The use of age-related D-dimers to rule out deep vein thrombosis

Date First Published:
August 17, 2014
Last Updated:
August 26, 2015
Report by:
Catherine Taylor, Medical Student (University of Manchester, University Hospital South Manchester, Manchester, UK )
Search checked by:
Laith Sultan, University of Manchester, University Hospital South Manchester, Manchester, UK
Three-Part Question:
In [adults with atraumatic unilateral swollen lower limb], does [the use of an age-adjusted D-dimer cut-off or a single dichotomous cut-off] have [greater sensitivity and specificity for the diagnosis of DVT]?
Clinical Scenario:
An elderly lady who is otherwise well presents at your Emergency Department (ED) with a swollen, red leg. You suspect deep vein thrombosis (DVT), assess her Wells’ score, which is found to be low, and request a D-dimer level. The D-dimer level is marginally raised using the standard diagnostic cut-off. Your current clinical protocol mandates ultrasound scanning in this situation but you have recently heard that there is a natural rise in D-dimer levels with age. The usual normal range may therefore give a high false positive rate in older people. You wonder whether the use of an age-adjusted D-dimer cut-off might allow you to safely rule out DVT without requesting a scan.
Search Strategy:
Medline 1946-Week 1 2014 using the OVID interface:

(exp Venous Thrombosis/OR exp Thrombosis/OR thrombosis.mp. OR venous thrombosis.mp. OR DVT.mp.) AND (exp Fibrin Fibrinogen Degradation Products/OR d dimer$.mp. OR d-dimer$.mp.) AND (age-related.mp. OR age-adjusted.mp. OR age dependent.mp.) Limited to Human and English Language.

Embase 1974–2014 week 22 using OVID interface (exp thrombosis/OR exp vein thrombosis/OR thrombosis.mp. OR venous thrombosis.mp. OR thrombo$.mp. OR dvt.mp. OR deep vein thrombosis.mp.) AND (exp D dimer/OR exp fibrin degradation product/OR d dimer$.mp. OR d-dimer$.mp. OR dimer$.mp.) AND (age-related.mp. OR age-adjusted.mp. OR age-dependent.mp.) Limited to Human and English Language.
Outcome:
In total 59 separate papers were found of which 56 were of insufficient quality for inclusion or did not answer the question set. The remaining three papers are included in the table
Relevant Paper(s):
Study Title Patient Group Study type (level of evidence) Outcomes Key results Study Weaknesses
Validation of two age dependent D-dimer cut-off values for exclusion of deep vein thrombosis in suspected elderly patients in primary care: retrospective, cross sectional, diagnostic analysis. Schouten HJ, Koek HL, Oudega R et al. 2012 Netherlands 406 consecutive primary care patients aged >50 years with a low clinical probability of DVT (Wells’ score ≤1). Retrospective analysis of cohort studies Proportion of patients with D-dimer levels below the conventional (500 µg/L) or age-adjusted (age×x10 µg/L) cut-offs who had DVT excluded by an ultrasound scan repeated at day 7. All ages: Conventional: 42%, 95% CI 38.2% to 46% (n=272/647) Age-adjusted: 47.8%, 95% CI 43.9% to 51.7% (n=309/647). Two studies are used to give these data—the second study adds additional patients to the original cohort of the first to validate the results. No data is given on the sensitivity of either test.
False negative rate. All ages: Conventional: 0.3%, 95% CI 0.04% to 1.1% (n=2/647) Age-adjusted: 0.5%, 95% CI 0.01% to 1.3% (n=3/647).
Using an age-dependent D-dimer cut-off value increases the number of older patients in whom deep vein thrombosis can be safely excluded. Douma RA, Tan M, Schutgens REG et al. 2012 Netherlands Retrospective analysis of five cohorts of hospital outpatients with non-high clinical probability.
Cohort 1:472 patients with Wells’ score ≤2.
Cohort 2:419 patients with clinical probability estimated by treating doctor as being <80%.
Cohort 3:297 patients with Wells’ score ≤2.
Cohort 4:484 patients with Wells’ score ≤2.
Cohort 5:212 hospital outpatients with low clinical probability of DVT (Wells’ score ≤1) 		Retrospective analysis of 5 cohort studies Proportion of patients with low clinical probability who had normal D-dimer levels according to the conventional (<500 µg/L) and age-adjusted (age×10 µg/L in patients aged >50 years) cut-offs. Cohorts 1 to 4 (which classified patients as high or non-high clinical probability): Conventional 42%, 95% CI 40% to 45% (n=707/1672) Age-adjusted 51%, 95% CI 48% to 53% (n=850/1672)Cohort 5 (which classified patients as being likely or unlikely to have DVT): Conventional: 39.2%, n=83/212; Age-adjusted: 47%, n=100/212 Authors have included patients with history of VTE and active malignancy. In cohort 5, 20% of patients had previous venous thromboembolism.
3-month follow-up was by meeting with clinician or telephone.
Cohort 5: no CIs available for the patients below the D-dimer cut-off.
Cohort 5: no confidence intervals available for the patients below the d dimer cut off.

