Introduction and objectives The specificity of D-dimer testing in detecting venous thromboembolism (VTE) appears to fall in older patients as D-dimer concentrations increase with age. There has been interest in the use of age-adjusted D-dimer cut-off values to increase the specificity in older patients. The objective of this study was to review the diagnostic accuracy of D-dimer testing in older patients with suspected VTE in our population, comparing conventional and age-adjusted D-dimer cut-off values.

Methods A retrospective study from a large teaching hospital in the UK was undertaken. 389 data episodes were compiled from suspected VTE presentations during two months in 2013. Patients were assessed using a combination of clinical probability scores and D-dimer measurement (D-dimer HS assay, Instrumentation Laboratories). Conventional (230 ng/ml) and age-adjusted (age x 10 ng/ml) cut-off values were applied to patients ≥50 years, and specificity and sensitivity were calculated.

Results (Table 1) Of the 389 presentations, 229 (58.9%) were from patients aged ≥50 years. 13 (11.5%) patients with positive D-dimers using the conventional cut-off, had VTE as confirmed by imaging tests. The sensitivity of the conventional D-dimer cut-off value was 100% in this older cohort, with a specificity of 53.7%. The age x10-adjusted cut-off improved specificity to 84.7%; however sensitivity was markedly reduced to 76.9%, with 3 patients (23.1%) with non-high clinical probability of VTE missed. Further analysis suggested that an age-adjusted cut-off factor of x3 would maintain sensitivity at 100%; however specificity was only 47.7%.

Conclusions We have identified that an age-adjusted cut-off factor of x10 significantly increased D-dimer specificity in older patients; however the sensitivity of this test was unacceptably compromised. A cut-off factor of x3 maintained sensitivity, but specificity was unsatisfactory compared to conventional values, although still higher than in most published series. We conclude that we cannot use an age-adjusted cut-off of x10 in our ≥50 year old population using this assay. Further work is required to identify an appropriate cut-off, concentrating on the >75 year old patients only. This would help to reduce the number of unnecessary tests and anxiety in this vulnerable group of patients.