Thrombosis Research xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Thrombosis Research
journal homepage: www.elsevier.com/locate/thromres

Full Length Article

Diagnosis of deep-vein thrombosis

Enrico Bernardia,⁎, Giuseppe Camporeseb
a
Emergency Unit, Department of Critical Care, aULSS2 “Marca Trevigana”, distretto di Pieve di Soligo, via Brigata Bisagno, 4, 31015 Conegliano, Treviso, Italy
b
Unit of Angiology, Department of Cardiac, Thoracic and Vascular Sciences, University Hospital of Padua, via Giustiniani, 2, 35128 Padova, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: The diagnostic approach to suspected symptomatic deep-vein thrombosis of the lower extremities is usually
Venous thrombosis based on non-invasive methods, including the estimation of clinical probability, the measurement of D-dimer
Diagnosis levels, and ultrasonography. The present review discusses the evidence available from the literature about the
Clinical decision rule management of the first episode of suspected deep-vein-thrombosis.
Fibrin fibrinogen degradation products
Ultrasonography
Doppler
Color
Recurrence

1. Introduction
For descriptive, as well as diagnostic and prognostic purposes, the
deep veins network of the lower limbs is classically divided into two
regions: the proximal and the distal territory. Noteworthy, proximal
DVT is more frequently associated with pulmonary embolism (PE) [1],
and recurrence [2] than isolated distal DVT. The former includes the
femoral (common, superficial, deep or profunda) veins and the popli-
teal vein; the latter comprises the paired anterior and posterior tibial
veins, and the peroneal veins, cumulatively known as axial, plus the
muscular (gastrocnemius, soleal) veins [3]. The calf “trifurcation”,
formed by the joining of the tibial and peroneal veins, though formally
belonging to the distal venous district, is usually screened when prox-
imal CUS is performed. Finally, it is now commonplace to include in the
proximal territory also the last 3 cm of the superficial veins close to the
saphenous junctions.
In symptomatic patients, as shown by classic venographic studies,
DVT invariably develops in the venous-valves sinuses of the distal
network, extending to the proximal system, in the absence of prophy-
laxis or treatment, in 5–20% of the patients [4,5]. Conversely, in
asymptomatic patients, DVT may arise anywhere in the deep-vein
system [3]. This observation, along with the more common finding of
small and non-occlusive thrombi, may account for the lower sensitivity
of ultrasonography for asymptomatic DVT [6].
In recent studies, the prevalence of DVT in symptomatic patients
was around 10–15%, suggesting a low referral-threshold, as compared
with older (venographic) studies reporting figures as high as 35% [7].
Therefore, the main challenge for any diagnostic approach is to rule-out
DVT safely (i.e., low incidence of thromboembolic events at follow-up
in patients left untreated on the basis of normal findings, otherwise
“cleared” from DVT) and efficiently (i.e., the proportion of patients in
whom a given strategy may be safely applied). The commonly accepted
safety threshold is below 2%, corresponding to the follow-up pre-
valence of DVT in patients with a normal venography [8].
Venography, the official “gold standard” for the diagnosis of DVT, is
seldom used in everyday practice, being invasive, costly, technically
demanding, painful, contraindicated in case of allergy or renal in-
sufficiency, and difficult to interpret, with considerable inter- and intra-
observer variability [5]. Alternative (invasive) imaging approaches, i.e.:
CT- and MR-venography, not only share many of the same limitations of
venography, but also do not possess adequate accuracy to be used as
gold standard [5].
Current non-invasive diagnostic algorithms to rule-out suspected
symptomatic DVT include pretest probability estimation, D-dimer, and
ultrasonography [5,9,10]. We will discuss the relevant literature con-
cerning those different strategies in the following sections. Noteworthy,
only ultrasonography may be used a stand-alone test to rule-in or rule-
out DVT.
1.1. Pretest probability
Although useful to raise the clinical suspicion of DVT, individual
clinical features; such as, calf pain or swelling, warmth, tenderness,
erythema, oedema, difference in calf diameter, Homan's sign, history of

⁎
Corresponding author.
E-mail addresses: enrico.bernardi@aulss2.veneto.it (E. Bernardi), giuseppe.camporese@aopd.veneto.it (G. Camporese).

http://dx.doi.org/10.1016/j.thromres.2017.10.006
Received 21 April 2017; Received in revised form 17 September 2017; Accepted 6 October 2017
0049-3848/ © 2017 Elsevier Ltd. All rights reserved.

