Pathophysiology and Diagnosis of Deep Venous Thrombosis
Bruce R. Line
Lower-limb deep venous thrombosis (DVT) affects be-
tween 1% to 2% of hospitalized patients. These thrombi
disrupt the vascular integrity of the lower limbs and are
the source of emboli that kill approximately 200,000
patients each year in the United States. The causes of
thrombosis include vessel wall damage, stasis or low
flow, and hypercoagulability. These factors favor clot
formation by disrupting the balance of the opposing
coagulative and fibrinolytic systems. The symptoms
and signs of venous thrombosis are caused by obstruc-
tion to venous outflow, vascular inflammation, or pul-
monary embolization. About 70% of patients referred
for clinically suspected venous thrombosis, however, do
not have the diagnosis confirmed by objective testing.
Among the 30% w h o have venous thrombosis, about
85% have proximal vein thrombosis, and the remainder
have thrombosis confined to the calf. Physicians cannot
rely on signs and symptoms to make the diagnosis of
DVT and must depend on imaging studies to guide
treatment. Patients with proximal vein thrombosis w h o
are inadequately treated have a 47% frequency of recur-
rent venous thromboembolism over 3 months. In con-
trast, clinically detectable recurrence occurs in less than
2% of patients with proximal vein thrombosis if an
adequate anticoagulant response is achieved. Of the
diagnostic procedures for DVT, venography is the only
invasive test of proven value, and ultrasonographic (US)
OWER-LIMB DEEP venous thrombosis (DVT)
affects between 1% to 2% of hospitalized
patients.l The true incidence of DVT in the popu-
lation is unknown because many cases without
symptoms pass undetected. However, many fatal
emboli originate from thrombi in the proximal
lower limb or the pelvis, and each year in the
United States approximately 200,000 patients die
as a result of thromboembolism. 2 It is useful to first
explore the natural history of venous thrombosis
from a molecular and pathophysiologic perspective
to provide the basis for understanding the clinical
and diagnostic aspects of DVT.
NATURAL HISTORY OF VENOUS
THROMBOEMBOLISM
Most deep vein (Fig 1) thrombi develop in the
calf, with common sites of origin in the soleal
sinusoids and behind the venous valves. 3 Sevitt has
suggested that the eddy currents occurring as blood
passes a valve encourage the deposition of throm-
bus within the valve cusp. 4 Vascular compression
accomplished by exercising muscles may wash
away these small deposits. Alternatively, they may
provide a nidus for the generation of thrombin and
90 Seminars in Nuclear Medicine, Vol XXXl, No 2 (April), 2001: pp 90-101
studies are the most commonly used noninvasive mo-
dality. Other procedures are occasionally used to diag-
nose DVT, including impedance plethysmography, com-
puted tomography, and magnetic resonance imaging.
US examinations are noninvasive, they are rapidly ob-
tained, and they can be performed serially. In symptom-
atic patients, venous US is sensitive and specific for
proximal DVT; however, US is insensitive to calf vein
thrombosis and to asymptomatic DVT occurring after
surgery. Patients with symptoms of recurrent DVT also
can present a difficult diagnostic problem. Only about
20% to 30% of these individuals actually have the dis-
ease; the rest have symptoms arising from chronic
venous insufficiency or from any of the causes of lower
extremity pain. After an acute episode, up to 50% of
patients have compression ultrasound abnormalities for
6 months that are indistinguishable from the original
findings of DVT. Hence, there are a significant number of
patients and clinical circumstances in which the diagno-
sis of DVT is difficult, sSmTc-radiolabeled peptides that
target the molecular biology of thrombosis should aid in
the management of the disease, particularly in asymp-
tomatic patients at high risk, in patients with recurrent
symptoms, in patients with active DVT in the calf and/or
pelvis, and in patients with intermediate- or low-proba-
bility lung scans.
Copyright 9 2001 by W,B. Saunders Company
fibrin clots (Fig 2). Coagulation may be initiated
by tissue factor released by vascular wall damage, 5
by cytokine activation of endothelial ceils, or by
activated monocytes responding to vascular in-
j u r y . 6 The initial platelet cluster on the vessel
builds and grows toward the center of the vessel
lumen, alternating layers of fibrin and red cells that
are trapped between layers of platelets. 7 As the
thrombus grows out into the bloodstream, it is bent
in the direction of flow and extends across the
lumen. The flow around the thrombus becomes
turbulent and gradually decreases. When the vein
is completely occluded, there is usually rapid
extension of a jellylike red thrombus up to the
mouth of the next major tributary. If this tributary
becomes occluded, the propagated thrombus con-
From the University of Maryland Medical College, Division
of Nuclear Medicine, University of Maryland Medical Systems,
Baltimore, MD.
Address reprint requests to Bruce R. Line, MD, Department
of Diagnostic Radiology, 22 South Greene St, Baltimore, MD
21201-1595.
Copyright 9 2001 by W.B. Saunders Company
0001-2998/01/3102-0001535.00/0
doi:l O.lO53/snuc.2001.21406
DIAGNOSIS OF DVT
Postedor Antmtor
Fig 1. Anatomy of the deep venous system. The deep
veins of the lower limb consist of paired anterior tibiel,
posterior tibial, and peroneal veins that parallel the course of
the arterial system. These veins join the popliteal vein. Once
the popliteal vein passes through the adductor canal, it
becomes the femoral vein. The deep femoral vein is s branch
of the femoral vein.
tinues to extend proximally and may reach several
feet in length. 8
When a thrombus adheres to the endothelium, it
stimulates a hyperemic inflammatory response that
begins the processes of organization. The clot is
invaded by granulomatous host response cells and
is replaced by fibrous tissue. Where the thrombus
remains loose within the lumen, polymerization
and maturation of the fibrin within the thrombus
cause it to retract. Thrombus retraction and orga-
nization eventually lead to recanalization and en-
dothelialization. This process destroys all the
valves in the affected segment of the vein and is
accompanied by an enlargement of the collateral
venous channels. There is a considerable danger
of embolization until a nonadherent, nonocclusive
thrombus begins to contract. Contraction normally
occurs 5 to 10 days after thrombus formation. 8
Serial clinical studies have attempted to docu-
ment the natural history of DVT. The rate of clot
lysis or retraction and the incidence of venous
valvular incompetence have been evaluated by
duplex ultrasonic scanning. After 7 days, venous
91
segment clot lysis was noted in 44% of the limbs
with acute DVT, and almost all affected limbs
showed some lysis by 30 days. After 90 days, only
14% of the originally thrombosed segments con-
tinued to obstruct venous flow. Within 1 week of
the initial diagnostic study, 17% of limbs with
acute DVT showed reflux in 1 or more venous
segments. And by 1 year, two thirds of the limbs
had incompetent valves. 8
Kakkar et al prospectively followed 153 patients
for 2 years from the onset of DVT and found that
plethysmographic evidence of venous obstruction
correlated well with the distribution of disease by
venography. 9 Despite anticoagulation with heparin
or aggressive thrombolytic therapy, 60% of limbs
with thrombi confined to veins below the knee and
87% with combined calf and iliofemoral thrombo-
ses had moderate to severe hemodynamic abnor-
malities. Other investigators using noninvasive
techniques to follow patients treated with antico-
agulants for DVT also have reported the persis-
tence of venous obstruction and valvular incompe-
tence in a large proportion of limbs. 8.~~
CAUSE OF VENOUS THROMBOSIS
Virchow's triad of the causes for thrombosis
includes (1) changes in the lining of the vessel wall
(wall damage), (2) changes in the flow of blood
(stasis or low flow), and (3) changes in the con-
stituents of the blood (hypercoagulability).~3 These
factors favor clot formation by disrupting the
balance of the opposing coagulative and fibrino-
lytic systems.
