Guide détaillé à venir
Nous préparons un guide éducatif complet pour le Wells Score for DVT. Revenez bientôt pour des explications étape par étape, des formules, des exemples concrets et des conseils d'experts.
The Wells Score for Deep Vein Thrombosis (DVT) is the most widely used and internationally validated clinical prediction rule for estimating the pre-test probability of DVT in patients presenting with leg swelling, pain, or other symptoms suspicious for lower limb venous thrombosis. It was originally developed by Philip Wells and colleagues and published in The Lancet in 1997, with subsequent validation and refinement studies through 2003. The score integrates clinical history, risk factors, and physical examination findings to stratify patients into low, moderate, or high pre-test probability groups — or, using the simplified two-level approach, into 'DVT likely' versus 'DVT unlikely' categories. Deep vein thrombosis affects approximately 1 in 1,000 adults annually in Western populations, with over 500,000 new cases in the UK each year. Left untreated, DVT carries a significant risk of pulmonary embolism (PE) — the most feared complication — occurring in 15–30% of untreated proximal DVT cases. DVT also causes chronic post-thrombotic syndrome (PTS) in up to 50% of patients over 2 years, characterised by pain, swelling, skin changes, and in severe cases venous ulceration. The clinical utility of the Wells DVT score lies in its ability to direct diagnostic workup efficiently: patients classified as low probability (score ≤ 0) can be further evaluated with a D-dimer blood test — if D-dimer is negative, DVT is effectively excluded without the need for compression ultrasonography (CUS), safely reducing the burden of unnecessary imaging. Patients with high pre-test probability (score ≥ 3 original, or ≥ 2 on the two-level version) should proceed directly to imaging even if D-dimer is unavailable, as a negative D-dimer in high pre-test probability does not adequately exclude DVT. The score is embedded in NICE guideline NG158 (Venous Thromboembolic Disease, 2020) and SIGN guideline 122, which mandate its use in UK clinical practice for suspected DVT assessment.
Wells DVT Score = Sum of applicable clinical criteria
Criteria and Points:
1. Active cancer (treatment within 6 months or palliative) = +1 point
2. Paralysis, paresis, or recent plaster immobilisation of leg = +1 point
3. Bedridden ≥ 3 days, or major surgery within 12 weeks = +1 point
4. Localised tenderness along deep venous system = +1 point
5. Entire leg swollen = +1 point
6. Calf swelling ≥ 3 cm compared with asymptomatic leg = +1 point
(measured 10 cm below tibial tuberosity)
7. Pitting oedema confined to symptomatic leg = +1 point
8. Collateral superficial veins (non-varicose) = +1 point
9. Alternative diagnosis at least as likely as DVT = -2 points
Score Interpretation:
Original (three-tier):
Score ≤ 0 = Low probability (~5% DVT prevalence)
Score 1–2 = Moderate probability (~17% DVT prevalence)
Score ≥ 3 = High probability (~53% DVT prevalence)
Two-level version (NICE / clinical practice):
Score < 2 = DVT unlikely (proceed to D-dimer)
Score ≥ 2 = DVT likely (proceed to compression ultrasound)- 1Take a focused clinical history and perform a targeted physical examination of both lower limbs. Systematically assess each of the nine Wells criteria, recording each criterion as present (+1 or -2) or absent (0). Always examine the asymptomatic leg for comparison — calf circumference measurement requires a tape measure placed exactly 10 cm below the tibial tuberosity (bony prominence below the kneecap).
- 2Apply the 'alternative diagnosis' criterion rigorously — this is the most frequently misapplied criterion. If a clinician believes an alternative diagnosis (e.g., ruptured Baker's cyst, cellulitis, muscle tear, lymphoedema) is 'at least as likely' as DVT, deduct 2 points. This criterion was designed to downgrade the pre-test probability when a clear non-DVT explanation exists, preventing excessive downstream testing. Do not apply the deduction speculatively — it should reflect a genuine clinical assessment.
- 3Sum all applicable points to obtain the total Wells DVT score. Apply the two-level interpretation for practical clinical use: score ≥ 2 = 'DVT likely' (proceed to compression ultrasonography regardless of D-dimer); score < 2 = 'DVT unlikely' (request D-dimer first; if D-dimer is negative, DVT is safely excluded with no further imaging required).
