What is the predictive value of a low probability Ventilation-Perfusion (VQ) scan for ruling out Pulmonary Embolism (PE)?

Predictive Value of Low Probability VQ Scan

A low probability VQ scan combined with low clinical probability has a negative predictive value of 96% (4% rate of PE), which is insufficient to rule out PE on its own and requires additional testing such as D-dimer or lower extremity ultrasound to safely exclude pulmonary embolism. 1

Clinical Probability Dramatically Alters Predictive Value

The interpretation of a low probability VQ scan is critically dependent on pretest clinical probability:

• Low clinical probability patients: 4% rate of PE 1
• Intermediate clinical probability patients: 16% rate of PE 1
• High clinical probability patients: 40% rate of PE 1

This wide variation demonstrates that clinical risk stratification using validated tools (Wells score or revised Geneva score) is essential before interpreting VQ scan results. 1

When Low Probability VQ Scan Can Exclude PE

Level A Evidence (Strongest Recommendation)

A normal/near-normal perfusion scan reliably excludes clinically significant PE in patients with low-to-moderate pretest probability, with a negative likelihood ratio of approximately 0.1 and subsequent PE rates <1% on long-term follow-up. 1

Level B Evidence for Low Probability Scans

In patients with low-to-moderate pretest probability AND a low probability (non-diagnostic) VQ scan, PE can be excluded using one of these additional tests: 1

1. Negative quantitative D-dimer (turbidimetric or ELISA)
2. Negative whole blood qualitative D-dimer with Wells score ≤4
3. Negative single bilateral venous ultrasound (for low probability patients)
4. Negative serial bilateral venous ultrasound (for moderate probability patients)

The 2020 ESC guidelines support this approach, recommending that a non-diagnostic VQ scan combined with negative proximal compression ultrasound can exclude PE in patients with low clinical probability. 1

Very Low Probability Interpretation

Recent data from PIOPED II established "very low probability" criteria with improved predictive value: 2

Very low probability VQ scan criteria include:

• Nonsegmental perfusion abnormalities
• Perfusion defect smaller than corresponding chest X-ray lesion
• ≥2 matched V/Q defects with normal chest X-ray
• 1-3 small segmental perfusion defects (<25% of segment)
• Stripe sign around perfusion defects 3, 2

When combined with low clinical probability assessment, very low probability VQ scans have a positive predictive value of only 3.1% (negative predictive value 96.9%), which reliably excludes PE. 2 In women ≤40 years with low clinical probability, this drops to 2% PPV. 2

Critical Pitfalls to Avoid

Never use a low probability VQ scan alone to exclude PE in patients with intermediate or high clinical probability, as the PE rate reaches 16-40% in these groups despite the "low probability" scan result. 1

Do not skip lower extremity ultrasound in patients with non-diagnostic scans, as asymptomatic proximal DVT is found in a significant proportion of PE patients and provides de facto evidence of venous thromboembolism requiring anticoagulation. 1

Avoid misclassifying intermediate probability scans as low probability, as 60% of PIOPED patients with PE had scans read as "low" or "intermediate" probability, making accurate categorization essential. 1

Practical Algorithm

For patients with low probability VQ scan:

1. First, determine pretest clinical probability using Wells score or revised Geneva score 1

2. If low clinical probability (Wells <2):

   • Obtain quantitative D-dimer OR
   • Obtain bilateral lower extremity ultrasound
   • If negative → PE excluded 1

3. If intermediate clinical probability (Wells 2-6):

   • Obtain quantitative D-dimer AND/OR
   • Obtain serial bilateral lower extremity ultrasound
   • If both negative → PE excluded 1

4. If high clinical probability (Wells >6):

   • Low probability VQ scan is insufficient
   • Proceed to CT pulmonary angiography or pulmonary angiography 1

A retrospective study of 536 patients with low probability VQ scans found zero cases of PE on 6-month follow-up, suggesting that in appropriately selected low-risk patients, the negative predictive value approaches 100%. 1 However, this requires rigorous clinical probability assessment and appropriate use of adjunctive testing as outlined above.