Aggregation ~ Propagation ---,,- Embolization ~ r g a n i z a t i o n
Fig2. Netursl history of DVT. The eddy currents thet occur
as blood passes a valve encourage the deposition of throm-
bus within the valve cusp. The initial platelet cluster on the
vessel builds and grows toward the center of the vessel
lumen. When a thrombus adheres to the endothelium, it
stimulates a hyperemic inflammatory response that begins
the processes of organization. Thrombus retraction and orga-
nization destroys the valves in the affected segment of the
vein.
92
Flow Reduction and Stasis
Venous return from the legs is enhanced by
contraction of the calf muscles, which propels
blood upward from the extremities. Venous stasis
contributes to thrombogenesis by allowing acti-
vated coagulation factors to accumulate. 14 Venous
stasis is produced by immobility, venous obstruc-
tion, increased venous pressure, increased blood
viscosity, and venous dilation.
Prolonged immobility and bed rest have been
well documented as risk factors for the develop-
ment of DVT. 15-17 Eddy currents and vortices
within the valve pockets encourage platelet depo-
sition and thrombus formation. 18-2~ McLachlin et
al showed that contrast media remained in valve
sinuses for 30 to 60 minutes if calf muscles were
not exercised. 21 The calf pump does not function
during general anesthesia, and the reduction of calf
blood flow may persist for 7 days after an opera-
tion. 15,22 Long periods of immobilization may ex-
plain the "economy class syndrome," suggesting
that people who travel long distances by air are at
a higher risk for DVT and embolism, even in the
absence of a previous history of cardiovascular
problems.23,24
Immobility contributes to the high incidence of
postoperative venous thrombosis in patients who
have undergone hysterectomy, transabdominal
prostatectomy, hip surgery, knee surgery, or sur-
gery for fractures of the lower limb. Furthermore,
preoperative immobility is associated with a higher
frequency of perioperative venous thrombosis. The
prevalence of venous thrombosis found at autopsy
is markedly increased in individuals who were
confined to bed for more than a week before
death. 25
Venous obstruction contributes to the risk of
venous thrombosis in patients with pelvic tumors.
Furthermore, a raised central venous pressure pro-
duces venous stasis in the extremities, which may
explain the high prevalence of venous thrombosis
in patients with heart failure. Venous stasis can
be caused by venous dilation or increased blood
viscosity, due to an elevated hematocrit (polycy-
themia), hypergammaglobulinemia, dysproteine-
mias, or an increased fibrinogen concentration. The
ability of estrogen to cause venous dilation may
help explain the increased prevalence of thrombo-
sis during pregnancy, 26 or in individuals taking
estrogen-containing oral contraceptive pills z6 or
estrogen replacement therapy. 27
BRUCE R. LINE
Disorders of Coagulation
Coagulation Cascade
A low level of activation of blood coagulation
occurs continuously in healthy patients. 28 Under
normal circumstances, activated coagulation fac-
tors are diluted in the flowing blood and are
neutralized by inhibitors on the surface of endo-
thelial cells or by circulating antiproteinases. 29
However, activated clotting factors that escape
regulation can trigger the coagulation system,
thereby leading to fibrin formation. 3~
Coagulation in vivo is initiated via the extrinsic
or tissue factor pathway. In the tissue factor path-
way, activated factor VII binds to tissue factor at
sites of vascular injury. The tissue factor-factor
VIIa complex then activates the coagulation cas-
cade, which results in the conversion of prothrom-
bin to thrombin. Thrombin then converts fibrino-
gen to fibrin and activates the platelets. Tissue
factor is strongly expressed in all solid tissues,
forming a hemostatic envelope around blood ves-
sels; if a vessel is injured, the exposed tissue factor
triggers coagulation. 31
Hypercoagulability, either at a local or distant
site, is critically important in the initiation of DVT.
Clinical risk factors that predispose to venous
thromboembolism by activating blood coagulation
include malignancy, extensive surgery, trauma,
burns, myocardial infarction, and possibly local
hypoxia produced by venous stasis. 14 Activation of
blood coagulation also increases with age in pa-
tients over 45. Diseases that increase blood viscos-
ity also raise the risk for DVT. Such diseases
include polycythemia rubra vera, leukemia, multi-
ple myeloma, and other macroglobulinemias.
Postoperatively, there is an acute phase re-
sponse, which results in an increase in the platelet
count, a decrease in fibrinolytic potential, and an
increase in the concentration of coagulation factors
(including fibrinogen). 32 The fibrinogen half-life
also decreases, which is presumably a reflection of
the increased turnover of coagulation factors. 33'34
The rise in plasma fibrinogen concentration also
increases blood viscosity, which in turn slows
venous circulation. 35
In vivo activation also is increased above normal
in some patients with hereditary deficiencies of
antithrombin, protein C, and protein S, 28 as well as
in patients with activated protein C resistance. 36
Antithrombin deficiency (antithrombin III deft-
DIAGNOSIS OF DVT
ciency) appears to be transmitted as an autosomal
dominant and is estimated to occur in 1 of every
2,000 families. 37,38 Acquired antithrombin defi-
ciency can develop in patients with the nephritic
syndrome (ie, as a loss of the protein in the urine).
It can also be caused by increased consumption in
patients with an acute venous thrombosis and in
those receiving heparin. 39 It can develop postoper-
atively4~ and in women who are taking oral
contraceptives. 42 Activated protein C resistance
may be a very common contributor to risk for DVT.
Dahlbeck43 showed that activated protein C resis-
tance was present in 40% of a population with DVT.
Fibrinolytic System
The fibrinolytic system is responsible for throm-
bus and clot breakdown. The basic reaction of the
plasma fibrinolytic system is the conversion of
plasminogen to the active enzyme plasmin. Plas-
min hydrolyzes fibrin and various plasma coagu-
lation proteins, including fibrinogen. Two endoge-
nous plasminogen activators, tissue plasminogen
activator and urokinase, are synthesized by and
released from endothelial cells. Tissue plasmino-
gen activator is released in response to a number of
different stresses, including stasis, adrenaline, and
exercise. 44 Defective production or release of acti-
vator from the vein wall has been found to be
associated with recurrent DVT. 45
Decreased fibrinolytic activity occurs in the
early postoperative period 46 in individuals taking
oral contraceptives, 47 during the last trimester of
pregnancy, 47 and in obese patients. 48 Fibrinolytic
activity in the leg veins is less than that in the arm
veins, which may partly explain the greater ten-
dency for venous thrombosis to occur in the lower
extremities. The relative reduction in fibrinolytic
activity in the leg veins is more marked in the
elderly and declines with age as the risk of post-
operative DVT increases. 49
Vessel Wall Damage
Although the normal endothelium is nonthrom-
bogenic, damage or injury to the endothelium can
trigger the activation of platelets and coagulation.
