Background: Deep venous thrombosis (DVT) is an uncommon yet potentially serious condition in patients presenting to primary care physicians. When DVT is suspected, most patients are referred for further testing, including compression ultrasonography. Although ultrasonography is safe and easily available, it requires an additional patient visit and increases medical costs. The Wells rule is a clinical decision rule that is used in many hospitals. The rule combines patient history and physical examination, followed by d-dimer assay to determine which patients need ultrasonography. The combination of low clinical suspicion, based on the decision rule, and a negative d-dimer assay effectively rules out DVT and precludes the need for ultrasonography.
The accuracy of the Wells rule in primary care has been questioned. A validation study showed an unacceptable failure rate, even when the results of the d-dimer assay were applied. This led to the development and validation of a primary care–specific decision rule (see accompanying table) that included elements of patient medical history, physical examination, and d-dimer assay, but, unlike the Wells rule, it does not include the estimated probability of an alternative diagnosis. To determine the safety and effectiveness of the primary care rule, van der Velde and colleagues compared the Wells rule with the primary care rule in an unselected primary care population.
The Study: More than 300 primary care physicians in the Netherlands participated in this study. Eligible participants presented with suspected DVT. Patients who were younger than 18 years and those who had received low-molecular-weight heparin or vitamin K antagonists were excluded. The physicians collected appropriate clinical data to apply both the Wells and primary care rules, and also performed a rapid point-of-care d-dimer assay and managed the patient clinically based on the score of the primary care rule. Under the primary care rule, patients with scores of less than 4 do not require treatment or referral for compression ultrasonography, whereas those with scores of 4 or greater should be referred for ultrasonography. All data were forwarded to the investigators. The physicians followed up with their patients five to nine days later.
In the 90 days after study entry, all patients received a questionnaire about signs and symptoms of venous thromboembolism, and subsequent medical information was collected from the primary care physicians if DVT was suspected. The investigators applied the Wells and primary care rules to each patient, with and without inclusion of the d-dimer assay result, and patients were stratified into low- or high-risk groups depending on their score. The primary outcome was the incidence of missed diagnosis of fatal or nonfatal pulmonary embolism or DVT. The number of patients referred for ultrasonography was also noted for each rule.
Results: A total of 1,002 patients were evaluated for DVT. The mean age of participants was 58 years, and 37 percent were men. The most common clinical presentation was leg pain (87 percent) or leg swelling (78 percent) with a median duration of five days. DVT was diagnosed by objective testing in 129 patients who had high-risk scores and positive results on d-dimer assay. Of those patients, 23.2 percent were referred for ultrasonography based on the Wells rule, and 25.4 percent were referred based on the primary care rule. After three months, seven patients with low-risk scores and negative results on d-dimer assay were diagnosed with venous thromboembolism (1.6 percent from the Wells rule, 1.4 percent from the primary care rule), which is a missed diagnosis rate comparable to that of ultrasonography alone. For each 100 patients, using the Wells rule resulted in four more referrals for ultrasonography, whereas using the primary care rule resulted in 22 additional d-dimer assays. In this study, the costs were equivalent.
Conclusion: Both the Wells rule and the primary care rule are safe and effective for ruling out DVT in primary care; however, the primary care rule may be simpler to use.