- Email: [email protected]
Diagnosis of deep-vein thrombosis
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 b
Emergency Unit, Department of Critical Care, aULSS2 “Marca Trevigana”, distretto di Pieve di Soligo, via Brigata Bisagno, 4, 31015 Conegliano, Treviso, Italy 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: Venous thrombosis Diagnosis Clinical decision rule Fibrin fibrinogen degradation products Ultrasonography Doppler Color Recurrence
The diagnostic approach to suspected symptomatic deep-vein thrombosis of the lower extremities is usually based on non-invasive methods, including the estimation of clinical probability, the measurement of D-dimer levels, and ultrasonography. The present review discusses the evidence available from the literature about the management of the first episode of suspected deep-vein-thrombosis.
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 popliteal 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 proximal 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 prophylaxis 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 prevalence 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 insufficiency, and difficult to interpret, with considerable inter- and intraobserver 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 concerning those different strategies in the following sections. Noteworthy, only ultrasonography may be used a stand-alone test to rule-in or ruleout 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: [email protected] (E. Bernardi), [email protected] (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
Thrombosis Research xxx (xxxx) xxx–xxx
E. Bernardi, G. Camporese
Table 1 The modified Wells DVT rule [12].
Table 2 The primary care rule [19].
Clinical variable
Scorea
Diagnostic variables
Points for the rulea
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
1
Male gender Oral contraceptive use Presence of malignancy Recent surgery Absence of leg trauma Vein distension Calf difference ≥ 3 cm D-dimer abnormal
1 1 1 1 1 1 2 6
1 1 1 1 1
a High risk: 10–13; moderate risk: 7–9; low risk: 5–6; very low risk: 0–3. The post-test probability of DVT was reported to be: 51% in the high probability group, 22% in the moderate probability group, 4% in the low probability group, and < 1% in the very-low probability group, respectively [19].
1 1 1 −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 usefulness 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 accurate 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.
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 probability group, respectively [12]. Alternatively, pretest probability scores may be categorized 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 suspected 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, intermediate, high) set-up, is the Well's DVT rule (Table 1) [7,12,13]. According to a meta-analysis of 21 studies, assuming a 15% DVT prevalence, 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 probabilities, 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, allowing 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 challenged, 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 Ddimer 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 “moderate 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
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 dissection, 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-dependent D-dimer cut-off (age × 10 mcg/L) has been evaluated in patients 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, multicentre study (ADJUST-DVT) testing the safety of withholding treatment 2
Thrombosis Research xxx (xxxx) xxx–xxx
E. Bernardi, G. Camporese
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] Buller 2009 [21] Van der Velde 2011 [22]
Meta-analysis Prospective Prospective
Primary care Primary care Primary care
Schouten 2012 [28]
Retrospective
3.114b 1.002 495b 447b 2.086
Wells ≤ 1 + normal D-dimer Oudega ≤ 3 Oudega ≤ 3 Wells ≤ 1 + normal D-dimer Wells ≤ 1 + normal D-dimerc
Schutgens 2003 [40] Geersing 2014 [9]
Prospective Meta-analysis