The vascular endothelium can be damaged by
direct trauma, thrombin, or low oxygen tension. It
can also be damaged by exposure to endotoxin
or inflammatory cytokines, such as interleukin-I
(IL-1) and tumor necrosis factor.
Vascular injury leads to the expression of tissue
93
factor, either directly by endothelial cells or by
monocytes that are attracted to the site of damage.
Furthermore, the exposure of blood to subendothe-
lium leads to platelet adhesion and aggregation. In
addition to activating coagulation, tissue damage
can also lead to impaired fibrinolysis because
inflammatory cytokines induce endothelial cell
synthesis of plasminogen activator inhibitor-1. 46
Endothelial cell stimulation (seconds to min-
utes) causes a retraction of endothelial cells from
each other, which exposes the underlying suben-
dothelium. This is accompanied by translocation of
P-selectin, release of von Willebrand's factor, and
secretion of platelet activating factor. 5~ Endothelial
cell activation (4 to 6 hours) causes progressive
transcription of many genes, including cytokines
(IL-I, IL-6, IL-8) and tissue factor, which is the
main initiator of coagulation.
CLINICAL CONCEPTS
Symptoms and Signs of DVT
The symptoms and signs of venous thrombosis
are caused by obstruction to venous outflow, vas-
cular inflammation, or pulmonary embolization.
Unilateral or asymmetric swelling suggests venous
obstruction and is one of the best signs of DVT.
Bilaterally symmetric edema is more likely to be
caused by a systemic problem, such as congestive
heart failure, but it can occur in inferior vena cava
obstruction. Ordinarily, circumferences measured
at the same level in the lower extremities should
differ by no more than 1 cm. Edema caused by
acute venous thrombosis usually decreases when
the limb is elevated and increases when the limb is
dependent. Swelling may persist for weeks or
months after the initial episode and may never
disappear entirely. Easily pitting edema suggests
that the swelling is reversible and of relatively
short duration. 5j A palpable cord with localized
inflammation is indicative of superficial thrombo-
phlebitis; deep veins are seldom felt. Homan's
sign, defined as pain in the calf with dorsiflexion, is
not only insensitive to DVT but is also very
nonspecific, occurring in many patients on bed
rest. 52 Most limbs with acute DVT are nearly
normal in color or have a faintly cyanotic discol-
oration caused by congestion of the cutaneous
veins. 5~
In seriously ill hospitalized patients, physicians
make the correct diagnosis of DVT by using signs
and symptoms in half of their patients. 5~.53 Predic-
94
tion criteria based on clinical data 53 can be used to
place ambulatory patients in low- (5% probability),
medium- (33% probability), or high-risk (85%
probability) groups. 53 Although these probability
assessments can be used to classify a patient's risk,
they are not sufficiently exact to exclude or con-
firm DVT and cannot replace imaging studies.
Table 1 lists some of the conditions that may
mimic DVT. Because venous thrombosis can co-
exist with any of these conditions, objective testing
for DVT may be required even when an alternative
diagnosis appears likely. The conditions that most
frequently simulate venous thrombosis are a rup-
tured Baker's cyst, cellulitis, muscle tear, muscle
cramp, muscle hematoma, external venous com-
pression, superficial thrombophlebitis, and post-
phlebitic syndrome.
Risk Factors
The risk of DVT increases with age, but there is
no difference between the sexes, and the risk in
children is small, al,54 Between 10% and 40% of
patients over 40 who undergo major abdominal or
thoracic surgery develop calf vein thrombosis; 2%
to 8% develop proximal venous thrombosis; 1% to
8% develop clinically evident pulmonary embo-
lism; and 0.1% to 1% develop a fatal pulmonary
embolus. In high-risk patients and those undergo-
ing orthopedic surgery of the lower limbs, the
incidence of calf vein thrombosis and proximal
venous thrombosis jumps to 40% to 80% and 10%
to 20%, respectively. Approximately 4% to 10% of
patients in this category experience a pulmonary
Table 1. Conditions That May Mimic Acute DVT
Muscle strain or blunt trauma
Ruptured muscle with subfascial hematoma
Spontaneous hemorrhage or hematoma
Ruptured synovial cysts (Baker's cysts)
Arthritis, synovitis, or myositis
Cellulitis, lymphangitis, or inflammatory lymphedema
Superficial thrombophlebitis
Arterial insufficiency
Pregnancy or oral contraceptive use
Lymphedema
Lipedema
Chronic venous insufficiency or venous reflux syndromes
Extrinsic venous compression: lymphadenopathy, tumors,
lymphomas, hematomas, abscesses, right iliac artery
Systemic edema: congestive heart failure, metabolic,
nephritic syndrome, postarterial reconstruction
Dependency or leg immobilization (casts)
Arteriovenous fistula
BRUCE R. LINE
embolism, and 1% to 5 % die as a result. 8 Until
they embolize, the vast majority of these thrombi
are completely asymptomatic.
After 3 months of anticoagulant therapy, the
recurrence rate is 5% to 10% in the following
year. 55,56 Patients whose first episode of venous
thrombosis was idiopathic and those who have
ongoing risk factors, such as prolonged immobili-
zation or cancer, are at higher risk of recurrence.
About 70% of patients referred for clinically sus-
pected venous thrombosis do not have the diagno-
sis confirmed by objective testing. 57 Among the
30% who have venous thrombosis, about 85%
have proximal vein thrombosis and the rest have
thrombosis confined to the calf. 57
Clinical Complications of the Disease
Pulmonary Embolism
Pulmonary embolism (PE) is the most common
preventable cause of death in the hospital, s8 In the
absence of prophylaxis, the frequency of post-
operative fatal PE is 0.1% to 0.8% in patients
undergoing elective general surgery, 0.3% to 1.7%
in patients undergoing elective hip surgery, and 4%
to 7% in patients undergoing emergency hip sur-
gery. 58 Most cases of clinically significant and fatal
PE probably arise from thrombi in the proximal
veins of the legs. Although PE may result from calf
vein thrombosis, these emboli tend to be smaller and
occur less commonly than in patients with proximal
vein thrombosis. 59 Asymptomatic PE is detected by
perfusion lung scanning in about 50% of patients
with documented proximal vein thrombosisP 9
Untreated or inadequately treated venous throm-
bosis is associated with a high complication rate,
which can be decreased markedly by adequate
anticoagulant therapy. About 20% of untreated
silent calf vein thrombi and 20% to 30% of
untreated symptomatic calf vein thrombi extend
into the popliteal vein. When extension occurs, it is
associated with a 40% to 50% risk of clinically
detectable PE. 9
Acute Recurrent DVT
Patients with proximal vein thrombosis who are
inadequately treated have a 47% frequency of
recurrent venous thromboembolism over 3 months.
