RESUMO
BACKGROUND: Diagnostic algorithms for deep vein thrombosis (DVT) include D-dimer for its high negative predictive value, thus reducing the need for imaging. Small thrombi may be associated with low D-dimer levels, increasing false negatives. AIM: To assess the sensitivity and thus the false negative rates of standard and age-adjusted D-dimer cut offs for isolated distal DVT (IDDVT) in outpatients. MATERIALS AND METHODS: We enrolled consecutive outpatients with suspected DVT of the lower limbs referring to our vascular emergency department from 2009 to 2018. Patients underwent D-dimer testing (STA, Stago, cut-off: 500 µg/L), pretest clinical probability (PTP) evaluation and complete compression ultrasonography. Follow-up was 3 months. RESULTS: Among 3948 patients (M:1554-39%, median age 69), 486 proximal DVTs (12.3%) and 348 IDDVTs (8.8%) were diagnosed. Median D-dimer was higher in proximal than IDDVT (3960 vs 1400 µgr/L; p = 0.001). The false negative rate of the standard D-dimer cut-off was 2% (95%CI: 0.8-3.2%) for proximal DVT and 14.7% (95% CI: 11-81%) for IDDVT. The false negative rate of the age-adjusted cut-off was 4.9% (3-7%) for proximal DVT and 19.5% (95% CI: 15.4-24.7%) for IDDVT. CONCLUSIONS: Small calf thrombi are associated with low D-dimer levels, and age-adjusted D-dimer may be below the cut-off more frequently in subjects with IDDVT than standard cut-off D-dimer, although such D-dimer levels might exclude IDDVT that require treatment.