No data is given on the sensitivity of either test.
tProportion of patients with low clinical probability who had normal D-dimer levels according to the conventional (<500 µg/L) and age-adjusted (age×10 µg/L in patients aged >50 years) cut-offs. tCohorts 1 to 4 (which classified patients as high or non-high clinical probability): Conventional 42%, 95% CI 40% to 45% (n=707/1672) Age-adjusted 51%, 95% CI 48% to 53% (n=850/1672)Cohort 5 (which classified patients as being likely or unlikely to have DVT): Conventional: 39.2%, n=83/212; Age-adjusted: 47%, n=100/212 tAuthors have included patients with history of VTE and active malignancy. In cohort 5, 20% of patients had previous venous thromboembolism. 3-month follow-up was by meeting with clinician or telephone. Cohort 5: no CIs available for the patients below the D-dimer cut-off. False negative rate Cohorts 1 to 4: Conventional: 0%, 95% CI 0% to 3.5% (n=0/83) Age-adjusted: 0%, 95% CI 0% to 3.0% (n=0/100)Cohort 5: Conventional: 0.0%, 95% CI 0.0% to 3.5% (n=0/83) Age-adjusted: 0.0%, 95% CI 0.0% to 3.0% (n=0/100).
Evaluation of an age adjusted D-dimer test cut-off point in patients with clinically suspected acute deepvein thrombosis. Tan M, Mol GC, Del Sol AI, et al. 2011 Japan Retrospective study of data of 606 patients with suspected DVT Four hundred patients (66%) were older than 50 years of age. Retrospective cohort study. Proportion of patients who had DVT excluded using conventional and age-adjusted (age×10 µg/L) cut-offs. In patients aged over 50 years with low Wells’ score, conventional cut-off excluded 9.3% with age-adjusted cut-off, DVT excluded in 13.5%. Journal conference abstract only.
False negative rate. The 3 month VTE failure rates were 0% (95% CI 0% to 7.8%) for conventional and 0% (95% CI 0% to 5.4%) for the age adjusted D-dimer cut-off point.
Author Commentary:
There is an increase in the specificity when using the age ×10 μg/L when compared with the conventional 500 μg/L cut-off in patients aged over 50 with low clnical risk. There is no statistically significant increase in the false negative rate.
Bottom Line:
Using an age-adjusted D-dimer cut-off in patients aged over 50 years has been shown to increase the ability to safely exclude deep vein thrombosis in patients with low clinical probability without the need for ultrasound scanning.
References:
  1. Schouten HJ, Koek HL, Oudega R et al. . Validation of two age dependent D-dimer cut-off values for exclusion of deep vein thrombosis in suspected elderly patients in primary care: retrospective, cross sectional, diagnostic analysis.
  2. Douma RA, Tan M, Schutgens REG et al.. Using an age-dependent D-dimer cut-off value increases the number of older patients in whom deep vein thrombosis can be safely excluded.
  3. Tan M, Mol GC, Del Sol AI, et al. . Evaluation of an age adjusted D-dimer test cut-off point in patients with clinically suspected acute deepvein thrombosis.