Patients with symptomatic calf vein thrombosis
who are treated with a 5-day course of heparin
without continuing oral anticoagulants have a re-
currence rate greater than 20% over the next 3
DIAGNOSIS OF DVT
months. 1~ In contrast, clinically detectable recur-
rence occurs in less than 2% of patients with
proximal vein thrombosis if an adequate anticoag-
ulant response is achieved, and the recurrence rate
during the subsequent 3 months of treatment with
oral anticoagulants is 2% to 4%. 55.60
A patient with a history of DVT who returns
with acute swelling or pain in the leg, which
suggests a recurrent episode of thrombosis, pre-
sents a difficult diagnostic problem. Only about
20% to 30% of these patients actually have the
disease; the rest have symptoms arising from
chronic venous insufficiency or from any of the
causes of lower extremity pain (see Table l). 61"62
Anticoagulants are administered when fresh clots
are identified or when there is any question of clot
activity; otherwise, the patient is treated for
chronic venous insufficiency. A diagnostic prob-
lem arises, however, when the age of the thrombus
is uncertain.
Postphlebitic Syndrome
Postphlebitic syndrome (PPS) is a major cause
of morbidity that develops within 3 years in about
2 out of 3 patients who have acute DVT below the
knee. 63 About 20% of these patients have symp-
toms severe enough to interfere with dally life, and
10% to 14% are incapacitated to the extent that
they are unable to work. The symptoms of the
syndrome include pain in the calf that is relieved
with rest and leg elevation; pigmentation and
induration around the ankle and lower third of the
calf; and ulceration, which usually occurs in the
region of the medial malleolus. Patients with ex-
tensive thrombosis involving the iliofemoral vein
frequently have greater disability and may even
have venous claudication, which is characterized
by incapacitating, bursting pain with exercise. 64
PPS is caused by venous hypertension, which is
usually the result of valve destruction. 65 Valve
destruction results in malfunction of the muscular
pump mechanism, which leads to increased pres-
sure in the deep calf veins during ambulation. The
high pressure ultimately renders the perforating
veins of the calf incompetent, so that blood flow is
directed from the deep veins into the superficial
venous system during muscular contraction. This
leads to edema and impaired viability of subcuta-
neous tissues and can lead to venous ulceration. 66
The symptoms and signs of PPS may be delayed
for 5 to 10 years after the initial thrombotic event.
95
The risk of PPS in patients with symptomatic DVT
is 22.8% after 2 years, 28% after 5 years, and
29.1% after 8 years.' 2 When symptoms are acute or
subacute in onset, a diagnosis of PPS should be
considered only after recurrent venous thrombosis
has been excluded by objective tests. Other causes
of recurrent leg pain or swelling include recurrent
muscle strain, internal derangement of the knee or
hip, recurrent cellulitis, extrinsic compression of
the vein, lumbosacral disk disease, and sciatic pain.
Venous Thrombosis in Subclavian
or Axillary Veins
Although thrombosis of the subclavian or axil-
lary veins most frequently occurs as a complication
of a chronic indwelling catheter, it also may occur
as a consequence of local malignancy and is a
well-documented complication of mastectomy and
radiation therapy. 67 Subclavian vein thrombosis is
particularly common in children with indwelling
central venous catheters; this complication devel-
ops in up to 20% of children who have catheters
placed for chemotherapy or parenteral nutrition.
Patients with subclavian or axillary thrombosis
experience pain in the axilla and develop edema
and cyanosis of the ann. If thrombosis extends into
the superior vena cava, patients can develop edema
and cyanosis of the face, neck, and upper extrem-
ities.
DIAGNOSTIC MODALITIES AND PARADIGMS
Objective tests are necessary to confirm the
clinical diagnosis of DVT. Of these diagnostic
procedures, venography is the only invasive test of
proven value, and ultrasonographic (US) studies
are the most commonly used noninvasive modal-
ity. Other procedures are occasionally used to
diagnose DVT, including impedance plethysmog-
raphy, computed tomography (CT), and magnetic
resonance imaging (MRI).
Plethysmography measures electrical impedance
to detect calf blood volume changes induced by
inflation of a thigh cuff. Blood volume change is
reduced by obstruction of the popliteal or more
proximal veins. The test is not specific for throm-
botic obstruction to venous outflow. Thus, false-
positive results may be obtained during pregnancy
if a patient is positioned incorrectly, if the vein is
compressed by an extravascular mass, or if venous
outflow is impaired by increased central venous
pressure. The reported sensitivities and specifici-
96
ties for the diagnosis of symptomatic proximal
venous thrombosis range from as low as 65% to as
high as 95%. 68,69 The test may not detect large,
nonocclusive proximal vein thrombi and fails to
detect most thrombi within the calf. Because ve-
nous reflux is easily recognized with Doppler US
or duplex Scanning, these modalities are preferred
over impedance plethysmography.
CT may be used to diagnose venous thrombosis
in pelvic and abdominal veins, to determine the
cause of leg swelling, or to evaluate a soft tissue
mass, but it is rarely used to diagnose DVT. Like
venography, CT has a disadvantage in that it
requires the intravenous administration of a con-
trast medium. MRI shares with ultrasound the
ability to identify cysts, hematomas, tumors, and
other masses that may be responsible for venous
obstruction or leg swelling. Like US, MRI is
noninvasive, but it can image both extremities
simultaneously and examine veins in the pelvis,
abdomen, and thorax, unencumbered by the pres-
ence of gaseous interfaces. 7~ However, MRI is
expensive, it may not be available in emergency
room settings, and it is impractical for critically ill
patients.7O, 71
Contrast venography (CV) is considered the
gold standard for the detection of DVT. Lower-
limb CV is performed on a fluoroscopic table with
the patient in a semiupright position and bearing
weight on the leg that is not being examined. 72
During fluoroscopy, radiographic contrast material
is infused into the deep venous system via a
catheter placed in the foot. The primary veno-
graphic criterion for the diagnosis of DVT is an
intraluminal filling defect in the deep venous con-
trast column. A less specific finding is the abrupt
cutoff of a deep vein with the development of
collateral circulation. The accuracy of lower-limb
CV is difficult to determine. In one early study, z5
postmortem venograms showed a sensitivity of
95% and a specificity of 97%. However, observers
may disagree about the presence of a thrombus in
as many as 10% of examinations. 73 Extensive
collateral development and recanalization may
make phlebographic findings difficult to interpret.
The deep venous system may be inadequately
visualized in 20% to 25% of patients because of
previous thrombosis, dilution of contrast material
in the proximal lower limb, and difficulties with
venous access related to obesity, severe edema, or
cellulitis. TM Particularly difficult to show are the
BRUCE R. LINE
intramuscular sinusoids, the profunda femoris
vein, and the internal iliac vein, all of which are
areas that may harbor thrombi.
