See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/14942546

Distribution of thrombosis in patients with symptomatic deep vein

thrombosis: Implications for simplifying the diagnostic process with
compression ultrasound

Article  in  Archives of Internal Medicine · January 1994

DOI: 10.1001/archinte.153.24.2777 · Source: PubMed

CITATIONS READS

217 380

4 authors, including:

Anthonie W A Lensing Paolo Prandoni

Academisch Medisch Centrum Universiteit van Amsterdam University of Padova
240 PUBLICATIONS   24,666 CITATIONS    577 PUBLICATIONS   32,501 CITATIONS   

SEE PROFILE SEE PROFILE

Jack Hirsh
McMaster University
948 PUBLICATIONS   70,840 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

EINSTEIN-Jr. rivaroxaban program in children with VTE View project

Effect of Short-Term vs. Long-Term Blood storage on Mortality after Transfusion View project

All content following this page was uploaded by Jack Hirsh on 22 May 2014.

The user has requested enhancement of the downloaded file.

Distribution of Thrombosis in Patients With
Symptomatic Deep Vein Thrombosis
Implications for Simplifying the Diagnostic Process With Compression Ultrasound
Alberto Cogo, MD; Anthonie W. A. Lensing, MD, PhD; Paolo Prandoni, MD, PhD; Jack Hirsh, MD

Background: Two different diagnostic strategies are used
to perform compression (real-time) ultrasound for the di-
agnosis of clinically suspected deep-vein thrombosis. One
is to examine the entire proximal venous system from com-
mon femoral to distal popliteal vein; the other is a limited
examination of only the common femoral and the entire
popliteal vein. The latter strategy, which is less time\x=req-\
consuming and requires less expensive equipment, is based
on a strong impression from prospective studies using lim-
ited compression ultrasound that proximal vein thrombi
always involve the common femoral or popliteal vein. This
impression, which is supported by the demonstrated safety
at long-term follow-up of not treating patients whose lim-
ited compression ultrasound is normal at presentation and
then repeated within the next week, has not been tested
in a formal study. Therefore, we reviewed a large series of
venograms performed in consecutive patients with clini-
tive patients with a first episode of clinically suspected deep
vein thrombosis were adjudicated. Of these, 20 (3.6%) were
inadequate for interpretation. In the remaining 542, venous
thrombosis was demonstrated in 189 instances (prevalence,
35%; 95% confidence interval, 31% to 39%) and were lo-
cated in the proximal veins in 166 (88%; 95% confidence
interval, 82% to 92%) venograms. Isolated calf-vein throm-
bosis was present in the remaining 23 (12%; 95% confidence
interval, 8% to 18%) venograms. Proximal with concurrent
calf thrombosis was detected in 164 (99%) of the 166 pa-
tients. Proximal thrombi involved only the popliteal vein in
16 (10%); the popliteal and superficial femoral veins in 70
(42%); and the popliteal, superficial, and common femoral
vein in eight (5%); whereas thrombi involving the entire proxi-
mal deep venous system were detected in 58 (35%) veno-
grams. Isolated thrombosis of the superficial femoral, com-
mon femoral, and iliac vein was not observed. Proximal venous

cally suspected venous thrombosis to determine the dis-
tribution of venous thrombosis in symptomatic patients.
Methods: Venograms were performed using 150 mL of
radiographic contrast material. Before the study, a panel of
experts agreed on the standardized criteria for the assess-
ment of venograms. Venograms were adjudicated blindly
for the presence of deep vein thrombosis and to determine
the distribution of proximal vein thrombosis and isolated
calf-vein thrombosis, the size of proximal thrombi, and whether
they were occlusive or nonocclusive. Subsequently, the du-
ration of symptoms was related to the venographic findings.
thrombi were occlusive in 146 (88%) patients. No relation
between the duration of symptoms and the extent or the oc-
clusiveness of venous thrombi could be demonstrated.
Conclusions: Most symptomatic patients have exten-
sive occlusive proximal vein thrombosis at the time of
presentation. Thrombi isolated to the superficial femoral
or iliac vein were not observed in this large sample of
consecutive patients. Our data support the use of the rela-
tively simple, inexpensive, and rapid compression ultra-
sound method that limits the examination of the proxi-
mal veins to the common femoral and popliteal veins.

