How to rule out pulmonary embolism (PE)?

How to Rule Out Pulmonary Embolism

Use validated clinical prediction rules (Wells or revised Geneva score) to stratify pretest probability, then apply PERC criteria for low-risk patients or obtain high-sensitivity D-dimer testing (with age-adjusted cutoffs for patients >50 years) to safely exclude PE without imaging. 1

Step 1: Assess Pretest Clinical Probability

• Apply a validated clinical prediction rule such as the Wells score or revised Geneva score to categorize patients into low (10% PE prevalence), intermediate (30%), or high (~65%) probability groups 1, 2
• The revised Geneva score assigns points for: previous PE/DVT (1 point), heart rate 75-94 bpm (1 point) or ≥95 bpm (2 points), surgery/fracture within past month (1 point), hemoptysis (1 point), active cancer (1 point), unilateral lower-limb pain (1 point), pain on deep venous palpation with unilateral edema (1 point), and age >65 years (1 point) 1
• Clinical gestalt performs equally well as formal prediction rules when used by experienced clinicians for risk stratification 1
• Using the two-level classification, PE-unlikely is defined as Wells score 0-4 or revised Geneva score 0-2, with PE prevalence of approximately 12% 1

Step 2: Apply PERC Rule for Low-Risk Patients

• For patients with LOW pretest probability only, apply the Pulmonary Embolism Rule-Out Criteria (PERC) to identify those who need no further testing 1
• All eight PERC criteria must be met: age <50 years, pulse <100 bpm, SaO₂ >94%, no unilateral leg swelling, no hemoptysis, no recent trauma or surgery, no history of VTE, and no oral hormone use 1
• If all PERC criteria are met, PE can be safely excluded without D-dimer testing or imaging, as the risk of PE (<1%) is lower than the risks of testing 1
• If any PERC criterion is not met, proceed to D-dimer testing 1

Important Caveat About PERC

• PERC should only be applied to patients already determined to have LOW clinical probability by a prediction rule or gestalt assessment 1
• One external validation study found a 6.4% PE prevalence among PERC-negative patients with low pretest probability, suggesting PERC may not be universally safe in all populations, particularly those with higher baseline PE prevalence 3
• The pooled sensitivity of PERC across 12 studies was 97% with specificity of 22%, meaning PERC can safely avoid 22% of D-dimer tests 1

Step 3: D-Dimer Testing Strategy

For Low or Intermediate Pretest Probability Patients

• Obtain high-sensitivity D-dimer (ELISA or ELISA-derived assays with ≥95% sensitivity) as the initial diagnostic test in patients with low pretest probability who do not meet all PERC criteria, or in patients with intermediate pretest probability 1
• Do NOT use imaging studies as the initial test in patients with low or intermediate pretest probability 1

Age-Adjusted D-Dimer Cutoffs

• For patients >50 years old, use age-adjusted D-dimer thresholds (age × 10 ng/mL) rather than the generic 500 ng/mL cutoff 1
• Age-adjusted cutoffs increase the number of patients in whom PE can be excluded from 6.4% to 30% without additional false-negative findings, while maintaining sensitivity >97% 1, 4
• D-dimer specificity decreases to only 10% in patients >80 years using standard cutoffs, making age adjustment particularly important in elderly patients 1

Interpreting D-Dimer Results

• If D-dimer is below the age-adjusted cutoff (or <500 ng/mL if age ≤50 years), PE is excluded and no imaging is needed 1
• If D-dimer is elevated, proceed to imaging with CT pulmonary angiography 1
• The negative predictive value of normal D-dimer combined with low clinical probability is 99% for PE 4, 5

D-Dimer Limitations

• D-dimer has high sensitivity (96%) but low specificity (35%) for PE, making it excellent for ruling out but poor for ruling in PE 4
• D-dimer is frequently elevated regardless of PE status in hospitalized patients, post-surgical patients, pregnant women, cancer patients, and those with severe infection or inflammatory disease 1
• The number needed to test rises from 3 in general emergency department populations to >10 in these special populations 1

Step 4: Imaging for High Pretest Probability

• For patients with HIGH pretest probability, proceed directly to imaging without D-dimer testing, as a negative D-dimer will not obviate the need for imaging 1
• CT pulmonary angiography (CTPA) is the preferred imaging modality when available and no contraindication to contrast exists 1, 2
• Reserve ventilation-perfusion (V/Q) scans for patients with contraindication to CTPA (renal insufficiency, contrast allergy) or when CTPA is unavailable 1
• V/Q scanning is associated with lower radiation exposure than CTPA and may be preferred in younger patients and pregnancy 5

Common Pitfalls to Avoid

• Never apply PERC to patients with intermediate or high pretest probability - it is only validated for low-risk patients 1
• Never use a positive D-dimer alone to diagnose PE - confirmation with imaging is always required 4
• Do not obtain D-dimer in high pretest probability patients - proceed directly to imaging 1
• Avoid ordering D-dimer in populations where results are likely positive regardless of PE status (hospitalized, post-surgical, pregnant patients) as specificity is too low to be useful 1
• Do not use point-of-care D-dimer assays in high pretest probability patients due to lower sensitivity (88% vs 95% for ELISA) 4
• Remember that many population-level risk factors (family history, postpartum period, lower-extremity fracture, pregnancy itself, congestive heart failure, stroke) are not included in validated prediction rules because they did not add to predictive performance 1

Algorithm Summary

1. Assess pretest probability using Wells or revised Geneva score 1
2. If LOW probability: Apply PERC criteria 1
   • All 8 PERC criteria met → PE excluded, no further testing 1
   • Any PERC criterion not met → Obtain D-dimer 1
3. If INTERMEDIATE probability: Obtain D-dimer 1
4. D-dimer interpretation:
   • Normal (age-adjusted if >50 years) → PE excluded, no imaging 1
   • Elevated → Proceed to CTPA 1
5. If HIGH probability: Proceed directly to CTPA (or V/Q if contraindicated), do not obtain D-dimer 1