A rule out test only, applied after a clinical probability score
Mnemonic
“Score first, then test”:
- Low or unlikely clinical probability by the Wells score: a negative D-dimer excludes venous thromboembolism, and no imaging is needed
- High or likely probability: go straight to imaging. A negative D-dimer does not exclude the diagnosis and should not be requested
“D-dimer is a rule out test, never a rule in test”, because it is highly sensitive and poorly specific.
Age adjustment in patients over 50: use a threshold of age x 10 micrograms per litre, which improves specificity without losing sensitivity.
Raised by almost anything: age, pregnancy, infection, sepsis, surgery, trauma, malignancy, DIC, aortic dissection, liver and renal disease, heart failure. In a hospital inpatient it is raised so often that it becomes useless, which is why the probability score must come first.
Expansion
What it is: a fibrin degradation product, so it is raised whenever clot is being formed and broken down anywhere in the body.
Its properties: highly sensitive but poorly specific, so it has a good negative predictive value and a poor positive one. It is therefore a rule out test.
Correct use
- Assess the clinical probability first, with the Wells score for deep vein thrombosis or pulmonary embolism, or the Geneva score
- Low or unlikely probability: a negative D-dimer excludes the diagnosis and no imaging is needed
- High or likely probability: proceed straight to imaging. A negative D-dimer does not exclude the diagnosis here, and should not be requested
- PERC rule may exclude pulmonary embolism without any test in a very low risk patient
Age adjustment: in patients over 50, a threshold of age multiplied by 10 micrograms/l improves specificity without losing sensitivity.
Causes of a raised D-dimer other than venous thromboembolism: age, pregnancy, infection and sepsis, surgery and trauma, malignancy, disseminated intravascular coagulation, aortic dissection, liver disease, renal disease, heart failure, inflammatory disease, and recent thrombolysis. In a hospital inpatient it is raised so often that its usefulness collapses.
The commonest clinical error is requesting a D-dimer without a probability score, then being obliged to image a patient in whom the diagnosis was never plausible.