Silveira 2015 [18] Cogo 2005 [44] Bernardi 1998 [45] Kraijenhagen 2002 [46] Wells 1997 [47] Schellong 2003 [3] Subramaniam 2005 [44] Gottlieb 2003 [48] Johnson 2010 [49] Stevens 2013 [50] Ageno 2015 [53]
Prospective Prospective Prospective Prospective Prospective Prospective Prospective Randomized Meta-analysis Prospective Prospective
1.2 (0.7 to 1.8) 1.4 (0.6 to 2.9) 1.4 (0.6 to 3.0) 1.6 (0.7 to 3.3) 0.5 (0.01 to 1.3) 0 (0.0 to 15.0) 0.6 (0.02 to 3.1) 0.4 (0.3 to 0.6) 0.6 (0.4 to 0.9) 5.9 (3.0–11.3) 0.7 (0.3 to 1.2) 0.4 (0 to 0.9) 0.7 (0.3 to 1.6) 0.3 (0 to 1.7) 0.3 (0.1–0.8) 0.2 (0.01 to 1.3) 0.8 (0.1 to 2.7) 2.5 (0.9 to 6.1) 0.6 (0.02 to 3.3) 1.5 (0.5 to 4.3)
29 (20 49 (46 57 (54 70 (66 48 (44 35 (24 22 (19 21 (13 46 (35 12 (10 NA 72 (69 48 (45 54 (50 NA NA 42 (36 NA NA 51 (46
812 3225b 4400b 1.135 1.741 946 1.739 593 1.646 542 261b 10.090 183 410b
Primary care, all patients Age ≥ 80 years
Wells ≤ 2 + normal D-dimer Wells ≤ 1 + normal quantitative D-dimer Wells ≤ 1 + normal qualitative D-dimer Wells ≤ 0 Normal serial CUS Normal serial CUS + normal D-dimer Normal serial CUS + normal D-dimer Wells ≤ 0 + normal single CUS Normal WLUS Normal WLUS Selective WLUSd Normal WLUS Normal WLUS Wells ≥ 2 + abnormal D-dimer + normal WLUS
Outpatients Inpatients Outpatients Outpatients Outpatients Outpatients Outpatients Outpatients Outpatients Outpatients with Wells ≥ 3 Outpatients with Wells ≥ 2 Outpatients
to to to to to to to to to to
40) 52) 61) 73) 52) 49) 25) 32) 57) 14)
to 75) to 50) to 58)
to 49)
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 Ddimer cut-off (age × 10 mcg/L) is currently on-going (NCT02384135). Less frequently, patients with DVT may have false-negative D-dimer results, such as those who had been treated with LMWH, or with leg symptoms lasting for > 2–3 weeks [29–31]. Numerous D-dimer assays have been validated against a gold standard (venography or ultrasonography) [30,32–38]. In general, quantitative assays possess higher sensitivity and lower specificity while the contrary is true for qualitative, and point-of-care tests [6,32,38,39]. Furthermore, D-dimer assays are less sensitive for isolated distal as compared with proximal DVT, and for asymptomatic as compared with symptomatic DVT [6,39]; finally, the overall performance of D-dimer assays decreases with increasing DVT prevalence [36]. Thus, D-dimer should not be used as a stand-alone test for to rule-in or rule-out DVT [5,10]. Instead, there is sound clinical evidence that suspected symptomatic DVT can be safely excluded in patients with a non-high, or an unlikely probability according to Wells, and normal Ddimer results, both with a qualitative or a quantitative assay, without the need for further testing (Table 3) [40,41]. A recent meta-analysis of 13 studies, including > 10.000 patients, reported that in patients with an unlikely probability according to the Wells' rule, DVT may be safely excluded on the basis of normal D-dimer results, irrespective if the latter is either a qualitative point-of-care or a quantitative laboratorybased assay (Table 3) [9]; instead, in patients with a likely or a high probability level, ultrasound testing should be performed [5,10,12]. In conclusion, a normal D-dimer result in combination with an unlikely, or non-high pretest probability according to Wells, allows for the exclusion of suspected symptomatic DVT. To the contrary, abnormal D-dimer levels mandate further testing; namely, ultrasonography.
Table 4 Randomized studies comparing CUS and WLUS in symptomatic patients with suspected symptomatic deep-vein thrombosis. Author
Bernardi 2008 [52]
Gibson 2009 [51]
Pre-selection
No
Wells ≥ 2, or abnormal Ddimer a
Diagnostic strategies
Serial CUS + D-dimer
Single WLUS
Serial CUS
Single WLUS
Patient n. Initial DVT prevalence VTE at follow-upb
1045 23%
1053 26%
257 23%
264 38%
b
Absolute difference
0.9 (0.3 to 1.2 (0.5 to 1.8) 2.2) 0.3 (− 1.4 to 0.8)
2.0 (0.6 to 1.2 (0.2 to 5.1) 4.3) 0.8 (− 2.3 to 2.9)
DVT: deep-vein thrombosis; CUS: compression ultrasonography; WLUS: whole-leg ultrasonography; VTE: venous thromboembolism. a Patients with Wells DVT score of ≤1 did not undergo ultrasonography. b Expressed as: % (95% confidence interval).