Results: Five hundred sixty-two venograms from consecu- (Arch Intern Med. 1993;153:2777-2780)

From Clinica Medica II,
University of Padua (Italy) (Drs
Cogo and Prandoni); the Center
for Hemostasis, Thrombosis,
Atherosclerosis and Inflammation
Research, Academic Medical
Center, Amsterdam, the
Netherlands (Dr Lensing); and
the Hamilton Civic Hospitals
Research Center, Hamilton,
Ontario (Dr Hirsh).
THE
APPROACH to the diagno¬ plethysmography and then with compres¬
sis of venous thrombosis has sion (real-time) ultrasonography,
changed markedly over the
last two decades.1 Venogra-
phy became the accepted
standard after a number of studies dem¬
See Materials and Methods
onstrated that clinical diagnosis is insen¬ on next page
sitive and nonspecific.2"' Subsequently, non-
invasive testing, first with impedance

MATERIAES AND METHODS
VENGORAMS
All venograms performed in consecutive outpatients with a
first episode of clinically suspected deep vein thrombosis
during the period between 1984 and 1989 were eligible for
inclusion in the study and all were retrieved and adjudi¬
cated. Venography was performed using a previously de¬
scribed technique.lg,2° Patients were examined in a semi-
upright position on a tilt table, with the examined leg in a
non-weight-bearing position. A water-soluble nonionic low
osmolar contrast agent was injected into a dorsal foot vein.
An ankle tourniquet was applied in patients with obvious
venous insufficiency. Long-leg roentgenograms outlining the
complete deep venous system of the leg were obtained with¬
out fluoroscopic control. Two anteroposterior long-leg roent¬
genograms were obtained after the injection of 100 mL
followed by another 30 mL of contrast material. A lateral
long-leg roentgenogram was obtained using an additional
20 mL of contrast material. Finally, an exposure of the iliac
vein was obtained after tilting the table into the horizontal
position while simultaneously elevating the leg. This veno-
graphic technique has been demonstrated to be associated
with a low percentage of inadequate results and a high in¬
terobserver agreement.19
INTERPRETATION OF VENOGRAMS
Venograms were interpreted by two observers with exper¬
tise in the interpretation of venograms. Diagnostic criteria
were established and agreed on prior to adjudication. In the
event of a disagreement, the venogram was reviewed by a
third reviewer and the final diagnosis was made by majority
became favored approaches in many centers.6"10 Two dif¬
ferent diagnostic strategies are used to perform compres¬
sion ultrasonography. The first is to examine the entire
proximal venous system from common femoral to distal
popliteal vein.11'13 The second approach, which is sim¬
pler and less time-consuming, limits the examination to
the common femoral vein and the entire popliteal vein.1415
The first more comprehensive approach requires equip¬
ment with (color) Doppler capability to assist in the iden¬
tification of the superficial femoral vein as it passes through
the adductor canal. It has the potential advantage of iden¬
tifying thrombi localized only to the superficial femoral
vein, but has the disadvantages of requiring more expen¬
sive equipment, of being more time-consuming than the
simplified approach, and of occasionally producing false-
positive results due to difficulty in compressing the su¬
perficial femoral vein in the adductor canal.16 The sim¬
plified approach, which also requires less expensive
equipment, has been shown to have great sensitivity and
decision. A venogram was considered normal if both pairs
of peroneal and posterior tibial veins and at least one of the
anterior tibial veins were clearly seen as well as the popliteal,
superficial femoral, common femoral, and iliac veins. A di¬
agnosis of deep vein thrombosis was made if a constant in-
traluminal filling defect was present in more than one pro¬
jection. In addition, nonopacification of a proximal vein (or
venous segment) despite repeated injections of contrast ma¬
terial with either filling of a more proximal segment of the
vein or filling of proximal collaterals was considered to in¬
dicate venous thrombosis. All other test findings were clas¬
sified as inadequate for interpretation. The venograms were
coded as normal, as positive for thrombosis, or as inad¬
equate for interpretation. Thrombi were classified as proxi¬
mal if they were proximal to the trifurcation of the calf veins
and involved at least the popliteal, superficial lemoral, com¬
mon femoral, or iliac vein (with or without concurrent calf-
vein thrombosis), or as isolated calf-vein thrombosis. The
location of proximal thrombi was recorded and they were
classified further as being occlusive or nonocclusive. A throm¬
bus was considered to be nonocclusive if contrast material
was seen between the thrombus and the vessel wall along
the entire length of the thrombus. In addition, the relation
between the time elapsed between the onset of symptoms
and diagnostic testing and the size and degree of occlusive
ness of the thrombus was assessed.
-
ANALYSIS
The 95% confidence intervals (CIs) were calculated accord¬
ing to the binomial distribution. The relationship among
the length of time between symptoms, presentation, and
the extent or occlusiveness of the thrombus was calculated
using the one-way analysis of variance. Two-sided P<.05
were considered statistically significant.
specificity for proximal vein thrombosis when compared
with venography and to be safe on long-term follow-up
in symptomatic patients when a normal result at presen¬
tation is repeated within the next week.14,15,17,18
The safety and effectiveness of the simplified ap¬
proach is based on the concept that proximal-vein thrombi
always involve the popliteal or common femoral vein. Al¬
though results of prospective follow-up studies of symp¬
tomatic patients who have a normal test result on pre¬
sentation support this contention,1718 to our knowledge,
it has not been tested in a formal study.
To determine the distribution of venous thrombosis
in symptomatic patients, we reviewed all venograms in a
series of 562 consecutive patients who were referred to
our center because of clinically suspected deep-vein throm¬
bosis. In addition, we assessed the prevalence of calf-vein
thrombosis in this study population and determined the
relation between the extent and occlusiveness of venous
thrombosis and the duration of patients' symptoms.
 VENOGRAMS

