Patients with a low risk Wells’s score and a negative moderately-sensitive D-dimer assay (e.g. whole blood agglutination – SimpliRed) have less than a 1% chance of deep vein thrombosis (DVT). Patients with a moderate risk Wells’s score and a negative highly-sensitive D-dimer assay (e.g. ELISA) have a 1% chance of DVT.
You are working a busy ED shift at your hospital, when a 60 year old male presents with a two day history of a red, hot, swollen, left calf. Ten days ago the patient cut the bottom of his foot. Your physical examination reveals a well, afebrile patient with inguinal adenopathy and pitting edema of the affected leg.
It is past banker hours – i.e. you do not have immediate access to ultrasound – do you:A. Make a diagnosis of cellulitis, begin antibiotic therapy and arrange outpatient follow up;
B. Bring in an ultrasound technician to perform an after-hours study to exclude DVT;
C. Start the patient on antibiotics, low molecular weight heparin and arrange an outpatient ultrasound tomorrow; or
D. Calculate the patient’s risk of DVT and
determine whether you can use a D-dimer assay to exclude the diagnosis of DVT?
Wells PS, Owen C, Doucette S, Ferguson D, Tran H, “Does this patient have deep vein thrombosis?” JAMA 2006 Jan 11;295(2):199-207.
If you did not consider option #4, hopefully, this article can convince you of a new diagnostic option for a clinical scenario emergency physicians routinely experience.
Wells, the creator of the DVT clinical prediction rule, authored a meta-analysis on the use of a formal score plus a D-dimer assay to rule out DVT. Essentially, this study involved an extensive and systematic review of the literature to find all potentially relevant studies on the topic. The authors then used formal criteria to select relevant articles, choosing only high quality studies. The results of the studies were statistically combined to determine a common answer.
From the 11 articles involving nearly 5700 patients, the following conclusions were made.
A patient with a low risk score (<=0 points) has a 5% risk of DVT. This decreases to 0.9% if a moderately-sensitive (negative Likelihood Ratio 0.20) D-dimer assay (e.g. whole blood agglutination) is negative and 0.5% if a highly-sensitive (negative Likelihood Ratio 0.10) D-dimer assay (e.g. ELISA) is negative.
In patients with a moderate risk score (1-2 points), the risk of DVT is 17%. If a highly-sensitive D-dimer assay is negative (negative Likelihood Ratio 0.05) the risk decreases to 1%.
In patients with a high risk score (>2 points) no D-dimer assay can safely rule out DVT.
Should I only use highly-sensitive D-dimer assays? The problem with highly-sensitive assays is that they can be falsely positive. Old age, comorbid illness and other conditions can lead to a positive test, even in the absence of venous thromboembolic disease. Using a highly-sensitive rather than a moderately-sensitive assay in low risk patients, may lead to more false positive tests requiring further diagnostic testing (e.g. ultrasound). So, you will not be any further ahead!
If you selected option #1 as a solution for this case, you may be entirely correct. Wells’s study does not suggest that every swollen leg requires that DVT be excluded. As the treating physician, only you can determine the appropriate differential diagnosis. If you do decide to incorporate this diagnostic approach into your practice, it is essential that you calculate a formal Wells’s score on your patient. The medical evidence does not support using your clinical gestalt to categorize a patient as low or medium risk prior to ordering a D-dimer assay. Finally, a positive D-dimer does not mean the patient has a DVT. Rather, it indicates that the patient requires further definitive testing – e.g. a compression duplex ultrasound. It should also be stressed that this diagnostic approach is only for lower extremity and not for upper extremity DVT.
Deciding that a DVT is important to exclude in this patient, a formal Wells’s score is calculated. The score is 0, making the patient low risk for DVT. A SimpliRed D-dimer is ordered. When the assay is reported as negative, the diagnosis of DVT is safely excluded and the patient is appropriately treated for cellulitis without the need for arranging ultrasound or initiating anticoagulation. Your shift just got easier!
The Wells’s clinical prediction rule for DVT is not easy to remember. You can readily find PDA or paper versions to carry with you. However, be cautious when Googling Wells’s score, ensuring that you are using the Wells’s score for DVT and NOT for pulmonary embolism.
Jonathan Sherbino, MD, MEd, FRCPC, is on the BEEM faculty and is an assistant professor, Division of Emergency Medicine, McMaster University