ARRS 2022 Abstracts


1800. The Inter-Reader Agreement of Radiology Residents in Using the ACR TI-RADS System Before and After Training
Authors * Denotes Presenting Author
  1. Meredith Bara *; University of Alberta
  2. Yang Du; University of Alberta
  3. Mitchell Wilson; University of Alberta
  4. Jonathan Porter; University of Alberta
  5. Katrin Resch; University of Alberta
  6. Roger Croutze; University of Alberta
  7. Gavin Low; University of Alberta
This study aims to evaluate the inter-reader agreement of radiology residents in using American College of Radiology Thyroid Imaging Reporting and Data System (ACR TI-RADS) before and after training

Materials and Methods:
A single center retrospective cohort study evaluating 50 thyroid nodules in 50 patients of varying TI-RADS levels was performed. Reference standard TI-RADS scores were established through a consensus panel of three fellowship-trained staff radiologists. Three blinded independent PGY-4 residents were evaluated to obtain the inter-reader agreement before and after TI-RADS training performed 6 weeks apart. Fleiss kappa was used to measure the pooled inter-reader agreement.

There were 42 women and 8 men with a mean age of 55.9 ± 12.7 years. The mean nodule size was 19 ± 13.7 mm (range 5 - 63 mm). A statistically significant superior inter-reader agreement was found on the post-training assessment compared to the pre-training assessment for the following variables: shape (k of 0.09 [slight] pre-training vs. 0.67 [substantial] post-training, P<0.001); echogenic foci (k of 0.28 [fair] pre-training vs. 0.45 [moderate] post-training, P=0.004); TI-RADS leve’ (k of 0.14 [slight] pre-training vs. 0.36 [fair] post-training, P <0.001); and recommendations (k of 0.36 [fair] pre-training vs. 0.50 [moderate] post-training, P =0.02). No significant differences between the pre- and post-training assessments were found for the variables spongiform, composition, echogenicity, and margins.

The study supports the use of training in improving the inter-reader agreement of radiology residents when using ACR TI-RADS.