinical sequence.
tion. Am Rev Respir Dis 1990;141:1356–72. Clin Chest Med 1990;11:633–56.
[21] Lyon R, Haque AK, Asmuth DM, et al. Changing [36] Kamineni R, Alpert JS. Acute coronary syn-
patterns of infections in patients with AIDS: dromes: initial evaluation and risk stratification.
a study of 279 autopsies of prison inmates and Prog Cardiovasc Dis 2004;46:379–92.
nonincarcerated patients at a university hospital [37] Pope JH, Ruthazer R, Beshansky JR, et al. Clinical
in eastern Texas, 1984–1993. Clin Infect Dis 1996; features of emergency department patients pre-
23:241–7. senting with symptoms suggestive of acute
[22] Guckel C, Benz-Bohm G, Widemann B. Myco- cardiac ischemia; a multicenter study. J Thromb
plasmal pneumonias in childhood. Roentgen Thrombolysis 1998;6:63–74.
features, differential diagnosis and review of lit- [38] Pope JH, Selker HP. Diagnosis of acute cardiac
erature. Pediatr Radiol 1989;19(8):499–503. ischemia. Emerg Med Clin North Am 2003;21:
[23] Goodman PC. Pneumocystis carinii pneumonia. 27–59.
J Thorac Imaging 1991;6(4):16–21. [39] Coronado BE, Pope JH, Griffith JL, et al. Clinical
[24] Goodman PC. The chest film in AIDS. In: Sande features, triage, and outcome of patients present-
MA, Volberding P, editors. The medical manage- ing to the ED with suspected acute coronary syn-
ment of AIDS. 4th edition. Philadelphia: Saun- dromes but without pain: a multicenter study.
ders; 1995. p. 592–613. Am J Emerg Med 2004;22:568–74.
[25] Moe AA, Hardy WD. Pneumocystis carinii infec- [40] Selker HP, Zalenski RJ, Antman EM, et al. An
tion in the HIV-seropositive patient. Infect Dis evaluation of technologies for identifying acute
Clin North Am 1994;8:331–64. cardiac ischemia in the emergency department:
[26] Wasser LS, Brown E, Talavera W. Miliary PCP in executive summary of a national heart attack
AIDS. Chest 1989;96(3):693–5. alert program working group report. Ann Emerg
[27] Chaffey MH, Klein JS, Gamsu G, et al. Radio- Med 1997;29:1–12.
graphic distribution of Pneumocystis carinii [41] Balk EM, Ioannidis JPA, Salem D, et al. Accuracy
pneumonia in patients with AIDS treated with of biomarkers to diagnose acute cardiac ischemia
prophylactic inhaled pentamidine. Radiology in the emergency department: a meta-analysis.
1990;175(3):715–9. Ann Emerg Med 2001;37:478–94.
[28] Baughman RP, Dohn MN, Shipley R, et al. [42] Newby LK. Markers of cardiac ischemia, injury,
Increased Pneumocystis carinii recovery from the and inflammation. Prog Cardiovasc Dis 2004;
upper lobes in Pneumocystis pneumonia: the ef- 46:404–16.
fect of aerosol pentamidine prophylaxis. Chest [43] Ioannidis JPA, Salem D, Chew PW, et al. Accu-
1993;103(2):426–32. racy of imaging technologies in the diagnosis of
[29] Hanson DL, Chu SY, Farizo KM, et al. Distribu- acute cardiac ischemia in the emergency depart-
tion of CD4 lymphocytes at diagnosis of acquired ment: a meta-analysis. Ann Emerg Med 2001;37:
immunodeficiency syndrome-defining and other 471–7.
human immunodeficiency virus-related illnesses. [44] Lee LC, Shah K. Clinical manifestation of pul-
Arch Intern Med 1995;155:1537–42. monary embolism. Emerg Med Clin North Am
[30] Primack SL, Müller NL. HRCT in acute diffuse 2001;19:925–42.
178 Butler & Swencki
[45] Rodger M, Wells PS. Diagnosis of pulmonary tion and treatment of infectious and other
embolism. Thromb Res 2001;103:V225–38. causes. Postgrad Med 2004;115(3):67–70, 73–5.
[46] Morgenthaler TI, Ryu JH. Clinical characteristics [63] Levy PY, Corey R, Berger P, et al. Etiologic diag-
of fatal pulmonary embolism in a referral hos- nosis of 204 pericardial effusions. Medicine
pital. Mayo Clin Proc 1995;70:417–24. 2003;82:385–91.
