What is the sensitivity of ventilation-perfusion (V/Q) scintigraphy for detecting acute pulmonary thromboembolism?

Sensitivity of V/Q Scintigraphy for Acute Pulmonary Embolism

The sensitivity of V/Q scintigraphy for detecting acute pulmonary embolism varies substantially depending on the reporting criteria used, ranging from approximately 60-95%, with high-probability scans achieving 77-95% sensitivity but only identifying PE in a minority of patients. 1, 2

Performance Characteristics by Scan Category

High-Probability Scans

• High-probability V/Q scans correctly identify pulmonary embolism in 86-92% of cases when PE is present 1
• The PIOPED II study demonstrated a sensitivity of 77.4% (95% CI: 69.7-85.0%) for high-probability scan findings 2
• However, only approximately 30-50% of patients with suspected PE receive a diagnostic result (either normal or high probability), with the remainder falling into non-diagnostic categories 1

Normal/Low-Probability Scans

• Normal perfusion scans have 96% accuracy in excluding PE, making them highly reliable for ruling out the diagnosis 1
• Low-probability scans demonstrate 86% accuracy in excluding PE 1
• The specificity of very low probability or normal scan findings reaches 97.7% (95% CI: 96.4-98.9%) 2

Critical Limitation: Non-Diagnostic Results

The major weakness of planar V/Q scintigraphy is the high rate of non-diagnostic scans (approximately 50% of cases), which provide no discriminatory value between PE and non-PE. 1

• In large studies using conventional reporting criteria, many patients fell into the indeterminate category requiring further imaging 1
• Agreement among scan readers is good for high-probability and normal scans (>90% agreement) but poor for indeterminate and low-probability scans (70-75% agreement) 1

Improved Performance with Modern Techniques

V/Q SPECT

• V/Q SPECT substantially improves diagnostic performance compared to planar imaging, with sensitivity of 92-100% and specificity of 87-91% 3, 4
• The rate of non-diagnostic scans decreases to less than 3% with SPECT imaging 1
• Using a diagnostic cutoff of at least 1 segmental or 2 subsegmental mismatches achieves sensitivity of 92% and specificity of 91% 3
• One study demonstrated V/Q SPECT sensitivity of 100% compared to only 64% for planar V/Q scintigraphy 4

Simplified Reporting Criteria

• Using simplified criteria (>0.5 segments of V/Q mismatch as positive) can achieve 95-98% sensitivity with improved inter-observer agreement (kappa 0.74-0.83) 5
• The PISAPED criteria using perfusion scanning alone showed lower sensitivity (60.8%) but eliminated non-diagnostic results 6
• Modified PISAPED criteria combining ventilation and perfusion improved sensitivity to 83.8% with 89.1% specificity 6

Clinical Context Matters

The predictive value of V/Q scintigraphy is heavily influenced by pre-test clinical probability:

• With a negative V/Q SPECT result, post-test probability of PE is only 1.0% for low clinical probability but rises to 11.9% for high clinical probability 3
• High-probability scans may not be sufficient to confirm PE in patients with low clinical probability, requiring further testing 1
• In patients with low-probability lung scans and low clinical probability, combined with negative proximal compression ultrasound, the 3-month thromboembolic risk is only 1.7% 1

Situations Where V/Q Scintigraphy May Be Misleading

Interpretation becomes difficult or unreliable in several clinical scenarios: 1

• Previous pulmonary embolism (unless follow-up scan performed)
• Left heart failure (causes regional perfusion variations)
• Chronic obstructive airways disease (local ventilation/perfusion variations)
• Lung fibrosis (patchy unmatched defects)
• Proximal lung cancer (vascular occlusion with preserved ventilation)

Preferred Clinical Applications

V/Q scintigraphy should preferentially be used in: 1

• Outpatients with low clinical probability and normal chest X-ray
• Young patients, particularly women (lower radiation exposure: ~2 mSv vs 3-10 mSv for CTPA)
• Pregnant women
• Patients with contrast medium-induced anaphylaxis history
• Severe renal failure patients
• Patients with contraindications to iodinated contrast

Bottom Line for Clinical Practice

While planar V/Q scintigraphy has moderate sensitivity (77-95% for high-probability scans), its clinical utility is severely limited by the 50% rate of non-diagnostic results 1. V/Q SPECT represents a substantial improvement with 92-100% sensitivity and <3% non-diagnostic rate, making it the preferred nuclear medicine approach when available 3, 4. However, CTPA remains the first-line imaging modality in most centers due to its excellent accuracy (83% sensitivity, 96% specificity), low rate of inconclusive results (3-5%), and ability to provide alternative diagnoses 1.