- 4For patients classified as 'DVT unlikely' (score < 2): order a highly sensitive D-dimer assay (ELISA method, sensitivity > 95%). If D-dimer is negative, DVT is excluded and no further investigation is required. If D-dimer is positive, proceed to compression ultrasonography of the proximal leg veins (common femoral, femoral, and popliteal veins).
- 5For patients classified as 'DVT likely' (score ≥ 2): order compression ultrasonography directly. If CUS is positive (non-compressible vein), diagnose proximal DVT and initiate anticoagulation. If CUS is negative but clinical suspicion remains high, consider repeat CUS at 6–8 days or MR venography. Do not use D-dimer alone in high pre-test probability patients — sensitivity is inadequate in this group.
- 6Initiate anticoagulation promptly if DVT is confirmed. First-line treatment is a direct oral anticoagulant (DOAC) — apixaban (10 mg twice daily for 7 days then 5 mg twice daily) or rivaroxaban (15 mg twice daily for 21 days then 20 mg daily) are preferred per NICE NG158 and ESC guidelines. Warfarin (with LMWH bridging) remains appropriate in renal failure or patient preference. Treatment duration depends on provoked vs unprovoked DVT status and recurrence risk.
- 7Reassess the clinical context for recurrence risk and appropriate anticoagulation duration: provoked DVT (recent surgery, immobilisation, oestrogen therapy) — typically 3 months then stop; unprovoked DVT — 3–6 months minimum, then review risk-benefit for extended therapy; cancer-associated DVT — indefinite anticoagulation while cancer is active (LMWH or DOACs preferred over warfarin in cancer).
Proceed directly to compression ultrasonography — do not rely on D-dimer
A score of 4 places this patient firmly in the 'DVT likely' category with an approximate DVT prevalence of 50–60% in this pre-test probability group. Post-major orthopaedic surgery is one of the highest DVT-risk scenarios. D-dimer would almost certainly be elevated post-operatively regardless of DVT, making it non-discriminatory. Compression ultrasound is the appropriate next investigation and should be performed urgently given the clinical picture.
Order D-dimer — if negative, DVT excluded, no imaging needed
Despite the plausible travel history (long haul flight), the clinical picture is more consistent with muscle strain than DVT: the tenderness is diffuse over the muscle belly (not specifically along the femoral/popliteal vein course), calves are equal, and an alternative diagnosis (muscle strain) is at least as likely — scoring -2 for this criterion. A negative highly sensitive D-dimer will safely exclude DVT with a likelihood ratio approaching 0.05, avoiding the need for compression ultrasound.
Positive D-dimer with Wells <2 mandates ultrasound — DVT not excluded
The score of 1 places her in the 'DVT unlikely' category, triggering D-dimer testing. However, D-dimer is elevated at 820 ng/mL. A positive D-dimer in any pre-test probability category does NOT exclude DVT — it mandates progression to compression ultrasonography. Only a negative D-dimer in the 'DVT unlikely' group safely excludes DVT without imaging. Oral contraceptive use substantially raises DVT risk (3-fold) and should inform clinical index of suspicion.
Wells DVT score is validated for unilateral lower limb DVT suspicion; bilateral DVT in cancer warrants direct imaging
The Wells DVT score was derived for unilateral limb DVT assessment. In a cancer patient with bilateral symmetrical leg oedema, the score's comparative criteria (calf swelling 3 cm more than asymptomatic leg, pitting oedema confined to symptomatic leg) cannot be applied as intended. In this clinical context, bilateral CUS or CT venography is often indicated directly. Active cancer alone is a strong independent DVT risk factor and justifies low threshold for imaging in equivocal bilateral presentations.