All of the venograms obtained in 562 consecutive symp¬

tomatic patients with a first episode of deep vein throm¬
bosis were analyzed for this study. Venograms were judged
to be inadequate for interpretation in 20 (3.6%) patients
and were adequate for interpretation in the remaining 542
patients. Venous thrombosis was diagnosed in 189 pa¬
tients (prevalence, 35%; 95% CI, 31% to 39%) and was
Popliteal Popliteal, Popliteal, All Proximal Common
excluded in 353 patients. Superficial, Superficial, Veins ± Superficial
and Femoral and Common Femoral or Iliac
Femoral
DISTRIBUTION OF VENOUS THROMBI 10% 42% 5% 35% 8%
95% CI, 95% CI, 95% Cl, 95% Cl, 95% Cl,
6%-15% 35%-50% 2%-9% 28%-43% 5%-14%
Proximal vein thrombosis was seen in 166 (88%; 95%
CI, 82% to 92%) of the 189 patients with deep vein throm¬
bosis, while the remaining 23 (12%; 95% CI, 8% to 18%) The distribution of proximal venous thrombi in patients with clinically
patients had isolated calf-vein thrombosis. One hundred suspected deep vein thrombosis. CI indicates confidence interval.
sixty-four (99%) of the 166 patients with proximal vein
thrombosis had associated calf-vein thrombosis. thrombosis (median delay, 9.2 days; range, 1 to 40 days)
The distribution of proximal vein thrombi is shown compared with that for patients with nonocclusive proxi¬
in the Figure. Venograms showed thrombi of the popliteal mal vein thrombosis (median delay, 13.8 days; range, 1
vein only in 16 (10%) patients; thrombi involved the popliteal to 39 days; P=.08).
and superficial femoral veins in 70 (42%); the popliteal,
superficial, and common femoral vein in eight (5%); and
the entire proximal deep venous system in 58 (35%). The
popliteal veins were not involved in 14 (8%) venograms The main objective of this study was to assess the distri¬
with proximal vein thrombosis; two of these had throm¬ bution of proximal vein thrombi in consecutive symp¬
bosis of the common femoral and iliac vein only (these were tomatic patients to examine the potential safety of limit¬
the only venograms with proximal vein thrombi with nor¬ ing the evaluation of the proximal venous system with
mal calf veins). Isolated thrombosis of the superficial femo¬ compression ultrasonography to the popliteal and com¬
ral, common femoral, or iliac vein was not observed. mon femoral veins. Our findings in this large series of

