D-dimer testing, tailored to clinical pretest probability, to minimize initial and follow-up ultrasound testing in suspected deep vein thrombosis: A management study

A D-dimer test is a blood test that measures proteins that are formed when the body breaks down leg blood clots (e.g., deep vein thrombosis; DVT). If a D-dimer test is normal, it is very unlikely that a patient has a DVT. However, because other illnesses also increase D-dimer levels, an abnormal D-dimer level has been considered unhelpful. We recently completed a study that showed that, rather than labelling D-dimer results as "positive" and "negative" in the same way in all patients, it is better to use a higher D-dimer cut-off value (i.e., level that separates a negative from a positive result) if, after examining the patient, the doctor thinks that there is "Low DVT probability". This allowed DVT to be ruled-out, without performing an ultrasound, in 50% more patients. Analysis of that study's findings also suggested a number of other ways to improve how D-dimer testing is used to diagnose DVT. We will now use this new information to further improve DVT diagnosis in 1,500 patients from 10 hospitals. Patients will be assessed as Low, Moderate or High DVT probability, and have D-dimer measured. DVT will be ruled-out if D-dimer is 3,000 with Low or Moderate probability, or >1,500 with High probability. We will then follow patients for 3 months, expecting that very few will have thrombosis (