Document resource
Background The age-adjusted D-dimer (AADD) cut-off (age ×10 ng/mL) improves specificity in pulmonary embolism (PE) diagnosis but requires validation in heterogeneous populations. This study evaluates AADD’s performance in a diverse cohort and explores increasing the age-adjustment multiplier to enhance specificity.Methods A retrospective analysis included 277 patients (mean age 63.9 years; 18.1% PE prevalence) undergoing CT pulmonary angiography and D-dimer testing at Oxford University Hospitals (2018). Exclusions included trauma, perioperative states, and chronic PE. Conventional D-dimer (500 ng/mL) was compared to AADD (multipliers ×10 and ×11). Sensitivity, specificity, negative/positive predictive values (NPV/PPV), and accuracy were calculated.Results Both conventional D-dimer and AADD demonstrated 100% sensitivity and NPV. AADD (×10) increased specificity from 11.45% (95% CI: 7.6–16.3%) to 20.70% (15.6–26.6%), correctly reclassifying 9% of PE-negative patients. Raising the multiplier to ×11 further improved specificity to 26.43% (20.8–32.7%) without compromising sensitivity. Accuracy increased from 27.44% (conventional) to 39.71% (×11 multiplier). Multipliers >12 reduced sensitivity, but ×11 maintained 100% sensitivity.Abstract P279 Figure 1Discussion/Conclusion In a heterogeneous population, AADD safely ruled out PE while enhancing specificity, reducing unnecessary imaging. Increasing the multiplier to ×11 provided additional gains in specificity (14.98% vs. conventional; 5.73% vs. ×10) without missed diagnoses. These findings support refining AADD thresholds to optimize diagnostic efficiency. While single-centre data and sample size limit generalizability, results align with prior studies, advocating for multiplier adjustments to improve clinical utility. Adoption of AADD with a ×11 multiplier could enhance cost-effectiveness and patient safety, particularly in older populations vulnerable to contrast-related complications. Further multicentre studies are warranted to validate these findings.