Ultrasound
B-mode, duplex, and color-coded duplex US
examinations are the most commonly used proce-
dures to evaluate lower-limb DVT. The studies are
noninvasive, and they can be rapidly obtained and
performed serially. Nonocclusive clots and clots
with free-floating tails are readily identified and
easily distinguished from those that are completely
obstructive or adherent to the venous w a l l . 75,76
Extrinsic compression that is caused by hemato-
mas, lymphadenopathy, or tumor masses can be
differentiated from intrinsic obstruction; condi-
tions such as intramuscular bleeding and popliteal
cysts can be diagnosed even in the presence of
c o n c o m i t a n t D V T . 75,77-8~
US studies are performed with a 3- to 7-MHz
real-time transducer, with the appropriate gain
settings so that normal vessels are free of internal
echoes. The examination depends primarily on
vein compressibility to detect lower-limb DVT. 81
Normal veins have thin walls, appear to be empty,
and collapse completely when light pressure is
applied with the probe. 82,83 Moderate transducer
compression completely collapses the normal ve-
nous lumen, whereas DVT prevents this. Blood
flow often produces visible echoes during low-flow
states when acceleration or deceleration oc-
curs. 75,84 Diameter changes coinciding with respi-
ration may be observed. Valsalva's maneuver,
coughing, proximal limb compression, and down-
ward tilting all tend to increase the diameter of
normal veins. 75,84
Some consider the response to compression to
be the most sensitive and specific method of
identifying or ruling out venous thrombosis. 85,86
However, several important venous segments are
resistant to compression even when no clot is
present. These include the profunda femoris vein at
its junction with the common femoral vein, and the
superficial femoral vein within the adductor ca-
nal. 87-89 Compression US may be used best in
conjunction with a clinical prediction rule. Data
from a prospective trial of US highlight the advan-
tage of incorporating the clinical probability of
DVT in nonhospitalized patients. 9~ In a patient
with low clinical probability and negative US, the
probability of DVT was less than 1%. In high
DIAGNOSIS OF DVT
clinical probability and negative US, the likelihood
was 24% in the proximal lower limb. In low
clinical probability but positive US, the likelihood
was 57%, and in high clinical probability with
positive US, the likelihood was 100%.
Duplex and color Doppler US examinations
often are used to verify the findings obtained by the
compression technique. 88,89 One can use color flow
Doppler to differentiate blood vessels from non-
vascular structures and to readily distinguish veins
from arteries. Flow patterns are visible simulta-
neously in multiple vessels. Thus, color decreases
the need to assess venous compressibility and to
evaluate Doppler flow patterns.
Clot morphology and echogenicity can be used
to differentiate old from new thrombi. Fresh clots
have an acoustic density only slightly greater than
that of flowing blood and may be practically in-
visible. Shortly after a thrombus has formed, there
is a period of decreasing echogenicity, after which
the clot becomes progressively more reflective. 91.92
Clot retraction begins within a few hours of onset,
leaving a thin gap between the vein wall and the
thrombus. Aging clots contract, become quite firm,
and are totally incompressible. The diameter of a
chronically involved vein is frequently reduced,
and its wall appears to be thickened. Thickened,
contracted, rigid valve cusps, frozen in the open
position or adherent to the vein wall, are indicative
of chronic venous insufficiency.
Actual visualization of a clot is the most specific
sign of venous thrombosis. Unfortunately, fresh
clots are sometimes difficult to see because their
acoustic properties are similar to those of blood.
Failure to detect blood flow is diagnostic of venous
obstruction even though the vein appears normal
on the B-mode scan, which is a frequent occur-
rence in acutely thrombosed veins with hypoecho-
genic clots. A continuous flow pattern implies
obstruction of the venous channel above the site
being studied. Killewich et al found the presence or
absence of phasic flow to be the most sensitive
(92%) and specific (92%) criterion for establishing
the diagnosis of DVT with the duplex scanner. 75 In
the series of limbs that were studied, clot visual-
ization, although quite specific (92%), had a sen-
sitivity of only 50%. Compressibility was neither
very sensitive (79%) nor specific (67%).
US studies are limited under the circumstances
shown in Table 2. In symptomatic patients, venous
US is sensitive and specific for proximal DVT;
97
Table 2. Errors and Limitations of US Studies
Technically difficult in obese patients and patients with
swollen limbs. Other factors that may interfere with the
exam include muscle spasm or contraction, gross edema,
and excess obesity.
Accuracy below the knee is poor. Because of their small
size, complicated anatomy, and depth, calf veins are less
easily examined than thigh veins; consequently, full
visualization may be inordinately time consuming and
frustrating.
Low sensitivity in asymptomatic, high-risk patients.
Difficulty differentiating between acute recurrent venous
thrombosis and old organized thrombus.
May miss DVT in duplicate veins.
Clot visualization can be subjective. Fresh clot is often
anechogenic.
Presence of bowel gas may obscure veins lying proximal to
the inguinal ligament.
The inability to compress normal veins in the adductor canal
region can lead to false-positive errors unless this
limitation is recognized. Duplication of the superficial
femoral and popliteal veins is not uncommon.
False-negative results can occur when a patent vein is found
paralleling an occluded channel.
Studies are compromised when the patient is apprehensive
or in pain. A complete exam may not be possible in limbs
with fresh surgical wounds or when cases, traction, or
splints limit access to underlying veins.
Operator dependent. Accuracy depends on the quality of the
study. The necessary skills are not easily or quickly
acquired.
Probe compression as a method of diagnosing DVT carries
the very slight risk of dislodging a fresh clot.
however, US is insensitive to calf vein thrombosis
and to DVT occurring after surgery. 93 Most of
these thrombi are nonocclusive and not symptom-
atic. There is clear evidence from data combined
from multiple studies that US procedures show
reduced sensitivity in asymptomatic patients
(Table 3). Even with color Doppler US, DVT in
the calf usually remains undetected. This is also a
significant problem in postoperative patients who
may have thrombi in both pelvic and calf
veins. 94.95 The sensitivity of US for clinically im-
portant calf thrombosis can be increased, and the
safety of diagnostic strategies that do not include
venography in patients with suspected calf vein
thrombosis can be improved, if US is repeated 7
days after the initial study. This strategy detects the
10% to 30% of calf vein thrombi that extend prox-
imally. If the test remains negative after 7 days, the
risk of clinically important proximal extension is
negligible, and it is safe to withhold treatment. 57.68
Patients with symptoms of recurrent DVT can
present a difficult diagnostic problem. Up to 50 %
of patients have compression ultrasound abnormal-
98
Table 3. Accuracy of US in First DVT in Symptomatic and Asymptomatic Patients
Sensitivity
All Proximal Distal
Symptomatic 89 97 73
Asymptomatic 47 62 53
NOTE. Accuracy (comparison with CV) in prospective cohort studies and randomized comparison studies identified from Medline
search. Values are weighted mean of study results. Data from Kearon et al: Ann Intern Med 128:663-667, 1998.
ities that are indistinguishable from the original
findings of DVT for 6 months after the acute
episode. Thus, it may not be possible to distinguish
new acute DVT from old. Like other flow-depen-
dent methods, Doppler US can detect only those
thrombi that involve the major deep veins that exist
in continuity from the ankle (or wrist) to the heart.