Emergency department DVT diagnostic pathways — using Wells score to stratify patients into D-dimer testing (low probability) versus direct compression ultrasonography (high probability), reducing unnecessary imaging and wait times, where accurate wells dvt score analysis through the Wells Dvt Score supports evidence-based decision-making and quantitative rigor in professional workflows
Primary care and GP surgeries — applying the Wells score to patients presenting with unilateral leg swelling to determine whether same-day D-dimer testing or urgent 2-week wait DVT imaging referral is appropriate
Anticoagulation clinics — reviewing Wells scores retrospectively to audit diagnostic accuracy, ensure appropriate anticoagulation initiation, and assess DVT recurrence risk for treatment duration decisions, where accurate wells dvt score analysis through the Wells Dvt Score supports evidence-based decision-making and quantitative rigor in professional workflows
Pre-operative VTE prophylaxis planning — using Wells score principles to identify high-risk patients for extended post-discharge LMWH or DOAC prophylaxis after major orthopaedic surgery, where accurate wells dvt score analysis through the Wells Dvt Score supports evidence-based decision-making and quantitative rigor in professional workflows
Clinical research and guideline development — Wells DVT score as the reference standard pre-test probability tool in VTE diagnostic trials, enabling consistent patient stratification across international multicentre studies
DVT in pregnancy — Wells score limitations
Pregnant women are at 5-fold higher risk of DVT than non-pregnant age-matched controls, with highest risk in the third trimester and 6 weeks postpartum. The Wells DVT score was not derived in pregnant populations and its performance in pregnancy is uncertain — many criteria are confounded by normal pregnancy physiology (bilateral leg oedema, reduced mobility, abdominal surgery at delivery). NICE recommends using clinical judgement alongside Wells criteria in pregnancy, with a low threshold for compression ultrasonography. D-dimer rises physiologically in pregnancy, making it non-discriminatory. LMWH is the anticoagulant of choice throughout pregnancy and postpartum breastfeeding (DOACs are contraindicated). Seek specialist obstetric medicine or haematology input.
Upper limb DVT — Wells score does not apply
The Wells DVT score applies exclusively to lower limb deep vein thrombosis. Upper limb DVT (affecting subclavian, axillary, or brachial veins) is a separate clinical entity with different risk factors (central venous catheters, pacemaker leads, thoracic outlet syndrome, vigorous upper limb exercise 'effort thrombosis' / Paget-Schroetter syndrome) and different assessment pathways. Upper limb DVT assessment uses clinical suspicion, CUS of the arm veins, and CT venography; dedicated upper limb DVT scoring tools exist but are less well validated than the Wells DVT score for the lower limb.
Isolated distal (calf) DVT — controversial management
Thrombus confined to calf (tibial, peroneal, soleal) veins without proximal extension is clinically controversial: the risk of propagation to proximal veins is 15–25% if untreated, but PE risk from isolated distal DVT alone is low. Two management strategies exist: (1) Anticoagulate for 3 months (same as proximal DVT) to prevent propagation and reduce post-thrombotic syndrome — preferred when propagation risk is high (extensive thrombus, cancer, no identifiable precipitant); (2) Serial CUS monitoring at 7–14 days without anticoagulation — appropriate when DVT is minor, provoked, and patient can be reliably followed up. The decision requires individualised risk-benefit assessment; involvement of a haematologist or anticoagulation service is recommended.
Cancer-associated DVT — special considerations
Patients with active cancer have a 4–7 fold increased DVT risk (Trousseau's syndrome), and DVT may be the presenting feature of an occult malignancy. Cancer-associated DVT requires indefinite anticoagulation while cancer is active. LMWH was the historical standard (CLOT trial, 2003), but DOACs (edoxaban in HOKUSAI-Cancer, rivaroxaban in SELECT-D) have demonstrated non-inferiority or superiority with improved convenience. DOACs are now recommended for most cancer-associated DVT except gastrointestinal malignancies (higher bleeding risk — prefer LMWH in GI cancers). Recurrence rates remain high in cancer-associated VTE (15–20%/year without anticoagulation). Always screen for occult malignancy in unprovoked DVT in patients over 40 years.
