What is the probability of pulmonary embolism (PE) with a negative D-dimer (D-dimer) result?

Probability of Pulmonary Embolism with Negative D-dimer

The probability of pulmonary embolism (PE) with a negative D-dimer result is extremely low at approximately 0.1-0.5% in patients with low to intermediate clinical probability, making it a safe exclusion strategy in these populations. 1

Clinical Probability Assessment and D-dimer Testing

The diagnostic approach to suspected PE requires two key components:

1. Clinical probability assessment using validated tools:

   • Wells score (traditional or simplified "PE unlikely/likely" version)
   • Revised Geneva score

2. D-dimer testing with consideration of assay sensitivity:

   • Highly sensitive assays (ELISA-based): Safe for excluding PE in low OR moderate probability patients
   • Moderately sensitive assays (latex agglutination): Safe only for low probability patients

Risk Stratification Results

When using the Wells score 1:

• Low probability: ~3.6% have PE
• Moderate probability: ~20.5% have PE
• High probability: ~66.7% have PE

With the dichotomized Wells approach ("PE unlikely" vs "PE likely") 1:

• PE unlikely (score ≤4): ~7.8% have PE
• PE likely (score >4): ~40.7% have PE

Negative Predictive Value of D-dimer Testing

The safety of excluding PE with a negative D-dimer varies by assay type and clinical probability:

• Highly sensitive assays (ELISA-based) 1, 2:

  • 3-month thromboembolic risk with negative result in low/moderate probability: 0.1% (95% CI: 0-0.5%)
  • Negative predictive value: 99.5-99.9%

• Moderately sensitive assays 1:

  • 3-month thromboembolic risk with negative result in low probability: 0.2-0.6%
  • Not validated for moderate probability patients

A systematic review of management outcome studies found that the 3-month thromboembolic risk in patients with non-high clinical probability and negative VIDAS D-dimer was only 0.14% (95% CI: 0.05-0.41%) 2.

Special Considerations

Age-Adjusted D-dimer

• For patients over 50 years, an age-adjusted D-dimer cutoff (age × 10 μg/L) improves specificity without reducing safety 3

Reduced Specificity Populations

D-dimer has reduced specificity (more false positives) in 1, 3:

• Elderly patients (>80 years: specificity may be as low as 10%)
• Cancer patients
• Hospitalized patients
• Pregnant women

High Clinical Probability

• Patients with high clinical probability should proceed directly to imaging regardless of D-dimer result 1

Practical Algorithm

1. Assess clinical probability using Wells score or Geneva score
2. If low or intermediate probability (or "PE unlikely"):
   • Perform D-dimer testing
   • If negative: Safely exclude PE (0.1-0.5% risk)
   • If positive: Proceed to imaging (CT pulmonary angiography)
3. If high probability (or "PE likely"):
   • Proceed directly to imaging regardless of D-dimer

Common Pitfalls

• Using D-dimer without clinical probability assessment
• Relying on moderately sensitive D-dimer assays for moderate probability patients
• Failing to consider age-adjusted D-dimer cutoffs in older patients
• Using D-dimer in patients with conditions that elevate results (cancer, pregnancy, hospitalization)
• Overreliance on D-dimer in high clinical probability patients

The most recent evidence from a 2019 study suggests that using higher D-dimer thresholds (up to 1000 ng/mL) may be safe in patients with low clinical probability, potentially reducing unnecessary imaging by 17.6% 4.