Isolated thrombi within tributaries, such as the
internal iliac, profunda femoris, gastrocnemial, and
soleal veins, will escape recognition. The problem
is more serious in the calf. Isolated calf vein
thrombi occur in 20% to 46% of limbs with DVT.
Unfortunately, the peroneal veins, which are in-
volved in 81% of limbs with calf vein thrombosis,
are difficult to examine.
I n d i c a t i o n s for V e n o u s S c i n t i g r a p h y
Although the diagnostic modalities for DVT
have greatly enhanced the management of patients
suspected of having the disease, there are a signif-
icant number of patients and clinical circumstances
in which the diagnosis of DVT is difficult. As is
discussed elsewhere in this issue, 99mTc-radiola-
beled peptides that target the molecular biology of
thrombosis may complement a negative or equiv-
ocal ultrasound study, particularly in patients who
have undergone surgery, patients with intermedi-
ate- or low-probability lung scans, and patients
with active DVT in the calf and or pelvis. Venous
scintigraphy may add specificity for the diagnosis
in symptomatic patients who have had previous
episodes of DVT and equivocal results on com-
pression US.
A scintigraphic molecular imaging approach
may be of significant value in the diagnosis of
acute recurrent DVT. The diagnosis is difficult to
make because the clinical manifestations of recur-
rence are nonspecific. In addition, all of the vali-
dated diagnostic tests for acute venous thrombosis
have limitations in this setting because the venous
occlusion that is produced by the initial episode of
venous thrombosis makes it difficult to identify
BRUCE R. LINE
Positive Negative
Specificity PredictiveValue Predictive Value
All All Proximal All Proximal
94 94 97 90 98
94 80 74 82 95
new abnormalities. Both acute and chronic thrombi
often are present in the same limb. In fact, Rollins
et al found ultrasonic evidence of chronic thrombi
in 48 (76%) of 63 limbs with acute venous throm-
bosis. 96 Only 13% of the patients in this study had
a history of DVT.
Venous scintigraphy should enhance the diagno-
sis of recurrent venous thrombosis and help to
reduce false-positive results caused by PPS. Most
patients with a history of venous thrombosis who
develop new-onset leg pain are not found to have
recurrent DVT when investigated with the appro-
priate diagnostic tests. Some of these patients have
PPS, whereas in others the pain is unrelated to
DVT. 61
Venous scintigraphy should allow improved
visualization of DVT that affects the calf veins.
Limiting a US study to the popliteal and more
proximal veins may simplify the examination, but
it sacrifices one of the unique advantages of imag-
ing, ie, the identification of calf vein thrombi. About
20% to 40% of clots that are responsible for symp-
tomatic DVT are isolated to the infrapopliteal
veins and would not be detected by this approach.
Calf vein thrombi are not benign. About 20%
propagate into the popliteal or more proximal
veins. 57,68 Even those that remain isolated may
occasionally be responsible for pulmonary emboli,
and perhaps a significant number may serve as the
nidus for subsequent, more extensive thrombosis.
Moreover, there is evidence that valvular destruc-
tion caused by calf vein thrombi may play a major
role in the development of chronic venous insuffi-
ciency. 65
Venous scintigraphy may play an important role
in patients suspected of having a pulmonary em-
bolism. It is disturbing that fewer than 10% of
patients dying of a pulmonary embolism have
clinical manifestations of DVT, and that Doppler
survey results are positive in only 23% to 61% of
patients with documented pulmonary emboli. 8 Be-
cause approximately 90% of these emboli originate
DIAGNOSIS OF DVT
in leg veins, clots are either being missed or,
having broken away, are leaving only small, non-
detectable residues. Cheely et al found abnormal
Doppler flow patterns in the legs of only 23% of
patients with angiographically shown PE. Simi-
larly, only 7 of 16 patients (44%) with anglo-
graphically proven PE had leg vein thrombi dem-
onstrated color flow duplex scanning in the study
REFERENCES
1. Anderson FA, Jr, Wheeler HB: Physician practices in the
management of venous thromboembolism: A community-wide
survey. J Vasc Surg 16:707-14,1992
2. Matzdorff AC, Green D: Deep vein thrombosis and
pulmonary embolism: Prevention, diagnosis, and treatment.
Geriatrics 47:48-63,1992
3. Cronan J J: Venous thromboembolic disease: The role of
US. Radiology 186:619-630,1993
4. Sevitt S. The structure and growth of valve pocket
thrombi in femoral veins. J Clin Pathol 27:517-528, 1974
5. Bauer KA, Kass BL, ten Care H, et al: Factor IX is
activated in vivo by the tissue factor mechanism. Blood
76:731-736, 1990
6. Stewart GL, Ritchie WG, et al: Venous endothelial
leukocytes. Am J Pathol 74:507-532, 1974
7. Zahn W: Untersuchungen iiber thrombose. Zentralbl Med
Wissenschaften 10:129, 1872
8. Salzman EW, Hirsh J: The epidemiology, pathogenesis
and natural history of venous thrombosis, in Colman RW, Hirsh
J, Marder VJ, et al (eds): Hemostasis and Thrombosis: Basic
Principles and Clinical Practice, ed 3. Philadelphia, PA, Lip-
pincott, 1994
9. Kakkar VV, Howe CT, Flanc C, et al: Natural history of
postoperative deep-vein thrombosis. Lancet 2:230-232, 1969
10. Lagerstadt CI, Fagher BO, Olsson CG, et al: Need for
long-term anticoagulant treatment in symptomatic calf-vein
thrombosis. Lancet 2:515-518, 1985
11. Ochsner A, DeBaker ME, DeCapt PT: Venous thrombo-
sis: Analysis of 580 cases. Surgery 29:24-43, 1951
12. Prandoni P, Lensing AW, Cogo A, et al: The long-term
clinical course of acute deep venous thrombosis. Ann Intern
Med 125:1-7, 1996
13. Virchow R: Die verstopfund den lungenarteries und ihre
folgen. Betir Exp Pathol Physiol 21, 1846
14. Ogawa S, Gerlach H, Eposito C, et al: Hypoxia mediates
the barrier and coagulant function of bovine endothelium:
Increased monolayer permeability and induction of procoagu-
lant properties. J Clin Invest 85:1090-1098, 1980
15. Browse NL: Effect of bedrest on resting calf blood flow
of healthy adult males. Br Med J 1:1721, 1962
16. Canavarro K: Early post-operative ambulation. Am Surg
124:180, 1946
17. Gibbs NM: Venous thrombosis in the lower limbs with
particular reference to bedrest. Br J Surg 45:209-236, 1957
18. Atik N, Broghamer W: The impact of prophylactic
measures in fatal pulmonary embolism. Arch Surg 114:366-
369, 1979
19. Smith R, Thick E, Coalston J, et al: Thrombus produc-
tion by turbulence. J AppI Physiol 32:261-264, 1972
99
reported by Kiilewich et al. 75 Thus, despite the
high sensitivity of this modality, most patients with
pulmonary emboli had negative noninvasive find-
ings. If the venous coagulation disorder persists
and requires therapy, venous scintigraphy should
allow detection of the ongoing thrombus well
before the clot mass causes symptoms, obstructs
vascular flow, or compromises venous valves.
