Abstract
OBJECTIVES: The aim of this study was to compare the observer performance of a cohort of radiographers in reporting CT sinus and facial bone investigations against a reference standard and alternative comparator of summary data from peer-reviewed literature. METHODS: The participants (n = 6) completed a 9-month part-time distance learning training programme prior to reporting an examination bank (n = 25 cases) from a retrospectively collected and anonymized digital imaging and communications in medicine archive of CT examinations with referral histories and clinical reports. A literature search was performed to identify an additional alternative comparison reference standard from studies reporting observer performance data in CT sinus and facial bone investigations of both trauma and sinus pathology (target conditions). The data analyses used to measure observer performance and determine differences between the cohort and the reference standards used statistical assessment models including accuracy, sensitivity, specificity, kappa (κ) and summary receiver-operating characteristic curves with estimated area under the curve (AUC). RESULTS: The cohort of radiographer sensitivity was 97.5%, specificity 93.6% and accuracy 95%, with p < 0.000, and a κ = 0.9121 score of agreement. The mean radiographer AUC was 0.9822. The summary reported data of the alternative literature reference standard comparator were AUC 0.9533 for sinus and 0.9374 for trauma. CONCLUSIONS: The results suggest that this cohort of radiographers at the end of a period training in CT sinus and facial bones are able to clinically report comparably high standards.