RETURN TO ABSTRACT LISTING


Objective:
The widened mediastinum is a well-known and widely accepted indicator of vascular injury on chest radiographs (CXR). As early as the 1960s, peer-reviewed papers in the radiology literature have reported size and width criteria for the mediastinum in cases of thoracic trauma, however there is scant evidence validating its use. Both technical and physiologic factors are known to affect the mediastinal width on chest radiographs. For example, it is known that low intravascular volume states can present as a narrowed mediastinum and high intravascular volume states will result in a widened mediastinum. Might this be an example of ‘eminence-based medicine,’ where a medical belief is based on commonly accepted beliefs espoused by authority figures, rather than evidence-based medicine?

Materials and Methods:
A literature review of the radiographic widened mediastinum in aortic injury, blunt chest trauma, and mediastinal injury was conducted to understand the development of the concept and how it became widely believed, tracing the earliest reports of widened mediastinum in case reports to the earliest research studies and through to the most recent scholarship on the subject.

Results:
The earliest reports we could find of widened mediastinum as a radiographic sign of traumatic vascular injury come from Wyman’s case series in 1953, where it is proposed that a widened mediastinum on radiography should exist in aortic injuries, "were it possible to obtain satisfactory films." Other case studies followed with Blazek declaring in 1964, “The correct diagnosis is almost entirely dependent upon roentgenographic examination of the chest. Widening of the mediastinum with or without rib fracture is the cardinal sign [of acute traumatic rupture of the thoracic aorta].” Cohort studies in the 1970s generated more data in support of the widened mediastinum and began to promote measurement thresholds. Subsequent literature published through the 1990s and 2000s, however, began to cast doubt on the validity of the appearance of a widened mediastinum on CXR for the diagnosis of aortic injury. In the past decade, contemporary studies have shown that the appearance of a widened mediastinum is neither sensitive nor specific (< 50%) for aortic injury nor mediastinal hemorrhage.

Conclusion:
Mediastinal widening as a radiographic sign of vascular injury seems to confirm a common-sense interpretation of anatomic understanding, that potential spaces swell when injured or when a hematoma forms. Because early reports promoted this ‘common-sense’ heuristic for an injury that necessitates emergent surgical intervention, this assertion was widely accepted without experimental evidence or validation. There is now, however, evidence suggesting that a widened mediastinum on CXR is neither sensitive nor specific for traumatic vascular injury. This may be an example where the availability of contradictory evidence seems to struggle against the inertia of commonly held traditional beliefs. Future directions involve a widespread educational update.