D-Dimer

D-dimer is a degradation product released when plasmin breaks down cross-linked fibrin, and its presence shows that coagulation and fibrinolysis have both been activated. It is measured on citrated plasma by latex enhanced immunoturbidimetric or enzyme immunoassay methods.

Its main use is as a rule out test. A negative result with a low or intermediate pre-test probability makes acute venous thromboembolism (deep vein thrombosis and pulmonary embolism) very unlikely, with a negative predictive value over 95%.

A positive result is not diagnostic, because many conditions raise it.

Units. Results are reported as fibrinogen equivalent units (FEU) or D-dimer units (DDU). Numerically, FEU values are about double DDU values. The cut off must match the unit, and a mismatch is a source of serious error.

Specificity falls with age, with only a small minority of people over 80 years having a result under the conventional cut off. It is also low in hospital inpatients, cancer, infection, pregnancy, trauma and after surgery.

Specimen. Citrated plasma. Samples are stable for about 4 to 8 hours at room temperature, and longer in some assays. Haemolysis, lipaemia and icterus can interfere.

Principle. A clinical probability score is calculated first, and D-dimer is used only when the probability is not high.

Clinical probability.

• Two level Wells score for DVT: unlikely 1 or less, likely 2 or more.
• Two level Wells score for pulmonary embolism: unlikely 4 or less, likely over 4.
• Low gestalt probability under 15% with all PERC criteria negative (age under 50, heart rate under 100, oxygen saturation 95% or more, no haemoptysis, no oestrogen use, no previous VTE, no unilateral leg swelling, no recent surgery or trauma): no D-dimer is needed.

Action by result.

• Unlikely probability and negative D-dimer: venous thromboembolism excluded without imaging, with a 3-month failure rate under 1%.
• Unlikely probability and positive D-dimer: imaging. Compression ultrasound for DVT, and CT pulmonary angiography or ventilation perfusion scanning for pulmonary embolism.
• Likely or high probability: imaging regardless of D-dimer.

Assay requirements. A negative result is reliable only with a high sensitivity quantitative assay (sensitivity over 95%). Qualitative bedside assays and some point-of-care methods have lower sensitivity (about 85 to 90%) and are suited only to low probability.

Other uses.

• Suspected DIC: markedly raised values form part of the DIC scoring system, together with platelets, PT and fibrinogen.
• Suspected aortic dissection: a value under 500 ng/mL FEU with a low risk score (0 or 1) makes dissection unlikely. A normal result does not exclude dissection in high risk patients.
• Duration of anticoagulation after unprovoked venous thromboembolism: a raised value after stopping anticoagulant predicts recurrence.
• Suspected cerebral venous thrombosis in low probability patients, with lower sensitivity than for pulmonary embolism.

Conventional cut off. 500 ng/mL FEU (0.5 mg/L FEU), or about 250 ng/mL DDU.

Age adjusted cut off. For patients over 50 years, the cut off is the age in years multiplied by 10 ng/mL FEU (or by 5 ng/mL for DDU assays). It raises specificity in older adults and keeps sensitivity over 97% for pulmonary embolism.

Alternative cut offs. In non-pregnant patients, clinical item algorithms (the YEARS algorithm) use a cut off of 1000 ng/mL FEU when no item is present (clinical signs of DVT, haemoptysis, pulmonary embolism as the most likely diagnosis) and 500 ng/mL FEU when one or more is present. Pregnancy adapted algorithms use similar cut offs, and local policy governs their use.

Causes of raised D-dimer. Venous thromboembolism, DIC, aortic dissection, sepsis, malignancy, recent surgery, trauma, burns, pregnancy and the postpartum period, pre-eclampsia, placental abruption, advancing age, liver disease, renal impairment, inflammatory disease, acute stroke, myocardial infarction, sickle cell crisis, COVID-19, thrombolytic therapy, large haematoma, hospital admission, and heterophile antibodies or rheumatoid factor (false positive).

Causes of a falsely low or negative result. Anticoagulation for several days (it lowers D-dimer), symptoms present for over 1 week, small thrombus burden (distal calf DVT, subsegmental pulmonary embolism), and use of a low sensitivity assay.

Critical findings needing immediate action.

• Markedly raised values (over about 10 000 ng/mL FEU) with bleeding, thrombocytopenia and a prolonged PT indicate DIC.
• A raised value with sudden severe chest or back pain, pulse or blood pressure differences, or hypotension needs immediate imaging for aortic dissection.
• A positive result with a high clinical probability of pulmonary embolism, with hypoxia or haemodynamic compromise, needs urgent imaging and treatment, not repeat testing.

Pitfalls.

• Testing without a prior clinical probability score leads to over investigation, because most positive results are false positives.
• Using FEU cut offs with DDU assays, or the reverse, leads to errors of two fold.
• D-dimer is not useful in high probability, or in patients with many alternative causes of a rise (cancer, recent surgery, sepsis, late pregnancy).
• Levels fall over days to weeks after a thrombus, and with anticoagulation.

Special groups.

• Pregnancy: levels rise progressively with gestation, and many women have values over the conventional cut off by the third trimester. Pregnancy adapted algorithms exist, and a negative result remains useful.
• Older adults: use the age adjusted cut off.
• Children: pre-test probability scores are not validated, and D-dimer is of limited use.
• Neonates: baseline values are higher.
• Cancer and hospitalised patients: low specificity, with a negative result still informative.

Follow up tests. Compression ultrasound, CT pulmonary angiography, ventilation perfusion scan, CT aortography, echocardiography, clotting screen, fibrinogen, full blood count and film, troponin and natriuretic peptides for risk stratification in pulmonary embolism, thrombophilia screen when indicated after the acute event.