Helical CT of blunt diaphragmatic rupture

Abstract

OBJECTIVE. This study evaluated CT findings for signs of blunt diaphragmatic rupture. MATERIALS AND METHODS. CT examinations of 179 blunt trauma patients, including 11 with left-sided and five with right-sided blunt diaphragmatic rupture, were reviewed by two staff radiologists who first decided by consensus on the presence or absence of 11 published signs of blunt diaphragmatic rupture and then formulated the diagnosis in terms of absence of, presence of, or suggestion of blunt diaphragmatic rupture. The significance of the findings was assessed by multivariate logistic regression. Four other reviewers interpreted the CT findings independently. They were asked first to formulate a diagnosis in terms of absence of, presence of, or suggestion of blunt diaphragmatic rupture and then to enumerate the findings supporting a diagnosis or suggestion of blunt diaphragmatic rupture. These findings were compared with those of the staff radiologists. RESULTS. Diaphragmatic discontinuity, diaphragmatic thickening, segmental nonrecognition of the diaphragm, intrathoracic herniation of abdominal viscera, elevation of the diaphragm, and both hemothorax and hemoperitoneum were strong predictors of blunt diaphragmatic rupture (p < 0.001). The combination of the first three findings was 100% sensitive (16/16). The staff radiologists' sensitivity for diagnosing blunt diaphragmatic rupture was 100% (16/16). The four reviewers' sensitivities were 56.2% (9/16), 81.2% (13/16), 62.5% (10/16), and 87.5% (14/16). CONCLUSION. Six of 11 signs were good predictors of blunt diaphragmatic rupture. Despite diaphragmatic thickening, focal defect and segmental nonrecognition had 100% cumulative sensitivity; the reviewers formulating the diagnosis before analyzing CT signs overlooked blunt diaphragmatic rupture on CT in 12.5-43.8% of the patients.