Please cite this article as: Bernardi, E., Thrombosis Research (2017), http://dx.doi.org/10.1016/j.thromres.2017.10.006
E. Bernardi, G. Camporese
Table 1
The modified Wells DVT rule [12].
Clinical variable
Active cancer (treatment on-going or within previous 6 months or
palliative)
Paralysis, paresis, or recent plaster immobilization of the lower
extremities
Recently bedridden for 3 days or more, or major surgery within the
previous 12 weeks requiring general or regional anaesthesia
Localized tenderness along the distribution of the deep venous system
Entire leg swelling
Calf swelling at least 3 cm larger than that on the asymptomatic leg
(measured 10 cm below the tibial tuberosity)b
Pitting oedema confined to the symptomatic leg
Collateral superficial veins (non-varicose)
Previously documented DVT
Alternative diagnosis at least as likely as DVT
DVT: deep-vein thrombosis.
a
Pretest probability scores may be categorized as follows: “high”: ≥ 3; “moderate”:
1–2; “low”: ≤ 0. The post-test probability of DVT was reported to be: 53% in the high
probability group, 17% in the intermediate probability group, and 5% in the low prob-
ability group, respectively [12]. Alternatively, pretest probability scores may be cate-
gorized as: “likely”: ≥2, or “unlikely” ≤1; and the post-test probability of DVT was
reported to be: 27% in the likely probability group, and 4% in the unlikely probability
group [41], respectively.
b
In patients with symptoms in both legs, the more symptomatic leg was used.
DVT, immobilization, recent surgery, malignancy, or obesity, are useful
but not sufficient to rule in/out DVT [6].
Structured forms clinical judgment (decision rules), based on the
cited individual clinical features, are instead valuable in that they allow
patients to be assigned a definite pretest probability level, being the
probability of DVT progressively higher in patients with higher scores
[6,9,11]. According to international guidelines, the assessment of
pretest probability should come first in the diagnostic pathway of sus-
pected proximal DVT in symptomatic outpatients [5,10].
The more thoroughly studied and validated clinical decision rule,
either in its dichotomized (likely, or unlikely) or tripartite (low, inter-
mediate, high) set-up, is the Well's DVT rule (Table 1) [7,12,13]. Ac-
cording to a meta-analysis of 21 studies, assuming a 15% DVT pre-
valence, the Wells' rule would categorize 18% of the patients as “high
risk” (Score: ≥ 3, DVT probability = 47%), 40% as “intermediate risk”
(Score: 1–2, DVT probability = 12%), and 42% as “low risk”
(Score ≤ 0, DVT probability = 4%) [6]. Given these post-test prob-
abilities, the Wells rule cannot be employed as a stand-alone test to
confirm or exclude DVT; thus, it is commonly associated with D-dimer
or ultrasonography [5,9,10]. If neither D-dimer nor ultrasonography is
readily available the Wells rule may be used to stratify patients, al-
lowing for delayed testing in low- and moderate-risk patients, who may
be safely and quickly discharged [14].
The value of the Wells DVT rule in both the primary care and the
inpatient setting, as well as in patients with suspected isolated distal
DVT, is disputed [6,13,15–18]. Particularly, in a study of primary care
patients with suspected DVT, the safety of the Wells rule was chal-
lenged, being the probability of DVT as high as 12% in the low-risk
group, as compared with 3% in the original Wells study; furthermore,
despite the combination of a low score with a normal D-dimer, the
observed DVT incidence crossed the standard 2% safety margin (2.3%,
95% CI, 1.9 to 2.7) [16]. Consequently, a group of 110 Dutch primary
care practices proposed a different rule (the primary care, or Oudega
rule, Table 2), combining clinical items with point-of-care qualitative D-
dimer testing, for the exclusion of DVT in that setting [19]. In the
original derivation study, the rule categorized 21% of the patients as
“high risk” (Score: 10–13, DVT probability = 51.3%), 51% as “mod-
erate risk” (Score: 7–9, DVT probability = 21.7%), 5% as “low risk”
(Score 5–6, DVT probability = 4.5%), and 23% as “very low risk”
(Score 0–3, DVT probability = 0.7%) [19]. These findings were
Thrombosis Research xxx (xxxx) xxx–xxx
Table 2
The primary care rule [19].
Scorea Diagnostic variables Points for the rulea
1 Male gender 1
Oral contraceptive use 1
1 Presence of malignancy 1
Recent surgery 1
1 Absence of leg trauma 1
Vein distension 1
1 Calf difference ≥ 3 cm 2
1 D-dimer abnormal 6
1
a
High risk: 10–13; moderate risk: 7–9; low risk: 5–6; very low risk: 0–3.
1 The post-test probability of DVT was reported to be: 51% in the high prob-
1 ability group, 22% in the moderate probability group, 4% in the low prob-
1 ability group, and < 1% in the very-low probability group, respectively [19].
−2
subsequently externally validated in independent cohorts (Table 3)
[20,21]. Furthermore, a recent head-to-head comparison study and a
meta-analysis found that both rules are similarly safe in that setting
(Table 3), the discrepancy being accounted for by the inclusion of a new
item (history of DVT) in the Wells rule [9,22]. Accordingly, primary
care guidelines for the management of patients with suspected DVT
endorse the use of the Wells rule [10].
In hospitalized patients, the Wells rule performs poorly, since the
probability of DVT in low-risk patients is disappointingly high, ranging
between 6% and 12% (Table 3) [15,18]. Similarly, in patients with
isolated distal DVT the rule displays unsatisfactory sensitivity, as the
probability of DVT in low-risk patients ranges between 8% and 14%
[13,15,17]. In addition, a recent meta-analysis challenged the useful-
ness of the Wells DVT rule in outpatients with malignancy, being the
failure rate almost 2-fold as compared with non-cancer patients, and the
efficiency lower than 10% [9]. Since D-dimer testing also possesses
limited accuracy in inpatients, in patients with isolated distal DVT, as
well as in those with cancer, it is feasible that such patients would be
better off managed on the basis of ultrasonography alone [6,17].
Finally, experienced physicians or nurses may formulate quite ac-
curate estimates of pretest probability, employing implicit or “gestalt”
clinical judgment [13]; however, the gestalt approach obviously lacks
reproducibility [6].
In conclusion, the Wells DVT rule is useful to stratify symptomatic
outpatients for subsequent testing; namely: D-dimer if the probability
level is either unlikely or non-high, or ultrasonography, if either a likely
or high pretest probability is assigned. Hospitalized patients, as well as
outpatients with malignancy, are probably better managed on the basis
of ultrasonography.
1.2. D-dimers
D-dimers are specific cross-linked derivatives of fibrin, produced
when fibrin is degraded by plasmin, so concentrations are raised in
patients with venous thrombosis [23]. Numerous other conditions, such
as older age, cancer, infection, inflammation, ischemic heart disease,
stroke, peripheral artery disease, ruptured aneurysm or aortic dissec-
tion, pregnancy, and recent trauma or surgery yield increased D-dimer
levels, limiting the efficiency of D-dimer-based approaches [9,24,25].
In particular, it is noteworthy that normal D-dimer levels may be found
in only 56% of healthy subjects with ≥70 years, as compared
to > 90% of the general population under 50 years [26]. An age-de-
pendent D-dimer cut-off (age × 10 mcg/L) has been evaluated in pa-
tients with suspected DVT and either an unlikely, or a non-high Wells
score, [27,28]. The age-dependent cut-off is used in patients with >
50 years instead of the conventional 500 mcg/L cut-off, doubling the
number of patients with ≥ 80 years in whom DVT can be excluded,
with acceptable safety (Table 3) [27,28]. A large, prospective, multi-
centre study (ADJUST-DVT) testing the safety of withholding treatment
2
E. Bernardi, G. Camporese Thrombosis Research xxx (xxxx) xxx–xxx

Table 3
Performance of different strategies for the exclusion of deep-vein thrombosis in symptomatic patients.

