[Validation of a clinical prediction rule for the diagnosis of deep vein thrombosis of the lower limbs in primary care]. / Validation d'un score de prédiction clinique de la thrombose veineuse profonde des membres inférieurs spécifique à la médecine générale.

PURPOSE: Patients with suspected deep vein thrombosis (DVT) are often managed on an outpatient basis. The aim of the study was to validate a clinical prediction rule specifically for use in primary care to help physicians in their decision to start anticoagulant therapy while awaiting ultrasound examination. PATIENTS AND METHODS: Between September 2007 and October 2008, 194 general practitioners prospectively included patients with clinically suspected DVT without clinically suspected pulmonary embolism. All patients underwent a standardized clinical assessment in order to collect items included in the clinical prediction rule (personal history of venous thromboembolism +1, immobilization in previous month+1, estrogen contraceptive+2, active malignancy+3, swelling of the calf+1, the presence of an alternative diagnosis more likely than that of DVT-3. DVT unlikely if score<2, likely if score≥2). RESULTS: Among the 164 included patients, 56 (34%) had DVT of them 28 (17%) had a proximal DVT. Proportions of confirmed DVT were 29% in the unlikely group and 43% in the likely group against 26% and 63% respectively in the derivation study. CONCLUSIONS: This clinical prediction rule might not fulfill the required conditions to be considered as a usable help in the ambulatory management of DVT. Variations of the cut-off value could enhance its performance.