A novel approach to improving the interpretation of CT brain in trauma

Computed tomography of the brain (CTB) has a fundamental role in the diagnosis and management of traumatic brain injury (TBI). There may be substantial discordance between initial CTB interpretation by emergency clinicians and the final radiology report. This study aimed to assess the utility of a structured reporting template in improving the accuracy of CTB interpretation by emergency clinicians.

A prospective pre- and post-intervention cohort study was undertaken using a study population of emergency medicine trainees. The CTB reporting template was created with consultation from radiology, emergency medicine and trauma specialists. Participants reported on a set of randomly selected trauma CTBs first without, and then with, the reporting template. Each case was independently assessed for concordance with the radiology report by two blinded assessors (including a radiologist) and the proportion of concordant reports in each phase calculated.

There were 26 participants recruited to the study who reported on a total of 320 CTBs. In the pre-intervention phase, 121 (76%) cases were concordant with the radiology report compared to 147 (92%) post-intervention (p<0.01). The AUROC was 0.84 (95% CI: 0.78-0.89) pre-intervention and improved to 0.94 (95% CI: 0.88-0.99) with the intervention (p=0.01). A higher level of baseline accuracy was observed in advanced trainees (78%) compared to basic trainees (72%), but both improved to a similar level of 92% with the use of the CTB reporting template. There was a marked reduction in false negative errors, with increased identification of critical diagnoses such as cerebral herniation and diffuse axonal injury.

The use of the CTB reporting template significantly increased the accuracy of emergency medicine trainees and reduced the number of missed critical diagnoses. Reporting templates may represent an effective strategy to improve CTB interpretation and enhance the initial care of head injured patients.