Author Design N Setting Strategy to exclude DVT Safetya Efficiencya

Geersing 2014 [9] Meta-analysis 3.114b Primary care Wells ≤ 1 + normal D-dimer 1.2 (0.7 to 1.8) 29 (20 to 40)
Buller 2009 [21] Prospective 1.002 Primary care Oudega ≤ 3 1.4 (0.6 to 2.9) 49 (46 to 52)
Van der Velde 2011 [22] Prospective 495b Primary care Oudega ≤ 3 1.4 (0.6 to 3.0) 57 (54 to 61)
447b Wells ≤ 1 + normal D-dimer 1.6 (0.7 to 3.3) 70 (66 to 73)
Schouten 2012 [28] Retrospective 2.086 Primary care, all patients Wells ≤ 1 + normal D-dimerc 0.5 (0.01 to 1.3) 48 (44 to 52)
Age ≥ 80 years 0 (0.0 to 15.0) 35 (24 to 49)
Schutgens 2003 [40] Prospective 812 Wells ≤ 2 + normal D-dimer 0.6 (0.02 to 3.1) 22 (19 to 25)
Geersing 2014 [9] Meta-analysis 3225b Outpatients Wells ≤ 1 + normal quantitative D-dimer 0.4 (0.3 to 0.6) 21 (13 to 32)
4400b Wells ≤ 1 + normal qualitative D-dimer 0.6 (0.4 to 0.9) 46 (35 to 57)
Silveira 2015 [18] Prospective 1.135 Inpatients Wells ≤ 0 5.9 (3.0–11.3) 12 (10 to 14)
Cogo 2005 [44] Prospective 1.741 Outpatients Normal serial CUS 0.7 (0.3 to 1.2) NA
Bernardi 1998 [45] Prospective 946 Outpatients Normal serial CUS + normal D-dimer 0.4 (0 to 0.9) 72 (69 to 75)
Kraijenhagen 2002 [46] Prospective 1.739 Outpatients Normal serial CUS + normal D-dimer 0.7 (0.3 to 1.6) 48 (45 to 50)
Wells 1997 [47] Prospective 593 Outpatients Wells ≤ 0 + normal single CUS 0.3 (0 to 1.7) 54 (50 to 58)
Schellong 2003 [3] Prospective 1.646 Outpatients Normal WLUS 0.3 (0.1–0.8) NA
Subramaniam 2005 [44] Prospective 542 Outpatients Normal WLUS 0.2 (0.01 to 1.3) NA
Gottlieb 2003 [48] Randomized 261b Outpatients Selective WLUSd 0.8 (0.1 to 2.7) 42 (36 to 49)
Johnson 2010 [49] Meta-analysis 10.090 Outpatients with Wells ≥ 3 Normal WLUS 2.5 (0.9 to 6.1) NA
Stevens 2013 [50] Prospective 183 Outpatients with Wells ≥ 2 Normal WLUS 0.6 (0.02 to 3.3) NA
Ageno 2015 [53] Prospective 410b Outpatients Wells ≥ 2 + abnormal D-dimer + normal WLUS 1.5 (0.5 to 4.3) 51 (46 to 56)

DVT: deep-vein thrombosis; CUS: compression ultrasonography; WLUS: whole-leg ultrasonography; NA: not applicable.
a
Safety: proportion of patients left untreated on the basis of normal testing with the proposed strategy who develop thromboembolic events during follow-up Efficiency: proportion of
patients in whom DVT can be excluded on the basis of the proposed strategy. Expressed as: % (95% confidence interval).
b
Subset of the data is reported.
c
Age-dependent D-dimer cut-off: if > 50 years, cut-off = age × 10 mcg/L.
d
WLUS was only performed if calf symptoms/signs were present, otherwise patients underwent CUS.

in patients with clinically suspected DVT and a normal age-adjusted D- Table 4

dimer cut-off (age × 10 mcg/L) is currently on-going (NCT02384135). Randomized studies comparing CUS and WLUS in symptomatic patients with suspected
symptomatic deep-vein thrombosis.
Less frequently, patients with DVT may have false-negative D-dimer
results, such as those who had been treated with LMWH, or with leg Author Bernardi 2008 [52] Gibson 2009 [51]
symptoms lasting for > 2–3 weeks [29–31].
Numerous D-dimer assays have been validated against a gold stan- Pre-selection No Wells ≥ 2, or abnormal D-
dard (venography or ultrasonography) [30,32–38]. In general, quanti- dimer a

