Background: Discussion still remains which polytraumatized patients require radiologic thoracolumbar spine (TL spine) screening. The purpose of this study is to determine whether pelvic fractures are associated with TL spine fractures after a blunt trauma. Additionally, the sensitivity of conventional TL spine radiographs and pelvic radiographs (PXRs) is evaluated.
Methods: We prospectively studied 721 consecutive patients who had sustained a high-energy blunt trauma. The diagnostic workup in these patients included routine conventional radiographs of the pelvis and TL spine followed by a computed tomography (CT) analysis. All patients with pelvic fractures and TL spine fractures identified on conventional radiographs and CT were analyzed. A relative risk (RR) was calculated for the association between pelvic fractures and TL spine fractures. The sensitivity for conventional TL spine radiographs and PXRs in identifying fractures was calculated.
Results: Of the 721 patients studied, 620 were included in our diagnostic high-energy trauma protocol. Of these 620 included patients, 86 (14%) suffered a pelvic fracture and 126 (20%) suffered a TL spine fracture. Thirty-three patients (5%) suffered both a pelvic fracture and a TL spine fracture. The RR for a TL spine fracture in the presence of a pelvic fracture identified on PXR is 2.14 (95% confidence interval, 1.54-2.98) and identified on CT this RR is 2.20 (95% confidence interval, 1.59-3.05). However, this association diminishes to a nonsignificant level when the transverse process and spinous process fractures are excluded. Overall sensitivity for conventional TL spine radiographs and PXRs is 22% and 69%, respectively.
Conclusion: Our data suggest that a pelvic fracture is not a predictor for clinically relevant TL spine fractures. Furthermore, our data confirm the superior sensitivity of CT for detecting TL spine injury and pelvic fractures.