What is the sensitivity of pulmonary scintography for diagnosing acute pulmonary embolism?

Sensitivity of Pulmonary Scintigraphy for Acute Pulmonary Embolism

Ventilation-perfusion (V/Q) lung scintigraphy has a high sensitivity of 98% for detecting acute pulmonary embolism, but with very low specificity of only 10% when considering all abnormal scans. 1

Diagnostic Performance by Scan Interpretation Category

The sensitivity of V/Q scanning varies significantly based on how the results are classified:

• High-probability scans:

  • Sensitivity: 41% (only a minority of PE patients have high-probability scans)
  • Specificity: 97%
  • Positive predictive value: 88% (102 of 116 patients with high-probability scans had PE) 1

• Normal perfusion scans:

  • Effectively rules out PE (safe to withhold anticoagulant therapy) 2
  • Negative predictive value: ~100% for normal perfusion scans 2

• Non-diagnostic scans (intermediate or low probability):

  • Major limitation: Up to 50% of V/Q scans are non-diagnostic 2
  • Intermediate-probability scans: 33% had PE (105 of 322 patients) 1
  • Low-probability scans: Approximately 12% had PE 1

Factors Affecting Diagnostic Performance

• Clinical probability integration:

  • Combining clinical assessment with V/Q scan results significantly improves diagnostic accuracy 1
  • A high-probability V/Q scan confirms PE in patients with high clinical probability 2
  • However, a high-probability scan may not be sufficient to confirm PE in patients with low clinical probability 2

• Technical considerations:

  • Conventional planar imaging has high interobserver variability (70-75% agreement) for indeterminate and low-probability scans 2
  • Agreement is better (>90%) for high-probability and normal scans 2

• Limitations in specific patient populations:

  • Previous pulmonary embolism (unless follow-up scan performed) 2
  • Left heart failure (causes regional variations in pulmonary perfusion) 2
  • Chronic obstructive airways disease 2
  • Lung fibrosis 2
  • Proximal lung cancer 2

Newer Techniques

• SPECT (Single-Photon Emission CT) imaging:

  • May reduce non-diagnostic scans to as low as 0-5% 2
  • Allows binary interpretation ("PE" versus "no PE") 2
  • However, validation in prospective management outcome studies is lacking 2

• Perfusion-only scanning:

  • May be acceptable in patients with normal chest X-ray 2
  • Any perfusion defect in this situation would be considered a mismatch 2

Comparison with Other Modalities

• CT Pulmonary Angiography (CTPA):

  • Now preferred over V/Q scanning in most settings 3
  • Better at detecting central emboli, but may miss small subsegmental emboli 4
  • Provides alternative diagnoses when PE is excluded 3

• Pulmonary Angiography:

  • Historical gold standard 2
  • Invasive with significant risks (0.5% mortality, 1% major complications) 2

Clinical Implications

• V/Q scanning remains valuable in specific populations:

  • Patients with contraindications to CTPA (severe renal failure, contrast allergy) 3
  • Pregnant women (lower radiation exposure than CTPA) 3
  • Young patients, particularly females 2

• The high sensitivity but low specificity of abnormal V/Q scans means:

  • A normal scan reliably excludes PE 2
  • Abnormal scans often require additional testing to confirm diagnosis 2

• When interpreting V/Q scans, the three-tier classification is recommended:

  • Normal scan (excluding PE)
  • High-probability scan (diagnostic of PE in most patients)
  • Non-diagnostic scan (requiring further testing) 2

In summary, while V/Q scanning has excellent sensitivity for detecting PE when considering all abnormal scans (98%), its clinical utility is limited by poor specificity and high rates of non-diagnostic results. The diagnostic value is maximized when integrated with clinical probability assessment and when newer techniques like SPECT are employed.