tative assays possess higher sensitivity and lower specificity while the Diagnostic strategies Serial CUS Single Serial CUS Single WLUS
contrary is true for qualitative, and point-of-care tests [6,32,38,39]. + D-dimer WLUS
Furthermore, D-dimer assays are less sensitive for isolated distal as
compared with proximal DVT, and for asymptomatic as compared with Patient n. 1045 1053 257 264
Initial DVT 23% 26% 23% 38%
symptomatic DVT [6,39]; finally, the overall performance of D-dimer
prevalence
assays decreases with increasing DVT prevalence [36]. VTE at follow-upb 0.9 (0.3 to 1.2 (0.5 to 2.0 (0.6 to 1.2 (0.2 to
Thus, D-dimer should not be used as a stand-alone test for to rule-in 1.8) 2.2) 5.1) 4.3)
b
or rule-out DVT [5,10]. Instead, there is sound clinical evidence that Absolute difference 0.3 (− 1.4 to 0.8) 0.8 (− 2.3 to 2.9)
suspected symptomatic DVT can be safely excluded in patients with a
DVT: deep-vein thrombosis; CUS: compression ultrasonography; WLUS: whole-leg ultra-
non-high, or an unlikely probability according to Wells, and normal D-
sonography; VTE: venous thromboembolism.
dimer results, both with a qualitative or a quantitative assay, without a
Patients with Wells DVT score of ≤1 did not undergo ultrasonography.
the need for further testing (Table 3) [40,41]. A recent meta-analysis of b
Expressed as: % (95% confidence interval).
13 studies, including > 10.000 patients, reported that in patients with
an unlikely probability according to the Wells' rule, DVT may be safely the common femoral vein at the groin and the popliteal vein in the
excluded on the basis of normal D-dimer results, irrespective if the popliteal fossa, including the so-called trifurcation (i.e.: the most
latter is either a qualitative point-of-care or a quantitative laboratory- proximal segments of the tibial and peroneal veins), are investigated
based assay (Table 3) [9]; instead, in patients with a likely or a high [42]. It is a simple, quick, bedside approach that does not require
probability level, ultrasound testing should be performed [5,10,12]. specialized operators or high-range technical equipment, widely im-
In conclusion, a normal D-dimer result in combination with an plemented in Emergency Departments, and hospital wards. Hence, CUS
unlikely, or non-high pretest probability according to Wells, allows for is usually available after-hours and during weekends. With WLUS, the
the exclusion of suspected symptomatic DVT. To the contrary, ab- whole deep-vein network of the leg is scanned in a continuous fashion,
normal D-dimer levels mandate further testing; namely, ultra- from the femoral veins, through the popliteal vein, to the distal veins. It
sonography. is usually performed by consultants in vascular medicine or vascular
surgery; thus, it is only available during office hours. Furthermore, it
1.3. Ultrasonography demands dedicated high-level machines and requires the patients to be
moved from the ED/ward to be evaluated [3].
Two distinct ultrasound approaches are available to rule in/out DVT Both approaches are based on the single diagnostic criterion of vein
in symptomatic patients (Table 4): compression ultrasonography (CUS; compressibility; broadly, if the veins are compressible, then DVT is
also, “2-point ultrasonography”, or “limited-US”), and whole-leg ul- ruled-out; if the veins are not compressible, then DVT is ruled-in [43].
trasonography (WLUS; also, “echo-color-Doppler”). Notably, while DVT can be safely excluded after a single normal WLUS
With CUS only two distinct spots of the proximal-vein system, i.e., [3,44], this is not the case with CUS. Indeed, as the distal veins are not