the common femoral vein at the groin and the popliteal vein in the popliteal fossa, including the so-called trifurcation (i.e.: the most proximal segments of the tibial and peroneal veins), are investigated [42]. It is a simple, quick, bedside approach that does not require specialized operators or high-range technical equipment, widely implemented in Emergency Departments, and hospital wards. Hence, CUS is usually available after-hours and during weekends. With WLUS, the whole deep-vein network of the leg is scanned in a continuous fashion, from the femoral veins, through the popliteal vein, to the distal veins. It is usually performed by consultants in vascular medicine or vascular surgery; thus, it is only available during office hours. Furthermore, it demands dedicated high-level machines and requires the patients to be moved from the ED/ward to be evaluated [3]. Both approaches are based on the single diagnostic criterion of vein compressibility; broadly, if the veins are compressible, then DVT is ruled-out; if the veins are not compressible, then DVT is ruled-in [43]. Notably, while DVT can be safely excluded after a single normal WLUS [3,44], this is not the case with CUS. Indeed, as the distal veins are not
1.3. Ultrasonography Two distinct ultrasound approaches are available to rule in/out DVT in symptomatic patients (Table 4): compression ultrasonography (CUS; also, “2-point ultrasonography”, or “limited-US”), and whole-leg ultrasonography (WLUS; also, “echo-color-Doppler”). With CUS only two distinct spots of the proximal-vein system, i.e., 3
Thrombosis Research xxx (xxxx) xxx–xxx
E. Bernardi, G. Camporese
Fig. 1. Suggested algorithm for the evaluation of suspected DVT in symptomatic patients. DVT: deep-vein thrombosis CUS: compression ultrasonography WLUS: whole-leg ultrasonography. * Suggested approach if the Wells DVT score is ≥ 3 (high probability).
A recent prospective, multicentre, cohort study devised a new sequential 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 evaluated with WLUS [53]. Only 35% of the patients required WLUS, and in half (49%) of them DVT was confirmed, indicating a high discriminatory power of the algorithm. The overall prevalence of thromboembolic 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 metaanalysis [49], although likely due to the small sample included, suggest the hypothesis that a single WLUS might not be sufficiently safe in highrisk 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 literature, and it's very unlikely to be performed based on the considerations 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 confirmation 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 ultrasonography 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-
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, confirmed 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 Ddimer, 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
E. Bernardi, G. Camporese
[13] S. Goodacre, A.J. Sutton, F.C. Sampson, Meta-analysis: the value of clinical assessment 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 sexdependent 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: comparison 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 Ddimer 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.
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 validated and accurate non-invasive imaging method in symptomatic patients, 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 investigated. Ultrasonography can be safely spared in patients with unlikely 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 evaluation 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 frequency 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 Evidencebased 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 Ddimer 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 suspected 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 thrombosis 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 thromboembolism 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
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 diagnostic 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,
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
6
K.S. Robinson, B. Lewandowski, Value of assessment of pretest probability of deepvein thrombosis in clinical management, Lancet 350 (9094) (1997) 1795–1798. 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. 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. 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. 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. 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 symptomatic deep vein thrombosis: a randomized controlled trial, JAMA 300 (14) (2008) 1653–1659, http://dx.doi.org/10.1001/jama.300.14.1653. 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, cohort study, Lancet Haematol. 2 (11) (2015) e474–80, http://dx.doi.org/10.1016/ S2352-3026(15)00190-8. 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. M.E. Sutter, S.D. Turnipseed, D.B. Diercks, P. Samuel, R.H. White, Venous ultrasound 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.