D-dimer age adjustment — improving specificity in older patients
D-dimer has notoriously low specificity in older patients — approximately 50% of patients over 70 years will have an elevated D-dimer due to comorbidity and physiological reasons unrelated to VTE. The age-adjusted D-dimer threshold (patient age × 10 ng/mL for patients over 50 years, e.g., a 70-year-old's threshold is 700 ng/mL instead of 500 ng/mL) has been validated in large prospective studies (ADJUST-PE and other trials) and significantly increases specificity without meaningfully reducing sensitivity. NICE NG158 endorses age-adjusted D-dimer thresholds for patients over 50 years old, reducing unnecessary imaging in older low-probability patients.
| Score (Two-Level) | Category | DVT Prevalence | D-Dimer if Negative | Recommended Next Step |
|---|---|---|---|---|
| < 2 | DVT unlikely | ~5% | DVT excluded — no imaging needed | High-sensitivity D-dimer first; if positive → CUS |
| ≥ 2 | DVT likely | ~28% | Does NOT exclude DVT | Proceed directly to compression ultrasonography |
| ≥ 3 (original) | High probability | ~53% | Does NOT exclude DVT | Urgent CUS; consider empirical anticoagulation if imaging delayed |
| CUS negative + DVT likely | Residual suspicion | ~2% distal DVT missed | — | Repeat CUS at 6–8 days or MR venography |
| CUS positive | DVT confirmed | — | — | Initiate anticoagulation; refer haematology if unprovoked or recurrent |
What is the Wells DVT score used for?
The Wells DVT score is a clinical prediction rule used to stratify the pre-test probability of deep vein thrombosis (DVT) in patients presenting with lower limb symptoms — typically unilateral leg swelling, pain, warmth, or redness. By categorising patients into low, moderate, or high probability (or 'DVT unlikely' vs 'DVT likely' in the two-level version), the score guides efficient diagnostic workup: directing low-probability patients to D-dimer testing first (avoiding unnecessary imaging), and directing high-probability patients directly to compression ultrasonography. It is mandated in UK DVT diagnostic pathways by NICE NG158.
What is the two-level Wells DVT score?
The two-level Wells DVT score simplifies the original three-tier system (low/moderate/high) into a binary classification: 'DVT unlikely' (score < 2) and 'DVT likely' (score ≥ 2). This simplification has been widely adopted in clinical practice because it aligns more clearly with diagnostic management pathways: 'DVT unlikely' patients proceed to D-dimer testing, while 'DVT likely' patients proceed directly to compression ultrasonography. The two-level approach is the version recommended by NICE NG158 and used in most UK hospitals.
Can D-dimer alone be used to rule out DVT?
D-dimer alone can only rule out DVT in the context of a low pre-test probability (Wells DVT score < 2 or 'DVT unlikely'). A highly sensitive D-dimer assay (ELISA method, sensitivity > 95%) that is negative in a patient with Wells score < 2 has a very high negative predictive value (>99%), safely excluding DVT without the need for imaging. However, D-dimer alone cannot rule out DVT in patients with a high pre-test probability (Wells ≥ 2 or 'DVT likely') — even a negative D-dimer does not exclude DVT with adequate certainty in this group, and compression ultrasonography is required.
What is compression ultrasonography and how is it used in DVT diagnosis?
Compression ultrasonography (CUS) is the primary imaging modality for DVT diagnosis. The sonographer applies graded pressure with the ultrasound probe over the venous system of the leg. A normal vein compresses completely with gentle probe pressure (the lumen collapses). A DVT-containing vein does not compress — the intraluminal thrombus prevents the walls from meeting. Whole-leg CUS examines proximal veins (common femoral, femoral, popliteal) and distal calf veins. Proximal-only CUS is faster and avoids the lower clinical significance of isolated distal calf DVT. NICE NG158 recommends proximal CUS initially with repeat at 6–8 days if initial imaging is negative in 'DVT likely' patients.
How long should anticoagulation be given for DVT?
Anticoagulation duration depends on DVT provocation and recurrence risk: (1) Provoked DVT (identifiable transient risk factor — surgery, immobilisation, oestrogen therapy, pregnancy): 3 months, then reassess. Recurrence risk after stopping anticoagulation is 5–10% per year. (2) Unprovoked DVT (no identifiable risk factor): minimum 3 months; then consider extended treatment based on bleeding risk, patient preference, and D-dimer result on treatment (D-dimer positive at 1 month off anticoagulation predicts higher recurrence). (3) Cancer-associated DVT: indefinite while cancer is active — LMWH or DOACs (edoxaban, rivaroxaban) are preferred over warfarin in cancer. (4) Second unprovoked DVT: generally indefinite anticoagulation is recommended.
What is post-thrombotic syndrome and how is it prevented?