[47] Investigators PIOPED. Value of the ventilation/ [64] Zayas R, Anguita M, Torres F, et al. Incidence
perfusion scan in acute pulmonary embolism. of specific etiology and role of methods for spe-
Results of the Prospective Investigation of Pul- cific etiologic diagnosis of primary acute peri-
monary Embolism Diagnosis study (PIOPED). carditis. Am J Cardiol 1995;75:378–82.
JAMA 1990;263:2753–9. [65] Troughton RW, Asher CR, Klein AL. Pericarditis.
[48] Laack TA, Goyal DG. Pulmonary embolism: an Lancet 2004;363:717–27.
unsuspected killer. Emerg Med Clin North Am [66] Spodick DH. Acute pericarditis: current concepts
2004;22:961–83. and practice. JAMA 2003;289:1150–3.
[49] Perrier A, Howarth N, Didier D, et al. Perfor- [67] Goyle KK, Walling AD. Diagnosing pericarditis.
mance of helical computed tomography in Am Fam Phys 2002;66:1695–702.
unselected outpatients with suspected pulmo- [68] Levy PY, Maotti JP, Gauduchon V, et al. Com-
nary embolism. Ann Intern Med 2001;135:88–97. parison of intuitive versus systematic strategies
[50] Riedel M. Acute pulmonary embolism: patho- for aetiological diagnosis of pericardial effusion.
physiology, clinical presentation, and diagnosis. Scand J Infect Dis 2005;37:216–20.
Heart 2001;85:229–40. [69] Spodick DH. Pericardial rub: prospective, multi-
[51] Tapson VF. Acute pulmonary embolism. Cardiol ple observer investigation of pericardial friction
Clin 2004;22:353–65. in 100 patients. Am J Cardiol 1975;35:357–62.
[52] Brilakis ES, Tajik AJ. 82-year-old man with recur- [70] Permanyer-Miralda G. Acute pericardial disease:
rent syncope. Mayo Clin Proc 1999;74:609–12. approach to the aetiologic diagnosis. Heart 2004;
[53] Sadosty AT, Boie ET, Stead LG. Pulmonary embo- 90(3):252–4.
lism. Emerg Med Clin North Am 2003;21:363–84. [71] Imazio M, Brunella D, Cacchi E, et al. Cardiac
[54] Wells PS, Ginsberg JS, Anderson DR, et al. Use troponin I in acute pericarditis. J Am Coll Car-
of a clinical model for safe management of pa- diol 2003;42:2144–8.
tients with suspected pulmonary embolism. Ann [72] Bonnefoy E, Godon G, Kirkorian M, et al. Serum
Intern Med 1998;129:997–1005. cardiac troponin I and ST-segment elevation in
[55] Weiner SG, Burstein JL. Nonspecific tests for patients with acute pericarditis. Eur Heart J 2000;
pulmonary embolism. Emerg Med Clin North 21:832–6.
Am 2001;19:943–55. [73] Brandt RR, Filzmaier K, Hanrath P. Circulating
[56] Wells PS, Anderson DR, Rodger M, et al. cardiac troponin I in acute pericarditis. Am J
Excluding pulmonary embolism at the bedside Cardiol 2001;87:1326–8.
without diagnostic imaging: management of pa- [74] Wang K, Asinger RW, Marriott HJL. ST-Segment
tients with suspected pulmonary embolism pre- elevation in conditions other than acute myo-
senting to the emergency department by using cardial infarction. N Engl J Med 2003;349:
a simple clinical model and D-dimer. Ann Intern 2128–35.
Med 2001;135:98–107. [75] Ginzton LE, Laks MM. The differential diagnosis
[57] Wicki J, Perneger TW, Junod AF, et al. Assessing or acute pericarditis from the normal variant:
clinical probability of pulmonary embolism in new electrocardiographic criteria. Circulation
the emergency ward: a simple score. Arch Intern 1982;65:1004–9.
Med 2001;161:92–7. [76] Spodick DH. Diagnostic electrocardiographic se-
[58] Carrier M, Wells PS, Rodger MA. Excluding pul- quences in acute pericarditis: significance of PR
monary embolism at the bedside with low pre- segment and PR vector changes. Circulation 1973;
test probability and D-dimer: safety and clinical 48:575–80.
utility of 4 methods to assign pre-test probability. [77] Surawicz B, Lasseter KC. Electrocardiogram in
Thromb Res 2005, in press. pericarditis. Am J Cardiol 1970;26:471–4.
