[Is clinical diagnosis of deep venous thrombosis really unreliable?]

The clinical diagnosis of deep venous thrombosis (DVT) is traditionally regarded as lacking of sensitivity and specificity. As we were not fully convinced by this statement, we assessed the value of two associated clinical signs in outpatients, namely the flapping of the calf and the modified Lowemberg's sign. The examiner has first put forth a diagnosis based on the data of history and/or the presence of clinical signs specific for another kind of disease. Then an objective test of the diagnosis was carried out with a duplex scanner performed by an expert. DVT was observed in 39 subjects out of 102 (38.2%) who where all referred by general practitioners (GP), which confirms the poor diagnostic value of clinics at GP level. Associated positiveness of these two signs yielded 87.2% sensitivity and 88.8% specificity with a positive predictive value of 82.9% and a negative predictive value of 91.8%. The addition of an overall clinical assessment later improved results to 87.2% sensitivity and 96.8% specificity, a positive predictive value of 94.4% and a negative predictive value of 92.4%. In this study, the performance of clinics, if assessed by an angiologist qualified for the diagnosis of DVT, proved to be better than in the common data of the literature. Clinical data, if perfect, should be included in the diagnostic management of DVT.