3
E. Bernardi, G. Camporese
investigated, repeat testing within one week is mandatory to identify
proximal extension of undetected distal DVT, the relative timing being
mandated by the clinical picture; that is, the patients' symptoms are
either stable or waning, then repeat CUS may be done after 7 days,
otherwise CUS should be performed as soon as possible (Table 3) [42].
Although safe, repeat testing of all patients with a normal baseline CUS
to detect only 3–5% extending distal DVT is inconvenient and expensive
[42]. Unnecessary repeat testing may be safely reduced (from 70 to
80% to 15–30%) by combining a single normal CUS test with either a
normal (quantitative, or qualitative) D-dimer, or with a low pretest
probability according to Wells (Table 3) [45–47]. On the other side,
selective use of WLUS, triggered by the presence of pain, redness or
swelling in the calf, was compared to routine WLUS in a randomized
prospective study (Table 3) [48]. A < 1% prevalence of events was
observed in both groups, and selective use of WLUS spared calf-vein
evaluation in > 40% of patients, statistically significantly reducing the
mean evaluation time (23 ± 14 min, vs 32 ± 14 min).
A meta-analysis of 7 studies, including > 10.000 patients, con-
firmed the safety of a single normal WLUS for the exclusion of DVT, but
reported surprisingly high (up to 3-fold the expected proportion) rates
of venous thromboembolism during follow-up in patients with a high
Wells score (Table 3) [49]. Conversely, both a recent small prospective
cohort study and a randomized clinical trial, investigating the safety of
withholding anticoagulation from patients with a likely or not-low
Wells score, reported a low rate of thromboembolic complication after a
single normal WLUS, ranging between 0.6 and 1.2% (Tables 3, 4)
[50,51].
Two randomized studies compared CUS and WLUS in symptomatic
outpatients with suspected DVT [51,52]. The first allocated all patients
to either serial CUS plus D-dimer, or a single (“one-shot”) WLUS; the
second randomized patients with a likely Wells score or an abnormal D-
dimer, to either serial CUS or one-shot WLUS. Both studies failed to
record statistically significant differences between the two ultrasound
approaches, despite a statistically significantly higher initial prevalence
of DVT in patients managed by WLUS, due to distal-vein testing
(Table 4). The authors concluded either strategy is safe to exclude
suspected symptomatic DVT; however, the debate on the optimal
management of isolated distal vein thrombosis is on-going [5].
4
Thrombosis Research xxx (xxxx) xxx–xxx
Fig. 1. Suggested algorithm for the evaluation of
suspected DVT in symptomatic patients. DVT:
deep-vein thrombosis CUS: compression ultra-
sonography WLUS: whole-leg ultrasonography. *
Suggested approach if the Wells DVT score is ≥ 3
(high probability).
A recent prospective, multicentre, cohort study devised a new se-
quential strategy, including the evaluation of pretest probability with
the Wells rule, D-dimer, CUS and WLUS, as follows [53]. All patients
underwent D-dimer testing; patients with unlikely probability and
normal D-dimer were spared further testing; those with either a likely
probability or abnormal D-dimer underwent a single CUS test; and
those with both a likely probability and abnormal D-dimer were eval-
uated with WLUS [53]. Only 35% of the patients required WLUS, and in
half (49%) of them DVT was confirmed, indicating a high dis-
criminatory power of the algorithm. The overall prevalence of throm-
boembolic events at follow-up in 950 untreated patients was < 1%,
confirming the safety of the sequential strategy; however, in high-risk
patients with both a likely Wells score and abnormal D-dimer, the upper
confidence limit for the prevalence was > 4% (Table 3).
The results reported by the latter study and by the cited meta-
analysis [49], although likely due to the small sample included, suggest
the hypothesis that a single WLUS might not be sufficiently safe in high-
risk patients. Interestingly, in a classic accuracy study of the Wells rule,
in which venography was planned for all cases of discrepancy between
pretest clinical probability and CUS, the negative predictive value of a
normal CUS was only 82% (95% CI, 59.7 to 94.8) in patients with high
pretest probability [47]. To our knowledge, a study evaluating WLUS
against venography in high-risk patients is not available from the lit-
erature, and it's very unlikely to be performed based on the con-
siderations stated in the introduction. None the less, we believe the
signal coming from this subgroup of high-risk patients might be further
investigated in a properly designed trial, perhaps using a combined
(instrumental and clinical) end-point.
In conclusion, venous ultrasonography allows for both safe con-
firmation and exclusion of suspected symptomatic DVT. Compression
ultrasonography is widely available, quick, but needs to be repeated in
all patients with normal baseline testing, unless combined with either a
low pretest probability or a normal D-dimer. Whole-leg ultra-
sonography allows for “one-shot” testing, but is rarely available outside
working-hours and requires highly skilled operators; furthermore, it
raises the point of isolated distal DVT, neglected by CUS, though with
equivalent safety. Never the less, investigating the whole leg allows also
prompt identification of other pathologic conditions (such as, muscle-
