What are the causes of elevated D-dimer (D-dimer) levels and what diagnostic tests should be performed?

Causes of Elevated D-dimer and Diagnostic Approach

D-dimer elevation occurs in numerous thrombotic and non-thrombotic conditions, and your diagnostic approach must be guided by clinical probability assessment using validated tools before ordering any imaging studies. 1

Understanding D-dimer Physiology

D-dimer is a fibrin degradation product resulting from plasmin-mediated breakdown of cross-linked fibrin, indicating simultaneous activation of coagulation and fibrinolysis with a half-life of approximately 16 hours. 1 While it has high sensitivity (96%) for thrombotic disease, specificity is very low (35%), meaning elevated levels commonly occur without venous thromboembolism (VTE). 1

Major Causes of D-dimer Elevation

Thrombotic Conditions

• Pulmonary embolism - Most common thrombotic cause, present in 32% of patients with extremely elevated D-dimer (>5000 μg/L) 2
• Deep vein thrombosis - Present in 13% of patients with extremely elevated D-dimer 2
• Acute aortic dissection - D-dimer >0.5 μg/mL has 94-100% sensitivity when measured within 24 hours of symptom onset 1, 3
• Cerebral venous thrombosis - Though D-dimer levels decline over time from symptom onset 1
• Arterial thrombosis including myocardial infarction 1

Non-Thrombotic Conditions

• Sepsis - Present in 24% of patients with extremely elevated D-dimer 2
• Active malignancy - Present in 29% of patients with extremely elevated D-dimer 2
• Disseminated intravascular coagulation (DIC) - Characterized by markedly elevated D-dimer 1
• Recent trauma or surgery - Present in 24% of patients with extremely elevated D-dimer 2
• COVID-19 - Elevated D-dimer predicts disease severity and mortality, with levels >2.12 μg/mL associated with death 1, 3
• Advanced age - D-dimer naturally increases with age, with specificity dropping to 10% in patients >80 years 1
• Pregnancy - Physiologic elevation occurs, peaking in third trimester (up to 2 μg/mL may be normal) 3
• Acute inflammatory states including ARDS 1
• Recent hospitalization or severe infection 1
• Liver disease with impaired clearance 1

Critical Clinical Thresholds

D-dimer levels 3-4 times above normal (>1500-2000 ng/mL) warrant hospital admission consideration even without severe symptoms, as this signifies substantial thrombin generation and increased mortality risk. 3 In the context of suspected VTE, 89% of patients with extremely elevated D-dimer (>5000 μg/L) had VTE, sepsis, and/or cancer. 2

Diagnostic Algorithm: What Tests to Perform

Step 1: Calculate Clinical Probability BEFORE Ordering D-dimer

Never order D-dimer without first assessing clinical probability—this is the most common pitfall. 3 Use validated scoring systems:

• Wells Score for DVT/PE - Assigns points for active cancer, paralysis/recent immobilization, bedridden status, localized tenderness, leg swelling, calf swelling, pitting edema, collateral veins, previous DVT, and whether alternative diagnosis is less likely 3
• Revised Geneva Score - Alternative validated tool 4, 1

Step 2: Apply D-dimer Testing Based on Clinical Probability

For LOW clinical probability patients:

• Order highly sensitive D-dimer (ELISA preferred with 97-98% sensitivity) 3
• If D-dimer <500 ng/mL (or age-adjusted cutoff for patients >50 years: age × 10 ng/mL), VTE is excluded with 3-month thromboembolic risk <1% 4, 1, 3
• If D-dimer positive, proceed to imaging 1

For INTERMEDIATE clinical probability patients:

• D-dimer can be used to exclude VTE if negative 3
• If positive or unavailable, proceed directly to imaging 3

For HIGH clinical probability patients:

• Proceed directly to imaging WITHOUT D-dimer testing, as D-dimer has insufficient negative predictive value in this population 1, 3

Step 3: Select Appropriate Imaging Based on Suspected Condition

For suspected pulmonary embolism:

• Multidetector CT pulmonary angiography (CTPA) is the imaging modality of choice 1, 3
• If D-dimer >2000 ng/mL, proceed directly to CTPA even in "unlikely" clinical probability patients due to 36% positive predictive value 3

For suspected deep vein thrombosis:

• Proximal compression ultrasonography or whole-leg ultrasound 3
• In pregnancy with elevated D-dimer, perform lower limb compression ultrasonography first 1

For suspected acute aortic dissection:

• CT angiography of chest/abdomen/pelvis 3
• D-dimer >0.5 μg/mL within 24 hours has 94-100% sensitivity, but never use D-dimer to rule out dissection in high-risk patients 1, 3
• High-risk features include Marfan syndrome, family history of aortic disease, known aortic valve disease, previous aortic surgery, abrupt severe ripping/tearing pain, pulse deficit, blood pressure differential, focal neurological deficit, new aortic murmur, or hypotension 1

For markedly elevated D-dimer (>5000 ng/mL) without clear source:

• Consider occult malignancy (present in 29% of cases), sepsis, or DIC 3, 2
• Assess complete blood count, coagulation studies (PT, aPTT), and fibrinogen levels to evaluate for DIC 3

Special Populations and Important Caveats

Avoid D-dimer testing in these populations due to dramatically reduced specificity:

• Hospitalized patients 1
• Cancer patients 1
• Post-surgical patients 1
• Elderly patients >80 years (specificity drops to 10%) 1, 3
• In these populations, proceed directly to imaging based on clinical probability alone 3

Age-adjusted cutoffs for patients >50 years:

• Use cutoff of age × 10 ng/mL to improve specificity from 34% to 46% while maintaining >97% sensitivity 1, 3
• This increases the proportion of patients in whom PE can be safely excluded from 6.4% to 30% 3

Critical pitfalls to avoid:

• D-dimer levels decline over time from symptom onset, potentially causing false-negative results in delayed presentations 1
• Thrombosed false lumen in aortic dissection or intramural hematoma may produce falsely low D-dimer 1
• Point-of-care D-dimer assays have lower sensitivity than laboratory-based tests and should only be used in low pretest probability patients 3
• Different laboratories use different reporting units (FEU vs DDU, with FEU approximately 2-fold higher), making cutoff values non-transferable between institutions 3
• Heterophilic antibodies can rarely cause falsely elevated D-dimer not conforming to clinical evidence 5