Evaluating accuracy of cervical spine computed tomography interpretation by emergency trainees with the use of a structured protocol.

Abstract

Objective: Radiological evaluation of cervical spine injury with computed tomography (CT) scanning is a fundamental component of the assessment of major trauma. Accurate interpretation of scans is essential for safe clearance or diagnosis of injuries. However, delays in radiologist reporting often result in prolonged spinal immobilisation. The aim of the present study was to evaluate a simple, structured reporting tool to improve assessment of CTs of the cervical spine by emergency medicine trainees.

Methods: A prospective pre- and post-intervention cohort study was undertaken within a major metropolitan ED. Participants in the pre-intervention phase interpreted a set of randomly selected cervical spine CTs. The post-intervention phase presented the same task with the additional provision of a structured cervical spine CT reporting template designed in collaboration with radiologists and emergency physicians. Interpretation by trainees was evaluated for concordance with the final radiology report by two blinded assessors.

Results: A total of 155 cervical spine CT scans were reported by the 46 participants. Participants in the cohorts were similar with regards to experience and country of primary medical degree. Concordance with the radiology report in the pre-intervention phase was 60% (95% CI 0.48-0.71), compared with a concordance of 54% (95% CI 0.42-0.65) in the post-intervention phase (P = 0.46).

Conclusions: Interpretation of cervical spine CT scans by trainees was inferior compared to radiologists and did not improve with a structured reporting template. Other innovative strategies towards timely reporting of CT scans by radiologists of the cervical spine are indicated for earlier definitive diagnosis.