symptomatic patients demonstrate that all of the proxi¬

EXTENT OF VENOUS THROMBI
IN RELATION TO DURATION OF SYMPTOMS
No difference in time elapsed between the onset of symp¬
toms and patient's referral was observed between patients
with isolated calf-vein thrombosis and patients with proxi¬
mal vein thrombosis (median delay before testing, 8.5 days;
range, 2 to 32 days, vs 9.3 days; range, 1 to 40 days,
respectively; P=.7). In addition, no significant relation¬
ship was noted between the duration of patients' symp¬
toms and the number of proximal venous segments in¬
volved (P=.36).
OCCLUSIVENESS OF PROXIMAL THROMBI
IN RELATION TO DURATION
OF PATIENTS' SYMPTOMS
mal vein thrombi were located in the segments examined
by the limited compression ultrasound approach. The find¬
ings are likely to be valid because bias was minimized by
ensuring that the venograms were interpreted by more
than one expert and the upper 95% confidence limit on
the observed frequency of (zero of 189) isolated super¬
ficial femoral vein thrombi is only 2.6%. Our results are
likely to be generalizable because unselected consecutive
patients were included and the prevalence of venous throm¬
bosis of 35% is similar to that reported in other studies.21
Most symptomatic patients with venous thrombosis (88%)
had proximal vein thrombosis. These proximal thrombi
were usually large and extended into the midthigh region
in more than 80% of patients, and almost all were asso¬
ciated with thrombosis of the calf veins. Of the 166 pa¬
tients with proximal vein thrombosis, the popliteal vein
was involved in 152 (92%). These findings are consistent