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 b
Emergency Unit, Department of Critical Care, aULSS2 “Marca Trevigana”, distretto di Pieve di Soligo, via Brigata Bisagno, 4, 31015 Conegliano, Treviso, Italy 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: Venous thrombosis Diagnosis Clinical decision rule Fibrin fibrinogen degradation products Ultrasonography Doppler Color Recurrence
The diagnostic approach to suspected symptomatic deep-vein thrombosis of the lower extremities is usually based on non-invasive methods, including the estimation of clinical probability, the measurement of D-dimer levels, and ultrasonography. The present review discusses the evidence available from the literature about the management of the first episode of suspected deep-vein-thrombosis.
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 popliteal 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 proximal 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 prophylaxis 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 prevalence 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 insufficiency, and difficult to interpret, with considerable inter- and intraobserver 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 concerning those different strategies in the following sections. Noteworthy, only ultrasonography may be used a stand-alone test to rule-in or ruleout 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: [email protected] (E. Bernardi), [email protected] (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
Thrombosis Research xxx (xxxx) xxx–xxx
E. Bernardi, G. Camporese
Table 1 The modified Wells DVT rule [12].
Table 2 The primary care rule [19].
Clinical variable
Scorea
Diagnostic variables
Points for the rulea
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
1
Male gender Oral contraceptive use Presence of malignancy Recent surgery Absence of leg trauma Vein distension Calf difference ≥ 3 cm D-dimer abnormal
1 1 1 1 1 1 2 6
1 1 1 1 1
a High risk: 10–13; moderate risk: 7–9; low risk: 5–6; very low risk: 0–3. The post-test probability of DVT was reported to be: 51% in the high probability group, 22% in the moderate probability group, 4% in the low probability group, and < 1% in the very-low probability group, respectively [19].
1 1 1 −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 usefulness 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 accurate 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.
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 probability group, respectively [12]. Alternatively, pretest probability scores may be categorized 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 suspected 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, intermediate, high) set-up, is the Well's DVT rule (Table 1) [7,12,13]. According to a meta-analysis of 21 studies, assuming a 15% DVT prevalence, 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 probabilities, 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, allowing 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 challenged, 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 Ddimer 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 “moderate 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
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 dissection, 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-dependent D-dimer cut-off (age × 10 mcg/L) has been evaluated in patients 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, multicentre study (ADJUST-DVT) testing the safety of withholding treatment 2
Thrombosis Research xxx (xxxx) xxx–xxx
E. Bernardi, G. Camporese
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] Buller 2009 [21] Van der Velde 2011 [22]
Meta-analysis Prospective Prospective
Primary care Primary care Primary care
Schouten 2012 [28]
Retrospective
3.114b 1.002 495b 447b 2.086
Wells ≤ 1 + normal D-dimer Oudega ≤ 3 Oudega ≤ 3 Wells ≤ 1 + normal D-dimer Wells ≤ 1 + normal D-dimerc
Schutgens 2003 [40] Geersing 2014 [9]
Prospective Meta-analysis
