Age-adjusted D-dimer for VTE

Clinical background

The D-dimer has high sensitivity but low specificity for venous thromboembolism, and its specificity falls further with increasing age. A substantial proportion of older patients have a raised D-dimer without thromboembolism, meaning that the conventional threshold of 500 ng/mL leads to a high proportion of false-positive results and unnecessary imaging in precisely the patient group at greatest risk. The age-adjusted D-dimer aims to increase specificity in patients over 50 years and thereby increase the proportion of patients in whom VTE can be excluded with the D-dimer and clinical probability alone, without the need for further imaging. The instrument is not a stand-alone diagnostic test and must be combined with an assessed clinical pre-test probability.

Calculating the age-adjusted D-dimer threshold

The threshold is calculated as:

$$\text{Threshold} = \begin{cases} 500 \text{ ng/mL} & \text{if age} \leq 50 \ \text{age} \times 10 \text{ ng/mL} & \text{if age} > 50 \end{cases}$$

The variables are age in whole years and the D-dimer (FEU) in ng/mL, expressed in fibrinogen equivalent units. A D-dimer value below the age-adjusted threshold is regarded as negative and can, at a low or intermediate pre-test probability, be used to exclude VTE. For assays reporting in DDU (D-dimer units), both the value and the thresholds are halved, so that 500 ng/mL FEU corresponds to 250 ng/mL DDU and the age-adjusted threshold becomes $\text{age} \times 5$ ng/mL DDU.

The derivation cohort for pulmonary embolism was the ADJUST-PE study, a prospective multicentre study in 19 centres in Belgium, France, the Netherlands and Switzerland (2010 to 2013) that included 3,346 consecutive emergency patients with suspected pulmonary embolism [1]. The prevalence of PE was 19%. Of the 2,898 patients with a low or unlikely clinical probability, 817 (28.2%) had a D-dimer below 500 and 337 (11.6%) had a D-dimer between 500 and their age-adjusted threshold. Of the patients in this zone, 331 were followed without anticoagulation and 1 developed thromboembolism over three months (failure rate 0.3%, 95% CI 0.1 to 1.7%) [1].

The derivation cohort for deep vein thrombosis was the ADJUST-DVT study (2026), a prospective multicentre study in 27 centres in Belgium, Canada, France and Switzerland (2015 to 2022) with 3,205 emergency patients with suspected DVT [2]. The prevalence of DVT was 14%, the median age 59 years and 54% were women. Of 2,169 patients with a low or unlikely clinical probability, 531 (24.5%) had a D-dimer below 500 and 161 (7.4%) fell within the age-adjusted zone. In this group there were no thromboembolic events over three months (0%, 95% CI 0 to 2.3%) [2].

Interpretation in practice

The age-adjusted D-dimer applies only at a low or intermediate pre-test probability. At a high pre-test probability the D-dimer should not be used to exclude VTE, and the patient proceeds directly to imaging regardless of the D-dimer value.

In practice, its use markedly increases the proportion of older patients in whom VTE can be excluded with the D-dimer alone. In ADJUST-PE, the proportion of patients aged 75 years and over with a low or unlikely probability in whom PE was excluded rose from 6.4% to 29.7% when the age-adjusted threshold was used instead of 500, with no additional false-negative results [1]. The corresponding increase in ADJUST-DVT was from 8.7% to 26.1% among patients aged 75 years and over [2].

Validation and performance

In ADJUST-PE the failure rate in the age-adjusted zone was 0.3% (1 of 331 patients) at three months of follow-up [1]. In ADJUST-DVT the corresponding failure rate was 0% (0 of 161 patients, 95% CI 0 to 2.3%) [2]. A post hoc analysis of the PALLADIO study, a multicentre study of 1,162 patients with suspected DVT, showed that age adjustment reduced the need for compression ultrasound by 5.1 percentage points compared with a fixed threshold, at a VTE incidence during follow-up of 0.24% in the group with a low probability and a negative D-dimer [3].

There is no consensus among international guidelines on which D-dimer threshold should be used, but the age-adjusted threshold has been incorporated into several of them. In its 2015 Best Practice Advice, the American College of Physicians recommends an age-adjusted threshold (age × 10 ng/mL) in patients over 50 years and advises against imaging at values below this threshold [4]. A 2024 review of international guidelines concludes that age-adjusted and pre-test-probability-adjusted D-dimer thresholds are regarded as safe across predefined patient subgroups, but the authors point out that clinicians must be aware that different D-dimer assays have different performance characteristics [5].

Limitations

The instrument requires a high-sensitivity D-dimer assay. The Clinical and Laboratory Standards Institute recommends immunoassays with a sensitivity of at least 97% and a negative predictive value of at least 98% for VTE to be safely excluded [5]. If an assay of lower sensitivity is used, the age-adjusted strategy risks missing diagnoses.

At a high pre-test probability the D-dimer should never be used for exclusion, whatever the value. A normal D-dimer in a patient with a high probability does not exclude VTE, and imaging should be performed regardless [4].

The threshold was derived and validated in emergency patients with suspected PE or DVT. It lacks validation in pregnancy, and since the D-dimer is physiologically raised during pregnancy, age adjustment is not applicable there. In patients with active cancer, severe inflammation, sepsis or recent surgery, the D-dimer may be raised for reasons other than thromboembolism, which lowers specificity but not sensitivity. Age adjustment does increase specificity in these groups as well, but the proportion of false positives remains high.

The D-dimer cannot be used to confirm VTE. A raised value, whether above the age-adjusted threshold or above 500, is not specific for thromboembolism and should not be interpreted as diagnostic without further imaging.

The age-adjusted strategy is particularly valuable in patients over 75 years, in whom the fixed threshold excludes only a small minority. For patients under 50 years, age adjustment offers no advantage and the conventional threshold of 500 ng/mL should be used unchanged.

References

1. Righini M et al. Age-adjusted D-dimer cutoff levels to rule out pulmonary embolism: the ADJUST-PE study. JAMA 2014;311(11):1117-1124. PMID: 24643601
2. Le Gal G et al. Age-adjusted D-dimer cutoff levels to rule out deep vein thrombosis. JAMA 2026;335(5):416-424. PMID: 41490105
3. Riva N et al. Age-adjusted D-dimer to rule out deep vein thrombosis: findings from the PALLADIO algorithm. J Thromb Haemost 2018;16(2):271-278. PMID: 29125695
4. Raja AS et al. Evaluation of patients with suspected acute pulmonary embolism: best practice advice from the Clinical Guidelines Committee of the American College of Physicians. Ann Intern Med 2015;163(9):701-711. PMID: 26414967
5. Fan BE, Lippi G, Favaloro EJ. D-dimer levels for the exclusion of pulmonary embolism: making sense of international guideline recommendations. J Thromb Haemost 2024;22(3):604-608. PMID: 38135252