Deep vein thrombosis: clinical examination and ultrasound can diagnose or exclude DVT.

Clinical bottom line (level 1a)

  1. Clinically scoring patients with suspected deep vein thrombosis aids diagnosis (high risk: (LR+16) ; low risk (LR+0.13) , but cannot rule it out.
  2. Low risk patients with a negative ultrasound scan have <1% chance of a deep vein thrombosis.
  3. A positive ultrasound scan diagnoses deep vein thrombosis in high- or moderate-risk patients (SpPin 99%).
  4. A venogram is still necessary in high- or moderate-risk patients with a negative ultrasound scan, and low-risk patients with a positive ultrasound scan (13% of patients).
Wells et al: Lancet 1995; 345: 1326-1330
Expires May 2003

The study

Setting: three teaching hospitals, Canada and Italy

529 patients (aged ?, ?% female) suspected deep venous thrombosis for <60 days

Excluded if
  • previous objectively documented DVT or pulmonary embolism
  • anticoagulation treatment for more than 48 hours
  • pregnant
  • contraindication to contrast media
  • below knee amputation
  • concomitant clinical suspicion of pulmonary embolism
  • refusal to give consent



    • Independent blinded reference standard, applied in all patients from a consecutive appropriate spectrum.
    Reference standard:
      • venogram- positive if constant intraluminal filling in two projections, or
      • ultrasound scan of common femoral vein at inguinal ligament, popliteal vein at knee joint line, followed to trifurcation of calf veins. All veins viewed only in transverse. Abnormal result in any vein showed lack of full compressibility
      • 'obviously clear' patients followed-up for six months
    Diagnostic test: clinical prediction rule
    • major risks
    • active cancer (on-going treatment or diagnosed within 6 months or palliative care)
    • paresis, paralysis or recent plaster cast immobilisation of lower extremity
    • recently bedridden for more than 3 days and/or major surgery within 4 weeks
    • localised tenderness over distribution of deep veins
    • thigh and calf swollen
    • calf swelling more than 3 cm compared with asymptomatic side, measured at 10 cm below tibial tubercle
    • family history of more than two primary relatives with DVT
    • localised tenderness over distribution of deep veins
    • minor risks
    • history of trauma in last 60 days to symptomatic leg
    • hospital stay in last 6 months
    • dilated superficial veins (non-varicose in symptomatic leg only)
    • erythema
    • pitting oedema symptomatic leg only

    • overall low risk, one of:
      • 1 major and 2 or less minor and alternative diagnosis
      • 1 major and 1 or less minor and no alternative diagnosis
      • 0 major and 3 or less minor and alternative diagnosis
      • 0 major and 2 or less minor and no alternative diagnosis
    • overall moderate risk:
      • all other categories
    • overall high risk:
      • 3 or more major and no alternative diagnosis
      • 2 or more major and 2 or more minor and no alternative diagnosis

    The evidence

    pre-test probability of DVT: 25%, (95% CI: 22% to 29%)
    pre-test probability of proximal DVT: 21%, (95% CI: 18% to 25%)

    diagnostic test DVT no DVT LR+
    (95% CI)    		 post-test probability LR-
    (95% CI)    		 post-test probability
    high risk 72 13 16
    (9.3 to 28)     		85% 0.48
    (0.40 to 0.58)     		14%
    moderate risk 47 96 1.4
    (1.1 to 1.9)     		33% 0.86
    (0.75 to 0.99)     		23%
    low risk 16 285 0.16
    (0.10 to 0.26)     		5% 3.2
    (2.7 to 3.8)     		52%
    total 135 394


    diagnostic test proximal DVT no proximal DVT LR+
    (95% CI)    		 post-test probability LR-
    (95% CI)    		 post-test probability
    high risk 69 16 16
    (9.6 to 26)     		81% 0.40
    (0.32 to 0.51)     		10%
    moderate risk 34 109 1.2
    (0.83 to 1.6)     		24% 0.95
    (0.83 to 1.1)     		20%
    low risk 10 291 0.13
    (0.070 to 0.23)     		3% 3.0
    (2.6 to 3.6)     		45%
    total 113 416


    diagnostic test DVT no DVT LR+
    (95% CI)    		 LR-
    (95% CI)
    ultrasound for all patients 39
    ( to )     		0.23
    ( to )
    ultrasound for high risk inf
    ( to )     		0.09
    ( to )
    ultrasound for moderate risk 53
    ( to )     		0.39
    ( to )
    ultrasound for low risk 24
    ( to )     		0.34
    ( to )
    total


    diagnostic test proximal DVT no proximal DVT LR+
    (95% CI)    		 LR-
    (95% CI)
    ultrasound for all patients 47
    ( to )     		0.11
    ( to )
    ultrasound for high risk inf
    ( to )     		0.09
    ( to )
    ultrasound for moderate risk 76
    ( to )     		0.17
    ( to )
    ultrasound for low risk 33
    ( to )     		0.20
    ( to )
    total

    Comments

    1. Pre-test probability of DVT was 42% in Italy and 22% in Canada- why the difference? Race differences? Are patients screened out more effectively in Italy by primary care physicians?
    2. Excellent correlation between clinicians for clinical scoring: K intraobserver= 0.85.

    Citation

    1. Wells PS, Hirsch J, Anderson DR, et al: Accuracy of clinical assessment of deep-vein thrombosis. Lancet 1995; 345: 1326-1330
    Search Terms: hand search
    Contributor: Chris Ball and Clare Wotton, May 2000
    Reviewer:

    Clinical Question.
    Patient suspected deep venous thrombosis
    Intervention or Exposure clinical prediction rule
    Outcome diagnosis