Silveira 2015 [18] Cogo 2005 [44] Bernardi 1998 [45] Kraijenhagen 2002 [46] Wells 1997 [47] Schellong 2003 [3] Subramaniam 2005 [44] Gottlieb 2003 [48] Johnson 2010 [49] Stevens 2013 [50] Ageno 2015 [53]
Prospective Prospective Prospective Prospective Prospective Prospective Prospective Randomized Meta-analysis Prospective Prospective
1.2 (0.7 to 1.8) 1.4 (0.6 to 2.9) 1.4 (0.6 to 3.0) 1.6 (0.7 to 3.3) 0.5 (0.01 to 1.3) 0 (0.0 to 15.0) 0.6 (0.02 to 3.1) 0.4 (0.3 to 0.6) 0.6 (0.4 to 0.9) 5.9 (3.0–11.3) 0.7 (0.3 to 1.2) 0.4 (0 to 0.9) 0.7 (0.3 to 1.6) 0.3 (0 to 1.7) 0.3 (0.1–0.8) 0.2 (0.01 to 1.3) 0.8 (0.1 to 2.7) 2.5 (0.9 to 6.1) 0.6 (0.02 to 3.3) 1.5 (0.5 to 4.3)
29 (20 49 (46 57 (54 70 (66 48 (44 35 (24 22 (19 21 (13 46 (35 12 (10 NA 72 (69 48 (45 54 (50 NA NA 42 (36 NA NA 51 (46
812 3225b 4400b 1.135 1.741 946 1.739 593 1.646 542 261b 10.090 183 410b
Primary care, all patients Age ≥ 80 years
Wells ≤ 2 + normal D-dimer Wells ≤ 1 + normal quantitative D-dimer Wells ≤ 1 + normal qualitative D-dimer Wells ≤ 0 Normal serial CUS Normal serial CUS + normal D-dimer Normal serial CUS + normal D-dimer Wells ≤ 0 + normal single CUS Normal WLUS Normal WLUS Selective WLUSd Normal WLUS Normal WLUS Wells ≥ 2 + abnormal D-dimer + normal WLUS
Outpatients Inpatients Outpatients Outpatients Outpatients Outpatients Outpatients Outpatients Outpatients Outpatients with Wells ≥ 3 Outpatients with Wells ≥ 2 Outpatients
to to to to to to to to to to
40) 52) 61) 73) 52) 49) 25) 32) 57) 14)
to 75) to 50) to 58)
to 49)
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 Ddimer cut-off (age × 10 mcg/L) is currently on-going (NCT02384135). Less frequently, patients with DVT may have false-negative D-dimer results, such as those who had been treated with LMWH, or with leg symptoms lasting for > 2–3 weeks [29–31]. Numerous D-dimer assays have been validated against a gold standard (venography or ultrasonography) [30,32–38]. In general, quantitative assays possess higher sensitivity and lower specificity while the contrary is true for qualitative, and point-of-care tests [6,32,38,39]. Furthermore, D-dimer assays are less sensitive for isolated distal as compared with proximal DVT, and for asymptomatic as compared with symptomatic DVT [6,39]; finally, the overall performance of D-dimer assays decreases with increasing DVT prevalence [36]. Thus, D-dimer should not be used as a stand-alone test for to rule-in or rule-out DVT [5,10]. Instead, there is sound clinical evidence that suspected symptomatic DVT can be safely excluded in patients with a non-high, or an unlikely probability according to Wells, and normal Ddimer results, both with a qualitative or a quantitative assay, without the need for further testing (Table 3) [40,41]. A recent meta-analysis of 13 studies, including > 10.000 patients, reported that in patients with an unlikely probability according to the Wells' rule, DVT may be safely excluded on the basis of normal D-dimer results, irrespective if the latter is either a qualitative point-of-care or a quantitative laboratorybased assay (Table 3) [9]; instead, in patients with a likely or a high probability level, ultrasound testing should be performed [5,10,12]. In conclusion, a normal D-dimer result in combination with an unlikely, or non-high pretest probability according to Wells, allows for the exclusion of suspected symptomatic DVT. To the contrary, abnormal D-dimer levels mandate further testing; namely, ultrasonography.
Table 4 Randomized studies comparing CUS and WLUS in symptomatic patients with suspected symptomatic deep-vein thrombosis. Author
Bernardi 2008 [52]
Gibson 2009 [51]
Pre-selection
No
Wells ≥ 2, or abnormal Ddimer a
Diagnostic strategies
Serial CUS + D-dimer
Single WLUS
Serial CUS
Single WLUS
Patient n. Initial DVT prevalence VTE at follow-upb
1045 23%
1053 26%
257 23%
264 38%
b
Absolute difference
0.9 (0.3 to 1.2 (0.5 to 1.8) 2.2) 0.3 (− 1.4 to 0.8)
2.0 (0.6 to 1.2 (0.2 to 5.1) 4.3) 0.8 (− 2.3 to 2.9)
DVT: deep-vein thrombosis; CUS: compression ultrasonography; WLUS: whole-leg ultrasonography; VTE: venous thromboembolism. a Patients with Wells DVT score of ≤1 did not undergo ultrasonography. b Expressed as: % (95% confidence interval).
