Reporting of CT cervical spine after office hours by radiology trainees—analysis of discrepancy rates and RADPEER scores

Purpose To review the errors made by radiology trainees in the reporting of cervical spine CTs (CCT) and to compare the discrepancy rates between the stages of training. Methods All CCTs reported by trainees after office hours between January 2015 and December 2015 were retrospectively reviewed by a team of five musculoskeletal consultants with experience ranging between 7 and 15 years. Discrepancies between the provisional report by the trainee and the findings by the musculoskeletal consultants were graded according to the RADPEER scoring system. Sensitivity and specificity of the trainees were assessed. Results Of 254 CCT provisional reports, there were 12 (4.7%) discrepancies, of which 5 (2.0%) discrepancies were likely to be clinically significant. We found a clinically significant difference between the stage of training of the trainee and RADPEER score ( P = 0.023). The sensitivity and specificity of the senior radiology trainees were 97.0 and 98.1%, respectively, and that of the junior radiology trainees were 80 and 98.0% respectively ( P = 0.039). Conditions misinterpreted as fractures include degenerative changes ( n = 2) and nutrient vessel ( n = 1). Other missed abnormalities include ossification of the posterior longitudinal ligament ( n = 1), fracture of the foramen transversarium ( n = 2), vertebral body fractures ( n = 2), articular facet fractures ( n = 2), and transverse process fractures ( n = 2). Conclusion Cervical spine CTs performed after office hours can be safely interpreted by senior radiology trainees to a reasonable degree, although a targeted intervention to improve diagnostic performance of junior radiology trainees may be of clinical benefit.