20. Stein P, Sabbah H: Measured turbulence and its effect on
thrombus formation. Circ Res 35:608-614, 1974
21. McLachlin AD, McLachlin JA, Jory TA, et al: Venous
stasis in the lower extremities. Ann Surg 152:678-685, 1960
22. Ludlam CA, Cash JD: 33-Thromboglobulin: A new tool
for the diagnosis of hypercoagulability?, in Neri Semer GG,
Prentic CRM. (eds): Haemostasis and Thrombosis. London:
Academis Press, 1978, pl59
23. Eklof B. Kistner RL, Masuda EM, et al: Venous throm-
boembolism in association with prolonged air travel. Dermatol
Surg 22:637-641. 1996
24. Milne R: Venous thromboembolism and travel: Is there
an association: J R Coil Physicians Lond 26:47-49, 1992
25. Lund F. Diener L, Ericsson JL: Postmortem intraosseous
phlebography as an aid in studies of venous thromboembolism.
Angiology 20:155-179, 1969
26. Lund F, Diener L, Ericsson JL: Postmortem intraosseous
phlebography as an aid in studies of venous thromboembolism.
Angiology 20:155-179, 1969
27. Grodstein F. Stampfer M J, Goldhaber SZ, et al: Prospec-
tive study of exogenous hormones and risk of pulmonary
embolism in women. Lancet 348:983-987, 1996
28. Bauer KA. Goodman TL. Kass BL, et al: Elevated factor
Xa activity in the blood of asymptomatic patients with congen-
ital antithrombin deficiency. J Clin Invest 76:826-836, 1985
29. Esmon CT: Protein C: The regulation of natural antico-
agulant pathways. Science 235:1348-1352, 1987
30. Wiman B. Ljungberg B, Chmielewska J, et al: The role
of the fibrinolytic system in deep vein thrombosis. J Lab Clin
Med 105:265-270, 1985
31. Drake TA, Morissey JH, Edgington TS: Selective cellu-
lar expression of tissue factor in human tissues. Implications for
disorders of hemostasis and thrombosis. Am J Pathol 134:1087-
1097, 1989
32, Davies JW, Forsberg K, Liljedaht SO, et al: The effect of
anticoagulants on postoperative fibrinogen metabolism. Sec-
ondary anaemias. XVI. Acta Med Scand 194:277-283, 1973
33. Davies JWL, Liljedahl SO, Reizenstein P: Fibrinogen
metabolism following injury and its surgical treatment. Injury
1:178-185, 1970
34. Hickman JA: A study of the metabolism of fibrinogen
after surgical operations. Clin Sci 41:141 - 152, 1971
35. Matsuda T, Murakami M: Relationship between fibrin-
ogen and blood viscosity. Thromb Res 8:25-33, 1976 (suppl 2)