the common femoral vein at the groin and the popliteal vein in the popliteal fossa, including the so-called trifurcation (i.e.: the most proximal segments of the tibial and peroneal veins), are investigated [42]. It is a simple, quick, bedside approach that does not require specialized operators or high-range technical equipment, widely implemented in Emergency Departments, and hospital wards. Hence, CUS is usually available after-hours and during weekends. With WLUS, the whole deep-vein network of the leg is scanned in a continuous fashion, from the femoral veins, through the popliteal vein, to the distal veins. It is usually performed by consultants in vascular medicine or vascular surgery; thus, it is only available during office hours. Furthermore, it demands dedicated high-level machines and requires the patients to be moved from the ED/ward to be evaluated [3]. Both approaches are based on the single diagnostic criterion of vein compressibility; broadly, if the veins are compressible, then DVT is ruled-out; if the veins are not compressible, then DVT is ruled-in [43]. Notably, while DVT can be safely excluded after a single normal WLUS [3,44], this is not the case with CUS. Indeed, as the distal veins are not
1.3. Ultrasonography Two distinct ultrasound approaches are available to rule in/out DVT in symptomatic patients (Table 4): compression ultrasonography (CUS; also, “2-point ultrasonography”, or “limited-US”), and whole-leg ultrasonography (WLUS; also, “echo-color-Doppler”). With CUS only two distinct spots of the proximal-vein system, i.e., 3
Thrombosis Research xxx (xxxx) xxx–xxx
E. Bernardi, G. Camporese
Fig. 1. Suggested algorithm for the evaluation of suspected DVT in symptomatic patients. DVT: deep-vein thrombosis CUS: compression ultrasonography WLUS: whole-leg ultrasonography. * Suggested approach if the Wells DVT score is ≥ 3 (high probability).
A recent prospective, multicentre, cohort study devised a new sequential 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 evaluated with WLUS [53]. Only 35% of the patients required WLUS, and in half (49%) of them DVT was confirmed, indicating a high discriminatory power of the algorithm. The overall prevalence of thromboembolic 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 metaanalysis [49], although likely due to the small sample included, suggest the hypothesis that a single WLUS might not be sufficiently safe in highrisk 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 literature, and it's very unlikely to be performed based on the considerations 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 confirmation 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 ultrasonography 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-
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, confirmed 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 Ddimer, 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
E. Bernardi, G. Camporese
[13] S. Goodacre, A.J. Sutton, F.C. Sampson, Meta-analysis: the value of clinical assessment 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 sexdependent 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: comparison 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 Ddimer 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.
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 validated and accurate non-invasive imaging method in symptomatic patients, 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 investigated. Ultrasonography can be safely spared in patients with unlikely 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 evaluation 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 frequency 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 Evidencebased 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 Ddimer 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 suspected 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 thrombosis 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 thromboembolism 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
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 diagnostic 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,
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
6
K.S. Robinson, B. Lewandowski, Value of assessment of pretest probability of deepvein thrombosis in clinical management, Lancet 350 (9094) (1997) 1795–1798. 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. 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. 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. 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. 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 symptomatic deep vein thrombosis: a randomized controlled trial, JAMA 300 (14) (2008) 1653–1659, http://dx.doi.org/10.1001/jama.300.14.1653. 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, cohort study, Lancet Haematol. 2 (11) (2015) e474–80, http://dx.doi.org/10.1016/ S2352-3026(15)00190-8. 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. M.E. Sutter, S.D. Turnipseed, D.B. Diercks, P. Samuel, R.H. White, Venous ultrasound 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.