Post-thrombotic syndrome (PTS) is a long-term complication of DVT occurring in up to 20–50% of patients within 2 years. It results from persistent venous hypertension due to valve damage and residual obstruction caused by thrombus. Features include chronic aching pain, oedema, varicosities, skin changes (lipodermatosclerosis), and in severe cases venous ulceration. Prevention strategies include: prompt and adequate anticoagulation to limit thrombus extension, early ambulation (bed rest does not reduce PE risk and delays recovery), graduated compression stockings (GCS) for symptomatic relief (though the SOX trial showed no benefit in preventing PTS), and weight management. Patients with iliofemoral DVT may benefit from catheter-directed thrombolysis in specialist centres.
What is the difference between the Wells DVT score and the Wells PE score?
Both are clinical prediction rules developed by Philip Wells, but they assess different conditions. The Wells DVT score estimates pre-test probability of deep vein thrombosis in the lower limb using 9 criteria including local examination findings. The Wells PE score estimates pre-test probability of pulmonary embolism using different criteria including clinical signs of DVT, heart rate, immobilisation/surgery, prior DVT/PE, haemoptysis, malignancy, and whether PE is the most likely diagnosis. While the two scores share some concepts (cancer, immobilisation), their criteria, weighting, and interpretation thresholds differ. They are used in complementary diagnostic pathways — a patient with DVT confirmed by Wells DVT + CUS may also need Wells PE assessment if they develop dyspnoea.
Which anticoagulants are preferred for DVT treatment?
Direct oral anticoagulants (DOACs) are the first-line preferred treatment for most patients with DVT, based on multiple non-inferiority randomised trials versus LMWH-warfarin. Apixaban (AMPLIFY trial) and rivaroxaban (EINSTEIN-DVT trial) are most widely used and can be initiated as monotherapy without LMWH bridging. Dabigatran and edoxaban require 5–10 days of LMWH before switching. DOACs have fewer drug interactions, no INR monitoring requirement, and similar or better safety profiles to warfarin. LMWH is preferred in cancer-associated DVT (CLOT and SELECT-D trials showed LMWH superiority over warfarin; newer data support DOACs in most cancer subtypes). Warfarin with LMWH bridging remains appropriate for patients with severe renal impairment (eGFR < 30 mL/min) or antiphospholipid syndrome.
Conseil Pro
The 'alternative diagnosis at least as likely as DVT' criterion (-2 points) is the most impactful and most frequently misapplied criterion in the Wells DVT score. It requires a genuine clinical assessment — not just a theoretical possibility — that an alternative diagnosis is plausible. The three most common alternative diagnoses that justify the deduction are: (1) ruptured Baker's cyst (posterior knee swelling with a history of knee osteoarthritis), (2) cellulitis (unilateral leg redness, warmth, and tenderness with skin changes starting at an entry point), and (3) acute muscle tear (localised muscle belly tenderness with a specific injury mechanism). Always document your clinical reasoning when applying or not applying this criterion.
Le saviez-vous?
Philip Wells, the Canadian physician who developed the DVT and PE scoring systems, first published the DVT score in 1997 in The Lancet when he was a junior researcher — the paper became one of the most cited clinical prediction rule papers in medical literature with over 4,000 citations. The score was so successful that Wells was subsequently asked to develop a companion score for pulmonary embolism (published in 2000 and 2003), creating the now-ubiquitous 'Wells PE score' that is used worldwide alongside the DVT score in complete venous thromboembolism diagnostic algorithms. Together, these two scores have fundamentally transformed how VTE is diagnosed, reducing both missed diagnoses and unnecessary anticoagulation.
Références
- ›Wells PS et al. Value of assessment of pretest probability of deep-vein thrombosis in clinical management — Lancet 1997
- ›Wells PS et al. Evaluation of D-dimer in the diagnosis of suspected deep-vein thrombosis — NEJM 2003
- ›NICE guideline NG158: Venous thromboembolic diseases: diagnosis, management and thrombophilia testing — 2020, updated 2023
- ›Konstantinides SV et al. 2019 ESC Guidelines for the diagnosis and management of acute pulmonary embolism — European Heart Journal 2020
- ›Righini M et al. Age-adjusted D-dimer cutoff levels to rule out pulmonary embolism: the ADJUST-PE study — JAMA 2014