OBJECTIVETo validate the value of age-adjusted D-dimer combined with clinical probability to confirm or exclude deep vein thrombosis (DVT) in elderly patients.

METHODElderly patients ( ≥ 65 years) suspected with DVT were evaluated by Wells score and D-dimer test. All patients underwent ultrasonography examination except for patients with Wells score < 2 and negative D-dimer test results. Conventional cut-off value is 500 µg/L, while age-adjusted cut-off value is set as patient's age×10 µg/L. We compared the sensitivity and specificity using the 2 cut-off values in confirming or excluding the diagnoses of DVT.

RESULTSThe study population consisted of 624 patients [mean age(76.4 ± 19.3) years], DVT was confirmed in 192 (30.8%) patients. Using Wells score model, 326 patients (52.2%) were scored as unlikely DVT and DVT was confirmed by ultrasonography in 44 patients (13.5%), and 298 patients as likely DVT patients and DVT was confirmed in 148 patients (55.0%). The sensitivity, specificity, positive predictive value, and negative predictive value by conventional and age-adjusted D-dimer cut-off value for diagnosing DVT in low-risk patients evaluated by Wells score model were 95.5%, 40.4%, 20.0%, 98.3% and 95.5%, 61.0%, 27.6%, 87.1%, respectively, and which were 89.9%, 67.3%, 73.1%, 87.1% and 89.2%, 89.3%, 89.2%, 89.3%, respectively, in high-risk patients evaluated by Wells score model. Thus, specificity increased about 20% using age-adjusted D-dimer cut-off value compared with conventional D-dimer cut-off value.

CONCLUSIONThe age-adjusted D-dimer cut-off value combined with clinical probability evaluation could increase diagnosing specificity of DVT in elderly patients.