E. Bernardi, G. Camporese
tear, joint-effusion, erysipelas, phlebitis, arthrosis, arthritis, chronic
venous insufficiency, lymphatic stasis, etc.), in up to 50% of patients in
whom suspected DVT has already been ruled out [48,54,55].
2. Conclusions
Invasive imaging techniques such as venography, albeit formally
regarded as “gold standard” for the diagnosis of DVT, are infrequently
used – if at all, in clinical practice. Ultrasonography is the most vali-
dated and accurate non-invasive imaging method in symptomatic pa-
tients, and may be performed either in a simplified fashion (CUS), in
which only two proximal venous spots are scanned; or in an extended
mode (WLUS), in which both the proximal and the distal veins are in-
vestigated. Ultrasonography can be safely spared in patients with un-
likely pretest probability for DVT and normal D-dimer levels.
Otherwise, if CUS is performed, all patients with normal baseline
findings need to undergo either a repeat CUS within one week, unless
they are at low pretest probability, or have normal D-dimer results. If
WLUS is chosen, no further testing is foreseen in patients with normal
baseline findings. Fig. 1 illustrates a suggested algorithm for the eva-
luation of suspected DVT in symptomatic patients, based on the current
review.
Conflicts of interest
None.
Role of funding source
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
References
[1] M. Meignan, J. Rosso, H. Gauthier, et al., Systematic lung scans reveal a high fre-
quency of silent pulmonary embolism in patients with proximal deep venous
thrombosis, Arch. Intern. Med. 160 (2000) 159–164.
[2] J.D. Douketis, C. Kearon, S. Bates, E.K. Duku, J.S. Ginsberg, Risk of fatal pulmonary
embolism in patients with treated venous thromboembolism, JAMA 279 (1998)
458–462.
[3] S.M. Schellong, T. Schwarz, K. Halbritter, J. Beyer, G. Siegert, W. Oettler,
B. Schmidt, H.E. Schroeder, Complete compression ultrasonography of the leg veins
as a single test for the diagnosis of deep vein thrombosis, Thromb. Haemost. 89 (2)
(2003) 228–234.
[4] M. Righini, H. Bounameaux, Clinical relevance of distal deep vein thrombosis, Curr.
Opin. Pulm. Med. 14 (5) (2008) 408–413.
[5] S.M. Bates, R. Jaeschke, S.M. Stevens, S. Goodacre, P.S. Wells, M.D. Stevenson,
C. Kearon, H.J. Schunemann, M. Crowther, S.G. Pauker, R. Makdissi, G.H. Guyatt,
American College of Chest Physicians, Diagnosis of DVT: antithrombotic therapy
and prevention of thrombosis, American College of Chest Physicians Evidence-
based Clinical Practice Guidelines, 9th ed, Chest, 141(2 Suppl) 2012, pp.
e351S–418S, , http://dx.doi.org/10.1378/chest.11-2299.
[6] S. Goodacre, F. Sampson, M. Stevenson, A. Wailoo, A. Sutton, S. Thomas, et al.,
Measurement of the clinical and cost-effectiveness of non-invasive diagnostic
testing strategies for deep vein thrombosis, Health Technol. Assess. 10 (15) (2006).
[7] A.J. Ten Cate-Hoek, M.H. Prins, Management studies using a combination of D-
dimer test result and clinical probability to rule out venous thromboembolism: a
systematic review, J. Thromb. Haemost. 3 (2005) 2465–2470.
[8] R. Hull, J. Hirsh, D.L. Sackett, D.W. Taylor, C. Carter, A.G. Turpie, P. Powers,
M. Gent, Clinical validity of a negative venogram in patients with clinically sus-
pected venous thrombosis, Circulation 64 (3) (1981) 622–625.
[9] G.J. Geersing, N.P.A. Zuithoff, C. Kearon, D.R. Anderson, A.J. ten Cate-Hoek,
J.L. Elf, S.M. Bates, A.W. Hoes, R.A. Kraaijenhagen, R. Oudega, R.E.G. Schutgens,
S.M. Stevens, S.C. Woller, P.S. Wells, Moons KGM, Exclusion of deep vein throm-
bosis using the Wells rule in clinically important subgroups: individual patient data
meta-analysis, BMJ 348 (2014) g1340, , http://dx.doi.org/10.1136/bmj.g1340.
[10] A. Quaseem, W. Snow, P. Barry, et al., Current diagnosis of venous thromboem-
bolism in primary care: a clinical practice guideline from the American Academy of
family physicians and the American College of Physicians, Ann. Fam. Med. 5 (2007)
57–62, http://dx.doi.org/10.1370/afm.667.
[11] S.M. Stevens, W. Ageno, Review: the wells rule is more useful than individual
clinical features for predicting risk of deep venous thrombosis, Evid. Based Med. 11
(2) (2006) 56.
[12] P.S. Wells, C. Owen, S. Doucette, D. Fergusson, H. Tran, Does this patient have
deep-vein thrombosis? JAMA 295 (2006) 199–207.
5
Thrombosis Research xxx (xxxx) xxx–xxx
[13] S. Goodacre, A.J. Sutton, F.C. Sampson, Meta-analysis: the value of clinical as-
sessment in the diagnosis of deep venous thrombosis, Ann. Intern. Med. 143 (2005)
129–139.
[14] D.R. Anderson, P.S. Wells, I. Stiell, B. MacLeod, M. Simms, L. Gray, K.S. Robinson,
J. Bormanis, M. Mitchell, B. Lewandowski, G. Flowerdew, Thrombosis in the
emergency department: use of a clinical diagnosis model to safely avoid the need for
urgent radiological investigation, Arch. Intern. Med. 159 (5) (1999) 477–482.
[15] R.P. Engelberger, D. Aujesky, L. Calanca, P. Staeger, O. Hugli, L. Mazzolai,
Comparison of the diagnostic performance of the original and modified Wells score
in inpatients and outpatients with suspected deep vein thrombosis, Thromb. Res.
127 (2011) 535–539, http://dx.doi.org/10.1016/j.thromres.2011.02.008.
[16] R. Oudega, A.W. Hoes, K.G. Moons, The Wells rule does not adequately rule out
deep venous thrombosis in primary care patients, Ann. Intern. Med. 143 (2005)
100–107.
[17] M. Sartori, B. Cosmi, C. Legnani, E. Favaretto, L. Valdré, G. Guazzaloca,
