What are the sensitivity and specificity of a widened mediastinum on a standard chest radiograph for detecting a thoracic aortic aneurysm?

Sensitivity and Specificity of Widened Mediastinum on CXR for Thoracic Aortic Aneurysm

Chest radiography has poor diagnostic accuracy for detecting thoracic aortic aneurysms, with a sensitivity of only 64% for widened mediastinum and 71% for abnormal aortic contour, and a specificity of 86%, making it inadequately sensitive to exclude significant aortic pathology. 1

Diagnostic Performance

The American College of Cardiology guidelines, based on pooled data from 10 studies, establish that:

• Sensitivity of widened mediastinum alone: 64% 1
• Sensitivity of abnormal aortic contour: 71% 1
• Sensitivity when including all abnormal radiographic findings: 90% 1
• Specificity for aortic pathology: 86% 1

This means that approximately 36% of thoracic aortic aneurysms will be missed if relying on widened mediastinum alone, and up to 29% will be missed even when looking for abnormal aortic contour. 1

Critical Limitations in Clinical Practice

Poor Negative Predictive Value

A completely normal chest X-ray does lower the likelihood of aortic disease but only in very low-risk patients—it cannot definitively exclude thoracic aortic pathology in intermediate or high-risk patients. 1, 2

Projection-Dependent Accuracy

Research demonstrates that posteroanterior (PA) films are significantly more accurate than anteroposterior (AP) films, with optimal cutoff values of:

• PA projection: Left mediastinal width ≥4.95 cm (sensitivity 90%, specificity 90%) 3
• AP projection: Mediastinal width ≥8.65 cm (sensitivity 72%, specificity 80%) 3

The lower diagnostic accuracy of AP films (commonly obtained in emergency settings) means even higher false-negative rates in acute presentations. 3

Location-Specific Sensitivity

Sensitivity varies dramatically by anatomic location:

• Ascending aorta pathology: 47% sensitivity 4
• Descending aorta pathology: 77% sensitivity 4

This is particularly problematic since ascending aortic aneurysms carry the highest mortality risk and require urgent surgical intervention. 4

High False-Positive Rate in Trauma

For traumatic aortic injury specifically, chest X-ray performs even worse:

• Sensitivity for blunt thoracic aortic injury: 41% 5
• Signs of hemomediastinum are more often false positive than true positive for aortic injury 1
• Inter-reader variability is substantial (kappa = 0.49 for mediastinal widening) 6

The American College of Cardiology explicitly states that chest X-ray is a poor screening test for aortic injury in trauma patients. 1

Clinical Algorithm

When CXR Shows Widened Mediastinum or Abnormal Aortic Contour:

Proceed immediately to definitive imaging with CT angiography (sensitivity approaching 100%, specificity 98-99%), MRI, or transesophageal echocardiography. 1, 7

When CXR is Normal:

• Low-risk patients: Normal CXR provides meaningful reassurance 1
• Intermediate or high-risk patients: Proceed to definitive imaging regardless of normal CXR, as up to 40% of acute aortic dissections lack widened mediastinum and 16% appear completely normal on chest film 2

Risk Stratification Factors Requiring Definitive Imaging:

• Sudden severe chest or back pain 2
• Pulse differential between extremities 2
• New aortic regurgitation murmur 2
• Hypotension or neurologic deficits 2
• Known connective tissue disorder or family history of aortic disease 1

Key Clinical Pitfall

The most dangerous error is using a normal chest X-ray to exclude thoracic aortic aneurysm or dissection in anything other than very low-risk patients. The American College of Radiology emphasizes that chest X-ray findings are neither sensitive nor specific enough to serve as a definitive screening tool, and definitive cross-sectional imaging should be obtained when clinical suspicion exists. 1, 7