Imaging of the thoracic and lumbar spine in a high volume level 1 trauma center: are reformatted images of the spine essential for screening in blunt trauma?

Reformatted CTs of the thoracic and lumbar spine (CT T/L) from CTs of the chest, abdomen, and pelvis (CT body) may be performed for screening the thoracolumbar spine in patients sustaining blunt trauma. The purpose of this study was to determine whether there was a difference in the rate of detection of spinal fractures on CTs of the body compared to the reformatted T/L spine. A secondary endpoint was to evaluate whether cases dictated by trainees improved fracture detection rate. We reviewed the records of 250 consecutive blunt trauma patients that received CTs of the chest, abdomen, and pelvis (CT body) with concurrent CT T/L reformats. Each report was reviewed to determine if there was a thoracolumbar fracture and whether a trainee had been involved in interpreting the CT body. If a fracture was identified on either report, then the number, type, and location of each fracture was documented. Sixty-nine fractures, from a total of 38 patients, were identified on either the CT of the body or the CT T/L. Sensitivity for CT body interpretations was 94 % (95 % CI: 86-98 %) compared to a 97 % (95 % CI: 89-100 %) sensitivity for the CT T/L (p > 0.5). Although the sensitivity was 97 % (95 % CI: 88-100 %) when a trainee was involved in interpreting the body CT, there was no statistically significant improvement. The results suggest that with careful scrutiny most spine fractures can be diagnosed on body CT images without the addition of spine reformats. The most commonly missed finding is an isolated non-displaced transverse process fracture, which does not require surgical intervention and does not alter clinical management. The results suggest that thin section reformats do not need to be routinely ordered in screening blunt trauma patients, unless a bony abnormality is identified on the thicker section body CT images.