[59] Hull RD, Hirsh J, Carter CJ, et al. Pulmonary [78] Bruce MA, Spodick DH. Atypical electrocardio-
angiography, ventilation lung scanning, and gram in acute pericarditis: characteristics and
venography for clinically suspected pulmonary prevalence. J Electrocardiol 1980;13:61–6.
embolism with abnormal perfusion lung scan. [79] Rienmüller R, Gröll R, Lipton MJ. CT and MR
Ann Intern Med 1983;98:891–9. imaging of pericardial disease. Radiol Clin North
[60] Rathbun SW, Rashkob GE, Whitsett TL. Sensitiv- Am 2004;42:587–601.
ity and specificity of helical computed tomogra- [80] Wang ZJ, Reddy GP, Gotway MB, et al. CT and
phy in the diagnosis of pulmonary embolism: MR imaging of pericardial disease. Radiographics
a systematic review. Ann Intern Med 2000; 2003;23:S167–80.
132(3):227–32. [81] Hagan PG, Nienaber CA, Isselbacher EM, et al.
[61] Lange RA, Hillis LD. Acute pericarditis. N Engl J The international registry of acute ortic dissection
Med 2004;351:2195–202. (IRAD): new insights into an old disease. JAMA
[62] Ross AM, Grauer SE, Acute pericarditis: evalua- 2000;283:897–903.
Chest Pain: A Clinical Assessment 179
[82] Klein DG. Thoracic aortic aneurysms. J Cardio- [91] Suzuki T, Katoh H, Watanabe M, et al. Novel
vasc Nurs 2005;20:245–50. biochemical diagnostic method for acute aortic
[83] von Kodolitsch Y, Schwartz AG, Nienaber CA. dissection: results of a prospective study using an
Clinical prediction of acute aortic dissection. immunoassay of smooth muscle myosin heavy
Arch Intern Med 2000;160:2977–82. chain. Circulation 1996;93:1244–9.
[84] Rogers RL, McCormack R. Aortic disasters. Emerg [92] Shinohara T, Suzuki K, Okada M, et al. Soluble
Med Clin North Am 2004;22:887–908. elastin fragments in serum are elevated in acute
[85] Khan IA, Nair CK. Clinical, diagnostic, and man- aortic dissection. Arterioscler Thromb Vasc Biol
agement perspectives of aortic dissection. Chest 2003;23:1839–44.
2002;122:311–28. [93] Armstrong WF, Bach DS, Carey LM, et al. Clinical
[86] Bushnell J, Brown J. Clinical assessment for and echocardiographic findings in patients with
acute thoracic aortic dissection. Ann Emerg Med suspected acute aortic dissection. Am Heart J
2005;46:90–2. 1998;136:1051–60.
[87] Chan K, Varon J, Wenker OC, et al. Acute tho- [94] Earnest F, Muhm JR, Sheedy PF. Roentgeno-
racic aortic dissection: the basics. J Emerg Med graphic findings in thoracic aortic dissection.
1997;15:859–67. Mayo Clin Proc 1979;54:43–50.
[88] Prêtre R, von Segesser LK. Aortic dissection. [95] Jagannath AS, Sos TA, Lockhart SH, et al. Aortic
Lancet 1997;349:1461–4. dissection: a statistical analysis of the usefulness
[89] Perron AD, Gibbs M. Thoracic aortic dissection of plain chest radiographic findings. AJR Am J
secondary to crack cocaine ingestion. Am J Emerg Roentgenol 1986;147:1123–6.
Med 1997;15:507–9. [96] Nienaber CA, von Kodolitsch Y, Nicholas V, et al.
[90] Hazui H, Fukumoto H, Negoro N, et al. Sim- The diagnosis of thoracic aortic dissection by
ple and useful tests for discriminating between noninvasive imaging procedures. N Engl J Med
acute aortic dissection of the ascending aorta and 1993;328:1–9.
acute myocardial infarction in the emergency [97] Nienaber CA, von Kodolitsch Y. Diseases of the
setting. Circ J 2005;69:677–82. aorta. Cardiol Clin 1998;16:295–314.