Determination of normal transverse mediastinal width and mediastinal-width to chest-width (M/C) ratio in control subjects: implications for subjects with aortic or brachiocephalic arterial injury.

Abstract

We measured transverse mediastinal width and mediastinal-width to chest-width (M/C) ratio on supine films of 100 nontraumatized controls. In 95% the transverse mediastinal width was less than 7.5 cm and the M/C ratio was less than 0.38. Thus a transverse width of 7.5 cm or more or an M/C ratio of 0.38 or more can be defined as abnormal with 95% confidence. Application of these values to determine abnormality in 32 patients with proven aortic or brachiocephalic injury showed that the transverse mediastinal width was within normal limits in 41% and M/C ratio was normal in 69%. Utilizing smaller values that would identify all abnormals resulted in false positive rates in the controls of 74% and 87%, respectively. However, one or more of eight specific signs of mediastinal abnormality related to hemorrhage or pseudoaneurysm formation were present in 94% of abnormals compared to only 11% of controls. Because of extreme overlap of transverse mediastinal width and M/C ratio between normals and abnormals, precise measurement of the mediastinum cannot reliably separate the two groups. The subjective assessment of anatomic mediastinal abnormality remains a superior plain film method in determining the need for aortography.