36. Zoller B, Holm J, Svenson P, et al: Elevated levels of
prothrombin activation fragment 1 +2 in plasma from patients
with heterozygous Arg 5~ to Gin mutation in the factor V gene
(APC-resistance) and/or inherited protein S deficiency. Thromb
Haemost 76:270-274, 1996
100
37. Rosenbeig RD: Actions and interactions of antithrombin
and heparin. N Engl J Med 292:146-151, 1975
38. Rosenberg RD: Heparin antithrombin and abnormal
clotting. Annu Rev Med 29:367-378, 1978
39. Von Kaulla E, Von KauUa KN: Deficiency of antithrom-
bin 111 activity associated with hereditary thrombosis ten-
dency. J Med 3:349-358, 1972
40. Olsson P: Variations in antithrombin activity in plasma
after major surgery. Acta Chit Scand 126:24-33, 1963
41. Stathatkis N, Papayannis AG, Gardikas C: Postoperative
antithrombin III concentration. Lancet 1:430, 1973
42. Von Kaulla E, Von Kaulla KM: Oral contraceptives and
low antithrombin 111 activity. Lancet 1:36, 1970
43. Dahlbach B: New molecular insights into the genetics of
thrombophilia. Resistance to activated protein C caused by
ArgS06 to Gin mutation in factor V as a pathogenic risk factor
for venous thrombosis. Thromb Haemost 74:139-148, 1995
44. Astrup T: Tissue activators of plasminogen. Fed Proc
25:42-51, 1996
45. Isacson S, Nilsson IM: Defective fibrinolysis in blood
vein walls in recurrent "idiopathic" venous thrombosis. Acta
Chir Scand 138:313-319, 1972
46. Prins MH, Hrish J: A critical review of the evidence
supporting a relationship between impaired fibrinolysis and
venous thromboembolism. Arch Inter Med 151:1721-1731, 1991
47. Menon IS, Peberdy M, Rannie RH, et al: A comparative
study of blood fibrinolytic activity in normal women, pregnant
women and women on oral contraceptives. J Obstet Gynaecol
Br Common 77:752-756, 1970
48. Ogston D, McAndrew GM: Fibrinolysis in obesity.
Lancet 2:1205-1207, 1964
49. Johnson PM, Berdon WE, Baker DH, et al: Thymic
uptake of gallium-67 citrate in a healthy 4-year-old boy. Pediatr
Radiol 7:243-244, 1978
50. Bach FH, Robson SC, Winkler H, et al: Barriers to
xenotransplantation. Nat Med 1:869-871, 1995
51. McLachlin J, Richards T, Paterson JC: An evaluation of
clinical signs in the diagnosis of venous thrombosis. Arch Surg
85:738-744, 1962
52. Haeger K: Problems of acute deep venous thrombosis. I.
The interpretation of signs and symptoms. Angiology 20:219-
223, 1969
53. Wells PS, Hirsh J, Anderson, et al: Accuracy of clinical
assessment of deep-vein thrombosis. Lancet 345:1326-1330,
1995
54. Harvey-Kemble JV: The incidence of deep vein throm-
bosis. Br J Hosp Med 6:721-726, 1971
55. Hull RD, Delmore T, Genton E, et al: Warfarin sodium
versus low-dose heparin in the long-term treatment of venous
thrombosis. N Eng] J IVied 301:855-858, 1979
56. Hull RD, Hirsh J, Jay RM, et al: Different intensities of
anticoagulation in the long-term treatment of proximal vein
thrombosis. N Engl J Med 307:1676-1681, 1982
57. Hull RD, Hirsh J, Carter C, et al: Diagnostic efficacy of
impedance plethysmography for clinically suspected deep-vein
thrombosis: A randomized trial. Ann Intern Med 102:21-28,
1985
58. Gallus AS, Salzman EW, Hirsh J, et al: Prevention of
venous thromboembolism, in Colman RW, Hirsh J, Marder VJ
(eds): Hemostasis and Thrombosis: Basic Principles and Clin-
ical Practice, ed 3. Lippincott, Philadelphia, PA, 1994
BRUCE R. LINE
59. Doyle DJ, Turpie AGG, Hirsh J, et aI: Adjusted subcu-
taneous heparin or continuous intravenous heparin in patients
with acute deep vein thrombosis: A randomized trial. Ann
Intern Med 107:441-449, 1987
60. Hull RD, Delmore T, Carter C, et al: Adjusted subcuta-
neous heparin vs warfarin sodium in the long-term treatment of
venous thrombosis. N Engl J Med 306:189-194, 1982
61. Hull RD, Carter C, Jay R: The diagnosis of acute
recurrent DVT: A diagnostic challenge. Circulation 67:901-
906, 1983
62. Huisman MV, Bueller HR, ten Cate JW: Utility of
impedance plethysmography in the diagnosis of recurrent deep-
vein thrombosis. Arch Int M e t 148:681-683, 1988
63. Strandness DE, Jr, Langlois Y, Cramer M, et al: Long-
term sequelae of acute venous thrombosis. JAMA 250:1289-
1292, 1983
64. Cockett FB, Lea Thomas M, Negus D: Iliac vein
compression: Its relation to iliofemoral thrombosis and the
post-thrombotic syndrome. Br Med J 2:14-19, 1967
65. Negus D: The post thrombotic syndrome. Ann R Coll
Surg Engl 47:92-105, 1970
66. Owens JC: The post-phlebitic syndrome management by
conservative means, in Bergman JJ, Yao JST (eds): Venous
Problems. Chicago, 1L, Year Book Medical Publishers, 1978,
pp 369-382
67. Urokinase Streptokinase Embolism Trial: Phase II re-
sults. A cooperative study. JAMA 229:1606-1613, 1974
68. Huisman MV, Buller HR, ten Cate CJ, et al: Serial
impedance plethysmography for suspected deep venous throm-
bosis in outpatients: The Amsterdam General Practitioner
Study. N Engl J Med 314:823-828, 1986
69. Wells PS, Hirsh J, Anderson DR, et al: Comparison of
the accuracy of impedance plethysmography and compression
ultrasonography in outpatients with clinically suspected deep
Vein thrombosis: A two center paired-design prospective trial.
Thromb Haemost 74:1423-1427, 1995
70. Spritzer CE, Sussman SK, Blinder RA, et al: DVT
evaluation with limited flip-angle, gradient refocused MRI:
Preliminary experience. Radiology 166:371-375, 1988
71. Erdman WA, Jayson HT, Redman HC, et al: DVT of
extremities. Role of MRI in the diagnosis. Radiology 174:425-
429, 1990
72. Wojtowycz M: Lower extremity venography, in Woj-
towycz M (ed): Interventional Radiology and Angiography.
Chicago, IL, Year Book Medical, 1990, pp 227-240
73. Wille-Jorgensen P, Borris L, Jorgensen LN, et al: Phle-
bography as the gold standard in thromboprophylactic studies?
A multicenter interobserver variation study. Acta Radiol 33:24-
28, 1992
74. White RH, McGahan JP, Saschbach MM, et al: Diagno-
sis of deep-vein thrombosis using duplex ultrasound. Ann
Intern Med 111:297-304, 1989
75. Killewich LA, Bedford GR, Beach KW, et al: Diagnosis
of deep venous thrombosis. A prospective study comparing duplex
scanning to contrast venography. Circulation 79:810-814, 1989
76. Langsfeld M, Hershey FP, Thorpe L, et al: Duplex
B-mode imaging for the diagnosis of deep venous thrombosis.
Arch Surg 122:587-591, 1987
77. Cavaye D, Kelly AT, Graham JC, et al: Duplex ultra-
sound diagnosis of lower limb venous thrombosis. Aust N Z
J Surg 60:283-288,1990
DIAGNOSIS OF DVT
78. Wright DJ, Shepard AD, McPharlin M, et al: Pitfalls in
lower extremity venous duplex scanning. J Vasc Surg 5:675-
679, 1990
79. Foley WD, Middleton WD, Lawson TL, et al: Color
Doppler ultrasound imaging of lower-extremity venous disease.
AJR Am J Roentgenol 152:371-376, 1991
80. Mussurakis S, Papaioannou S, Voros D, et al: Compres-
sion ultrasonography as a reliable imaging monitor in deep
venous thrombosis. Surg Gynecol Obstet 171:233-239, 1990
81. Cronan JJ: Ultrasound evaluation of deep venous throm-
bosis. Semin Roentgenol 27:39-52, 1992
82. Appelman PT, DeJong TE, Lampmann LE: Deep venous
thrombosis of the leg. US findings, Radiology 163:743-746,
1987
83. Sullivan ED, Peter DJ, Cranley JJ: Real-time B-mode
venous ultrasound. J Vasc Surg 1:465-471, 1984
84. Effeney DJ, Friedman MB, Gooding GA: Iliofemoral
venous thrombosis: Real-time ultrasound diagnosis, normal
criteria, and clinical application. Radiology 150:787-792, 1984
85. Barnes RW, Nix L, Barnes CL, et al: Perioperative
asymptomatic venous thrombosis: Role of duplex scanning
versus venography. J Vasc Surg 9:251-260, 1989
86. Comerota AJ, Katz ML, Grossi ILl, et al: The compara-
tive value of non-invasive testing for diagnosis and surveillance
of deep vein thrombosis. J Vasc Surg 7:40-49, 1988
87. Rose SC, Zwiebel WJ, Nelson BD, et al: Symptomatic
lower extremity deep venous thrombosis: Accuracy, limitations,
and role of color duplex flow imaging in diagnosis. Radiology
175:639-644, 1990
101
88. Foley WD, Middleton WD, Lawson TL, et al: Color
Doppler ultrasound imaging of lower-extremity venous disease.
AJR Am J Roentgenol 152:371-376, 1989
89. Cronan JJ, Froehlich JA, Dorfman GS: Image-directed
Doppler ultrasound: A screening technique for patients at
high risk to develop deep vein thrombosis. J Clin Ultrasound
19:133-138, 1991
90. Wells PS, Hirsh J, Anderson, et al: Accuracy of clinical
assessment of deep-vein thrombosis. Lancet 345:1326- 1330, 1995
91. Alanen A, Kormano M: Correlation of the echogenicity
and structure of clotted blood. J Ultrasound Med 4:421-425,
1985
92. Coelho JC, Siegel B, Ryva JC, et al: B-mode sonography
of blood clots. J Clin Ultrasound 10:323-327, 1982
93. Wells PS, Lensing AW, Davidson BL, et al: Accuracy of
ultrasound for the diagnosis of deep venous thrombosis in
asymptomatic patients after orthopedic surgery: A meta-analy-
sis. Ann lntem Med 122:47-53, 1995
94. Ginsberg JS, Caco CC, Brill-Edwards PA, et al: Venous
thrombosis in patients who have undergone major hip or knee
surgery: Detection with compression US and impedance pleth-
ysmography. Radiology 181:651-654, 1991
95. Davidson BE Elliot CG, Lensing AW: Low accuracy of
color Doppler ultrasound in detection of proximal leg vein
thrombosis in asymptomatic high-risk patients. Ann Intern Med
117:735-738, 1992
96. Rollins, DL, Semrow CM, Friedell ML, et al: Progress in
the diagnosis of deep venous thrombosis. J Vasc Surg 7:638-
641, 1988