The Role Of Wells Rule In Patients With Pulmonary Thromboembolism

Objective To compare the prediction capacity of the unstructured clinical estimate for pulmonary embolism(PE) with two clinical decision rules,the original Wells rule and the simplified Wells rule.The clinical probability was categorized as unlikely and likely.And to assess the safety and efficiency of excluding PE by a normal D-dimer combined with an unlikely clinical probability.Method All the patients with suspected PE were included prospectively from August 2007 to August 2008 in Sir run run shaw hospital of medical college of Zhejiang University.Patients were divided into the "unlikely group" when they had a score≤4.0 points by the Wells rule,and "the likely group" when a score＞4.0 points.All of them had an D-dimer test at the same time.Patients with a score≤4.0 points and a negative D-dimer test had no further tests,and PE was considered excluded temporarily.All of them were followed up for three months without anticoagulation or thrombolysis treatment.The other patients,with a score＞4.0 points or a positive D-dimer concentration had further imaging tests.Patients diagnosed as PE by the imaging tests had received anticoagulation or thrombolysis treatment.Patients undiagnosed were also followed up for three months.All the patients were evaluated by the unstructured clinical estimate and the simplified Wells rule.Result We collected 140 patients,36 of whom were diagnosed as PE.The overall prevalence of PE was 25.7%.The sensitivity,specificity and negative prediction value(NPV) was 61.1%,65.4%,75.9%for the unstructured clinical estimate,72.2%,53.8%,87.2%for the original Wells rule and 75.0%,42.3%,86.2%for the simplified Wells rule respectively.The area under the ROC(AUC)of the unstructured clinical estimate,the original and simplified Wells rule was 0.612(95% CI:0.489-0.732),0.749(95%CI:0.650-0.847) and 0.712(95%CI:0.603-0.820) respectively.The PE incidence at three months follow-up in the group of patients with a Wells score≤4 and a negative D-dimer(55 patients altogether) was 1.4%(2 patients altogether).The sensitivity,specificity and negative prediction value was 94.5%,50.9%, 96.4%for a Wells score≤4 and a negative D-dimer versus 80.0%,70.5%,91.4%for a negative D-dimer only.Conclusions This study validates that the original and simplified Wells rule have similar prediction capacity for PE,which perform better than the unstructured clinical estimate.Ruling out patients with suspected pulmonary embolism by a clinical decision rule or D-dimer test alone have considerable false negative possibility.And exclusion suspected pulmonary embolism by a negative D-dimer combined with an unlikely clinical probability appears to be a safe and efficient strategy.