Within the last decade, the diagnosis of venous thromboembolism (VTE) has evolved with standard clinical riskstratification models, D-dimer testing and helical computed pulmonary tomography, among others [1][2][3][4][5]. The demand for diagnostic image testing in patients with suspected VTE has increased along with the costs related to this approach. To address this problem, many non-invasive management strategies have been proposed to achieve a rapid, safe and costeffective diagnosis. Due to different VTE prevalence, D-dimer performance, and resources available, local validation of these strategies is recommended [6].Despite a few shortcomings, such as method variation, ageand setting-related sensitivity problems and a rather low specificity, D-dimer testing has been successfully employed to rule out VTE in low-risk outpatients [1,2]. Age of the clot, estimated by the duration of symptoms, is recognized as one of the potential clinical factors affecting the performance of Ddimer, but its cut-off limit has not been accurately defined [6]. In previous studies employing D-dimer, the cut-off point has ranged from 3 days to up to 43 days [6,7]. Some have proposed the duration of symptoms cut-off point to be between 7 and 15 days [8,9]. The British Committee for Standards in Haematology recommended that D-dimer testing should be used with caution if the patient had symptoms for over 2 weeks [6].Recently, we retrospectively evaluated the role of D-dimer in a non-invasive clinical approach to VTE. The D-dimer used was a rapid quantitative ELISA method (VIDAS D-dimer; BioMerieux, Marcy l'Etoile, France) with a cut-off level of 500 ng mL )1. Blood samples were collected prior to anticoagulant treatment. All measurements were carried out by an operator unaware of the results of imaging tests. Imprecision of replicate determinations of samples with D-dimer concentration of 550 ng mL )1 was 5.8%, similar to previous studies that employed this method [8,10].We were able to review the medical records from 335 adults admitted to an outpatient emergency clinic at a hospital located in Belo Horizonte, Brazil, between August 2002 and August 2004. We also reviewed the medical records of these patients during a follow-up period of 90 days. In 205 out of 335 patients (61%) a telephone interview was also successfully undertaken. Deep vein thrombosis (DVT) and pulmonary embolism (PE) were objectively confirmed by compression ultrasonography (CUS) and lung scan or helical chest computed tomography, respectively, following standard criteria [3][4][5]. A patient was considered as having a VTE event when suggestive symptoms associated with positive imaging test were present at a first examination or during the 90-day follow-up period. Therefore, a false-negative case would represent a confirmed VTE event in a patient with a negative D-dimer, and a false-positive case would result from a non-confirmed VTE event in a patient with a positive D-dimer.Population age was 64 ± 14.5 years (range 22-94 years). Male to female ratio was 0.35. The pr...