G. Rodorigo, M. Cini, G. Palareti, The Wells rule and D-dimer for the diagnosis of
isolated distal deep vein thrombosis, J. Thromb. Haemost. 10 (11) (2012)
2264–2269, http://dx.doi.org/10.1111/j.1538-7836.2012.04895.
[18] P.C. Silveira, I.K. Ip, S.Z. Goldhaber, G. Piazza, C.B. Benson, R. Khorasani,
Performance of Wells score for deep vein thrombosis in the inpatient setting, JAMA
Intern. Med. 175 (7) (2015) 1112–1117, http://dx.doi.org/10.1001/
jamainternmed.2015.1687.
[19] R. Oudega, K.G. Moons, A.W. Hoes, Ruling out deep venous thrombosis in primary
care. A simple diagnostic algorithm including D-dimer testing, Thromb. Haemost.
94 (2005) 200–205.
[20] D.B. Toll, R. Oudega, R.J. Bulten, A.W. Hoes, K.G. Moons, Excluding deep vein
thrombosis safely in primary care, J. Fam. Pract. 55 (2006) 613–618.
[21] H.R. Buller, A.J. Ten Cate-Hoek, A.W. Hoes, M.A. Joore, K.G. Moons, R. Oudega,
et al., Safely ruling out deep venous thrombosis in primary care, Ann. Intern. Med.
150 (2009) 229–235.
[22] E.F. Van der Velde, D.B. Toll, A.J. Ten Cate-Hoek, R. Oudega, H.E. Stoffers,
P.M. Bossuyt, et al., Comparing the diagnostic performance of 2 clinical decision
rules to rule out deep vein thrombosis in primary care patients, Ann. Fam. Med. 9
(2011) 31–36.
[23] J. Kelly, A. Rudd, R.R. Lewis, B.J. Hunt, Plasma D-dimers in the diagnosis of venous
thromboembolism, Arch. Intern. Med. 162 (2002) 747–756.
[24] P. Raimondi, O. Bongard, P. de Moerloose, G. Reber, F. Waldvogel, H. Bounameaux,
D-Dimer plasma concentration in various clinical conditions: implication for the use
of this test in the diagnostic approach of venous thromboembolism, Thromb. Res.
69 (1993) 125–130.
[25] G. Sodeck, H. Domanovits, M. Schillinger, M.P. Ehrlich, G. Endler, H. Herkner,
A. Laggner, D-dimer in ruling out acute aortic dissection: a systematic review and
prospective cohort study, Eur. Heart J. 28 (2007) 3067–3075.
[26] C. Haase, M. Joergensen, C. Ellervik, M.K. Joergensen, L. Bathum, Age- and sex-
dependent reference intervals for D-dimer: evidence for a marked increase by age,
Thromb. Res. 132 (6) (2013) 676–680, http://dx.doi.org/10.1016/j.thromres.2013.
09.033.
[27] R.A. Douma, M. Tan, R.E.G. Schutgens, S.M. Bates, A. Perrier, C. Legnani,
D.H. Biesma, J.A. Ginsberg, H. Bounameaux, G. Palareti, M. Carrier, G.C. Mol, G. Le
Gal, P.W. Kamphuisen, M. Righini, Using an age-dependent D-dimer cut-off value
increases the number of older patients in whom deep vein thrombosis can be safely
excluded, Haematologica 97 (10) (2012) 1507–1513, http://dx.doi.org/10.3324/
haematol.2011.060657.
[28] H.J. Schouten, H.L. Koek, R. Oudega, G.J. Geersing, K.J. Janssen, J.J. van Delden,
K.G. Moons, Validation of two age dependent D-dimer cut-off values for exclusion of
deep vein thrombosis in suspected elderly patients in primary care: retrospective,
cross sectional, diagnostic analysis, BMJ 344 (2012) e2985, , http://dx.doi.org/10.
1136/bmj.e2985.
[29] P.M. Baker, D.M.A. Keeling, Single dose of Low-molecular-weight Heparin (LMWH)
invalidates the use of D-dimer as part of a Deep Vein Thrombosis (DVT) diagnostic
algorithm, Int. J. Lab. Hematol. 39 (1) (2017) e17–e18, http://dx.doi.org/10.1111/
ijlh.12579.
[30] F. van der Graaf, H. van den Borne, M. van der Kolk, P.J. de Wild, G.W. Janssen,
S.H. van Uum, Exclusion of deep venous thrombosis with D-dimer testing: com-
parison of 13 D-dimer methods in 99 outpatients suspected of deep venous
thrombosis using venography as reference standard, Thromb. Haemost. 83 (2)
(2000) 191–198.
[31] R. Anastasio, L. Quartero, C. POrta, F. FAlaschi, S. Baratto, M.A. Bressan,
S. Siragusa, Diagnostic accuracy of D-dimer test is affected by thrombus site,
duration of symptoms, and heparin treatment, J. Thromb. Haemost. (Suppl. 1)
(2001) 748.
[32] P.D. Stein, R.D. Hull, K.C. Patel, R.E. Olson, W.A. Ghali, R. Brant, R.K. Biel,
V. Bharadia, N.K. Kalra, D-dimer for the exclusion of acute venous thrombosis and
pulmonary embolism: a systematic review, Ann. Intern. Med. 140 (8) (2004)
589–602.
[33] G.J. Geersing, K.J. Janssen, R. Oudega, L. Bax, A.W. Hoes, J.B. Reitsma,
K.G. Moons, Excluding venous thromboembolism using point of care D-dimer tests
in outpatients: a diagnostic meta-analysis, BMJ 339 (2009) b2990, , http://dx.doi.
org/10.1136/bmj.b2990.
[34] D. Neale, C. Tovey, A. Vali, S. Davies, K. Myers, M. Obiako, V. Ramkumar, A. Hafiz,
Evaluation of the simplify D-dimer assay as a screening test for the diagnosis of deep
vein thrombosis in an emergency department, Emerg. Med. J. 21 (6) (2004)
663–666.
[35] K. Boeer, R. Siegmund, D. Schmidt, T. Deufel, M. Kiehntopf, Comparison of six D-
dimer assays for the detection of clinically suspected deep venous thrombosis of the
lower extremities, Blood Coagul. Fibrinolysis 20 (2) (2009) 141–145, http://dx.doi.
E. Bernardi, G. Camporese
org/10.1097/MBC.0b013e3283255381.
[36] S.W. Heim, J.M. Schectman, M.S. Siadaty, J.T. Philbrick, D-dimer testing for deep
venous thrombosis: a meta-analysis, Clin. Chem. 50 (2004) 1136–1147.
[37] G.J. Geersing, D.B. Toll, K.J. Janssen, R. Oudega, M.J. Blikman, R. Wijland, K.M. de
Vooght, A.W. Hoes, K.G. Moons, Diagnostic accuracy and user-friendliness of 5
point-of-care D-dimer tests for the exclusion of deep vein thrombosis, Clin. Chem.
56 (11) (2010) 1758–1766, http://dx.doi.org/10.1373/clinchem.2010.147892.
[38] M. Righini, A. Perrier, P. De Moerloose, H. Bounameaux, D-dimer for venous
thromboembolism diagnosis: 20 years later, J. Thromb. Haemost. 6 (7) (2008)
