A retrospective review of chest radiographs from 205 patients with blunt chest trauma who also underwent aortography was performed. Forty-one of the 205 had aortographically proved aortic rupture. Discriminant analysis of 16 radiographic signs indicated that the most discriminating signs were loss of the aorticopulmonary window, abnormality of the aortic arch, rightward tracheal shift, and widening of the left paraspinal line without associated fracture. No single or combination of radiographic signs demonstrated sufficient sensitivity to indicate all cases of traumatic aortic rupture on plain chest radiographs without the performance of a large number of aortographically negative studies. The bedside anteroposterior "erect" view of the chest proved far more valuable than the supine view in detecting true-negative studies. Despite significant reader variability in the interpretation of the various radiographic signs, in general the analysis confirmed the role of chest radiography in this clinical situation, but suggests that its most beneficial use is in excluding the diagnosis and eliminating unwarranted aortography rather than in predicting aortic rupture.

Between 1996 and 1999, 54 patients with wrist pain had magnetic resonance imaging performed using a 1.5 Tesla scanner without a wrist coil. Wrist arthroscopy was performed using a standard technique. The findings were then compared. Magnetic resonance imaging had a low sensitivity for the detection of triangular fibrocartilage complex injuries (0.44) and scapholunate ligament injuries (0.11) when wrist arthroscopy was used as the standard of reference. We conclude that when a magnetoresonance technique that does not employ a dedicated wrist coil is used, a negative magnetic resonance imaging scan does not exclude these two significant injuries.

No abstract

No abstract