Diagnostic Testing for Pulmonary Embolism

Ventilation/Perfusion Scanning

Nuclear ventilation/perfusion scans are often used to investigate possible PE. Palmowski et al. (2014) reported the sensitivity and specificity of ventilation/perfusion scanning as 95.8% and 82.6%, respectively, with false negative rates of 4.2% and false positive rate of 17.3%. Hence, a normal scan virtually excludes a PE (high negative predictive value). Identified perfusion defects are non-specific and only represent true PE in about one-third of cases. The probability that a perfusion defect represents a PE increases with the size, shape and number of defects as well as the presence of a normal ventilation scan. Mismatched defects, normal ventilation scan reveals poor perfusion, but are large or segmental in size are “high probability” defects and are associated with an approximately 80% prevalence of PE (Kearon et al. 1998).

Table: Probability of Pulmonary Embolism Based on Ventilation-Perfusion Scan Results and Clinical Suspicion in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED Investigators) Study

Ventilation-Perfusion Scan Results

Clinical Suspicion of Pulmonary Embolism*

 

Low

Intermediate

High

High probability

56%

88%

96%

Intermediate probability

16%

28%

66%

Low probability

4%

16%

40%

Normal/near-normal probability

2%

6%

0%

* Percentage of patients with pulmonary embolism; Adapted from the PIOPED Investigators (Gill & Nahum 2000; PIOPED Investigators 1990).

 

The Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED Investigators) study demonstrated that a low-probability or normal ventilation-perfusion scan with a low clinical suspicion of PE essentially excludes the diagnosis of PE (negative predictive values of 96% and 98% respectively) (Gill & Nahum 2000; PIOPED Investigators 1990). When clinical suspicion is high and the scan indicates a high probability of PE, the positive predictive value is 96% (Gill & Nahum 2000; PIOPED Investigators 1990), and these patients should be treated. The majority of ventilation perfusion scans have non-diagnostic results, requiring further testing (PIOPED Investigators 1990). 

Pulmonary Angiography

Pulmonary angiography is the gold standard for diagnosis of PE (Gill & Nahum 2000). It involves percutaneous catheterization and injection of contrast dye into a pulmonary artery branch (Gill & Nahum 2000). Pulmonary angiography is most commonly used when ventilation-perfusion scanning is non-diagnostic but clinical suspicion remains high (Tapson et al. 1999). A negative pulmonary angiogram excludes clinically relevant PE (Gill & Nahum 2000; Tapson et al. 1999). It is expensive and is associated with the risk of significant complications. Relative contraindications include significant bleeding risk, allergy to contrast medium, and renal insufficiency (Gill & Nahum 2000).  

Spiral CT Scan

A spiral computed tomography scan is a non-invasive test, which can scan the entire thorax in one breath-hold. The sensitivity of a spinal computed tomography scan in diagnosing PE has been reported to range from 64% to 93%, with a specificity ranging between 89% and 100%. It is most accurate when the embolism is large. It actually visualizes the clot and has the added benefit of investigating other disease states in the differential diagnosis.