Proximal venous thrombi were occlusive in 146 (88%)
venograms and nonocclusive in the remaining 20 veno¬
grams. A nonsignificant trend was noted toward a shorter
time elapsed between the onset of symptoms and referral
to the hospital for patients with occlusive proximal vein
with (but do not prove) the hypothesis that thrombi in
symptomatic patients usually originate in the calf veins
and that most subsequently extend into the popliteal and
proximal veins.22"24
Surprisingly, the interval between the onset of symp-
toms and presentation for diagnostic testing was similar
for patients with isolated calf-vein thrombosis, limited proxi¬
mal vein thrombosis, and extensive proximal vein throm¬
bosis. Most proximal thrombi (88%) were occlusive and
patients with occlusive venous thrombosis showed a trend
for having a shorter duration of symptoms than those with
nonocclusive thrombi (median duration of symptoms, 9.2
days and 13.8 days, respectively; P=.08).
There are two main clinical implications of our find¬
ings. The first is that available noninvasive diagnostic tests
that are sensitive to proximal vein thrombosis but have
limited sensitivity to calf-vein thrombosis will fail to de¬
tect only a small proportion of symptomatic patients at
presentation because they have isolated calf-vein throm¬
bosis. Those calf thrombi that extend can be detected if
they extend by repeating the noninvasive test in the first
week after presentation. Other studies have shown that if
reliable tests are used to detect proximal vein thrombosis,
then it is safe to leave symptomatic patients untreated pro¬
vided that the noninvasive test result remains normal at
approximately 1 week after presentation.6717 The second
clinical implication is that the results of our study sup¬
port the use of a relatively simple, inexpensive, and rapid
compression ultrasound method for detecting proximal-
vein thrombosis in symptomatic patients in which only
the common femoral and popliteal veins are evaluated.
Accepted for publication May 6, 1993.
Reprint requests to Center for Hemostasis, Thrombosis,
Atherosclerosis and Inflammation Research, FA, Academic
Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, the
Netherlands (Dr Lensing).
1. Lensing AWA, Hirsh J, B\l=u"\llerHR. Diagnosis of deep vein thrombosis. In: Col-
man RW, Hirsh J, Marder VJ, Salzman EW, eds. Hemostasis and Thrombosis.
Philadelphia, Pa: JB Lippincott Co. In press.
2. Bauer G. A venographic study of thromboembolism problems. Acta Chir Scand.
1940;84(suppl 161):1-5.
3. Haeger K. Problems of acute deep venous thrombosis, I: the interpretation of
signs and symptoms. Angiology. 1969;20:219-223.
View publication stats
4. Cranley JJ, Canos AJ, Sull WJ. The diagnosis of deep vein thrombosis: falli-
bility of clinical signs and symptoms. Arch Surg. 1976;111:34-36.
5. O'Donnel TF, Abbott WM, Athanasoulis CA, et al. Diagnosis of deep vein throm-
bosis in the outpatient by venography. Surg Gynecol Obstet. 1980;150:69-74.
6. Hull RD, Hirsh J, Carter C, et al. Diagnostic efficacy of impedance plethys-
mography for clinically suspected deep-vein thrombosis: a randomized trial.
Ann Intern Med. 1985;102:21-26.
7. Huisman MV, B\l=u"\llerHR, ten Cate JW, Vreeken J. Serial impedance plethys-
mography for suspected deep venous thrombosis in outpatients. N Engl J Med.
1986;314:823-826.
8. Hull RD, van Aken WG, Hirsh J, et al. Impedance plethysmography using the
occlusive cuff technique in the diagnosis of venous thrombosis. Circulation.
1976;53:696-701.
9. White RH, McGahan JP, Dasenbach MM, Hartling RP. Diagnosis of deep vein
thrombosis using duplex ultrasound. Ann Intern Med. 1989;111:297-303.
10. Becker DM, Philbrick JT, Abbitt PL. Real-time ultrasonography for the diagnosis of
lower extremity deep venous thrombosis. Arch Intern Med. 1989;149;1731-1736.
11. Elias A, LeCorff G, Bouvier JL, et al. Value of real-time B-mode ultrasound
imaging in the diagnosis of deep vein thrombosis of the lower limbs. Int An-
giol. 1987;6:175-180.
12. Vogel P, Laing FC, Jeffrey RB, Wing VW. Deep venous thrombosis of the lower
extremity: US evaluation. Radiology. 1987;163:747-751.
13. Cronan JJ, Dorfman GS, Scola FH, Shepps B, Alexander J. Deep venous throm-
bosis: US assessment using vein compressibility. Radiology. 1987;162:191-196.
14. Appelman PT, de Jong TE, Lampman LE. Deep venous thrombosis of the leg:
US findings. Radiology. 1987;163:743-748.
15. Lensing AWA, Prandoni P, Brandjes D, et al. Detection of deep-vein throm-
bosis by real-time B-mode ultrasonography. N Engl J Med. 1989;320:342-345.
16. Ramshorst B, Legemate DA, Verzijlbergen JF, et al. Duplex scanning in the
diagnosis of acute deep vein thrombosis of the lower extremity. Eur J Vasc
Surg. 1991;5:255-260.
17. Heijboer H, Biiller HR, Lensing AWA, Turpie AGG, Colly LP, ten Cate JW. A
randomized comparison of the clinical utility of real-time compression ultra-
sonography versus impedance plethysmography in the diagnosis of deep-vein
thrombosis of the leg in symptomatic outpatients. N Engl J Med. In press.
18. Sluzewski M, Koopman MMW, Schuur KH, van Vroonhoven TJMV, Ruijs JHJ.
Influence of negative ultrasound findings on the management of in- and out-
patients with suspected deep-vein thrombosis. Eur J Radiol. 1991;13:174-177.
19. Lensing AWA, B\l=u"\llerHR, Prandoni P, et al. Contrast venography, the gold stan-
dard for the diagnosis of deep vein thrombosis: improvement in observer agree-
ment. Thromb Haemost. 1992;67:8-13.
20. Lensing AWA, Prandoni P, B\l=u"\llerHR, et al. Lower extremity venography with
iohexol: results and complications. Radiology. 1990;177:503-507.
21. B\l=u"\llerHR, Lensing AWA, Hirsh J, ten Cate JW. Deep venous thrombosis: new
noninvasive tests. Thromb Haemost. 1991;66:133-139.
22. Nicholaides AN, Kakkar VV, Field ES, Renney JTG. The origin of deep vein
thrombosis: a venographic study. Br J Radiol. 1971;44:653-663.
23. Thomas ML, O'Dwyer JA. Site of origin of deep vein thrombosis in the calf.
Acta Radiol. 1977;18:418-424.
24. Stamatakis JD, Kakkar VV, Lawrence D, Bentley PG. The origin of thrombi in the
deep veins of the lower limb: a venographic study. Br J Surg. 1978;65:449-451.