1059–1071, http://dx.doi.org/10.1111/j.1538-7836.2008.02981.x.
[39] S. Goodacre, F.C. Sampson, A.J. Sutton, S. Mason, F. Morris, Variation in the di-
agnostic performance of D-dimer for suspected deep vein thrombosis, QJM 98 (7)
(2005) 513–527.
[40] R.E. Schutgens, P. Ackermark, F.J. Haas, H.K. Nieuwenhuis, H.G. Peltenburg,
A.H. Pijlman, M. Pruijm, R. Oltmans, J.C. Kelder, D.H. Biesma, Combination of a
normal D-dimer concentration and a non-high pretest clinical probability score is a
safe strategy to exclude deep venous thrombosis, Circulation 107 (4) (2003)
593–597.
[41] P.S. Wells, D.R. Anderson, M. Rodger, M. Forgie, C. Kearon, J. Dreyer, G. Kovacs,
M. Mitchell, B. Lewandowski, M.J. Kovacs, Evaluation of D-dimer in the diagnosis
of suspected deep-vein thrombosis, N. Engl. J. Med. 349 (13) (2003) 1227–1235.
[42] A. Cogo, A.W. Lensing, M.M. Koopman, F. Piovella, S. Siragusa, P.S. Wells,
S. Villalta, H.R. Büller, A.G. Turpie, P. Prandoni, Compression ultrasonography for
diagnostic management of patients with clinically suspected deep vein thrombosis:
prospective cohort study, BMJ 316 (7124) (1998) 17–20.
[43] J.J. Cronan, G.S. Dorfman, F.H. Scola, B. Schepps, J. Alexander, Deep venous
thrombosis: US assessment using vein compression, Radiology 162 (1 Pt1) (1987)
191–194.
[44] R.M. Subramaniam, R. Heath, T. Chou, K. Cox, G. Davis, M. Swarbrick, Deep venous
thrombosis: withholding anticoagulation therapy after negative complete lower
limb US findings, Radiology 237 (1) (2005) 348–352.
[45] E. Bernardi, P. Prandoni, A.W. Lensing, G. Agnelli, G. Guazzaloca, G. Scannapieco,
F. Piovella, F. Verlato, C. Tomasi, M. Moia, L. Scarano, A. Girolami, D-dimer testing
as an adjunct to ultrasonography in patients with clinically suspected deep vein
thrombosis: prospective cohort study. The Multicentre Italian D-dimer Ultrasound
Study Investigators Group, BMJ 317 (7165) (1998) 1037–1040.
[46] R.A. Kraaijenhagen, F. Piovella, E. Bernardi, F. Verlato, E.A. Beckers,
M.M. Koopman, M. Barone, G. Camporese, B.J. Potter Van Loon, M.H. Prins,
P. Prandoni, H.R. Büller, Simplification of the diagnostic management of suspected
deep vein thrombosis, Arch. Intern. Med. 162 (8) (2002) 907–911.
[47] P.S. Wells, D.R. Anderson, J. Bormanis, F. Guy, M. Mitchell, L. Gray, C. Clement,
Thrombosis Research xxx (xxxx) xxx–xxx
K.S. Robinson, B. Lewandowski, Value of assessment of pretest probability of deep-
vein thrombosis in clinical management, Lancet 350 (9094) (1997) 1795–1798.
[48] R.H. Gottlieb, S.L. Voci, L. Syed, C. Shyu, P.J. Fultz, D.J. Rubens, J.G. Strang,
N. Carson, W.J. DiGrazio, C.W. Francis, Randomized prospective study comparing
routine versus selective use of sonography of the complete calf in patients with
suspected deep venous thrombosis, Am. J. Roentgenol. 180 (1) (2003) 241–245.
[49] S.A. Johnson, S.M. Stevens, S.C. Woller, E. Lake, M. Donadini, J. Cheng, J. Labarère,
J.D. Douketis, Risk of deep vein thrombosis following a single negative whole-leg
compression ultrasound: a systematic review and meta-analysis, JAMA 303 (5)
(2010) 438–445, http://dx.doi.org/10.1001/jama.2010.43.
[50] S.M. Stevens, S.C. Woller, K.K. Graves, V. Aston, J. Jones, G. Snow, C.G. Elliott,
Withholding anticoagulation following a single negative whole-leg ultrasound in
patients at high pretest probability for deep vein thrombosis, Clin. Appl. Thromb.
Hemost. 19 (1) (2013) 79–85, http://dx.doi.org/10.1177/1076029612445919.
[51] N.S. Gibson, S.M. Schellong, D.Y. Kheir, J. Beyer-Westendorf, A.S. Gallus, S. McRae,
R.E. Schutgens, F. Piovella, V.E. Gerdes, H.R. Buller, Safety and sensitivity of two
ultrasound strategies in patients with clinically suspected deep venous thrombosis:
a prospective management study, J. Thromb. Haemost. 7 (12) (2009) 2035–2041,
http://dx.doi.org/10.1111/j.1538-7836.2009.03635.x.
[52] E. Bernardi, G. Camporese, H.R. Büller, S. Siragusa, D. Imberti, A. Berchio,
A. Ghirarduzzi, F. Verlato, R. Anastasio, C. Prati, A. Piccioli, R. Pesavento, C. Bova,
P. Maltempi, N. Zanatta, A. Cogo, R. Cappelli, E. Bucherini, S. Cuppini, F. Noventa,
P. Prandoni, Erasmus Study Group, Serial 2-point ultrasonography plus D-dimer vs
whole-leg color-coded Doppler ultrasonography for diagnosing suspected sympto-
matic deep vein thrombosis: a randomized controlled trial, JAMA 300 (14) (2008)
1653–1659, http://dx.doi.org/10.1001/jama.300.14.1653.
[53] W. Ageno, G. Camporese, N. Riva, M. Iotti, E. Bucherini, M. Righini,
P.W. Kamphuisen, P. Verhamme, J.D. Douketis, C. Tonello, P. Prandoni,
P.A.L.L.A.D.I.O. Study Investigators, Analysis of an algorithm incorporating limited
and whole-leg assessment of the deep venous system in symptomatic outpatients
with suspected deep-vein thrombosis (PALLADIO): a prospective, multicentre, co-
hort study, Lancet Haematol. 2 (11) (2015) e474–80, http://dx.doi.org/10.1016/
S2352-3026(15)00190-8.
[54] G. Camporese, E. Bernardi, L. Scarano, A. Ghirarduzzi, D. Imberti, P. Prandoni, on
behalf of The PALIST Study Group, Outcome of patients with suspected lower limb
symptomatic deep vein thrombosis and a normal ultrasound-based initial diagnostic
workup: a prospective study, J. Thromb. Haemost. 10 (2012) 2605–2606.
[55] M.E. Sutter, S.D. Turnipseed, D.B. Diercks, P. Samuel, R.H. White, Venous ultra-
sound testing for suspected thrombosis: incidence of significant non-thrombotic
findings, J. Emerg. Med. 36 (1) (2009) 55–59, http://dx.doi.org/10.1016/j.
jemermed.2007.08.066.