Computed tomography in left-sided and right-sided blunt diaphragmatic rupture: experience with 43 patients

AimTo investigate differences in the radiographic signs for left and right-sided blunt diaphragmatic rupture (BDR) in order to provide guidance to avoid missing these injuries.Materials and methodsA retrospective review of the computed tomography (CT) examinations of 43 patients with BDR treated at our hospital between January 1995 and 2007 was undertaken. The presence of diaphragmatic discontinuity, diaphragmatic thickening, herniation of abdominal organs into the thoracic cavity, collar/hump sign, dependent viscera sign, abnormally elevated 4 cm or more above the dome of the other-sided hemi-diaphragm, and of associated injuries was recorded and their relationship to each other and to BDR diagnosis examined. A comparison between the use of axial and sagittal/coronal reconstruction images in diagnosis was also performed in 15 patients.ResultsOn axial imaging, left-sided diaphragmatic rupture occurred in 31 patients (72%) and right-sided in 12 (28%). Twenty-nine patients had associated injuries. More than 60% of the patients showed the “dependent viscera” sign, “abdominal organ herniation” sign, diaphragm thickening, or had a more than 4 cm elevation of one side of the diaphragm. “Diaphragmatic discontinuity” and “stomach herniation” were seen almost exclusively in left-sided rupture. Those with BDR and haemothorax had a significantly lower incidence of “diaphragm discontinuity” (p = 0.034) than those without haemothorax. Sagittal/coronal reconstruction slightly increased the number of band signs, diaphragmatic discontinuities and diaphragmatic thickenings seen.ConclusionsOf the CT signs examined in this study, when herniation of abdominal organs was used as a diagnostic marker, only a very small fraction of trauma patients identifiable by CT would be missed. Further, CT signs differ for left-sided and right-sided BDR, thus the possibility of BDR should be considered when any of the reported CT signs are present.