How high is the inter-observer reproducibility in the LIRADS reporting system?

Area of Science:

• Radiology
• Hepatology
• Oncology

Background:

• Liver nodules require accurate characterization to guide patient management.
• The Liver Imaging Reporting and Data System (LIRADS) was developed to standardize the reporting of liver nodules on imaging.
• Assessing the reproducibility of LIRADS is crucial for its effective clinical implementation.

Purpose of the Study:

• To evaluate the inter-observer reproducibility of the LIRADS v2014 reporting system.
• To identify areas of agreement and disagreement in the interpretation of liver nodules using LIRADS.
• To contribute to the wider adoption of LIRADS in clinical practice.

Main Methods:

• A retrospective analysis of 132 patients with liver nodules who underwent dynamic MRI/CT.
• Independent interpretation of liver nodules by five radiologists using LIRADS v2014 criteria.
• Calculation of inter-observer agreement using kappa statistics for different LIRADS categories and parameters.

Main Results:

• Higher inter-observer agreement was observed for LR-1, LR-5, and LR-5V categories (κ = 0.522–0.600) compared to LR-2, LR-3, and LR-4 categories (κ = 0.082–0.298).
• The parameter with the highest inter-observer agreement was venous thrombus (κ = 0.600).
• Overall, LIRADS demonstrated acceptable inter-observer reproducibility but was insufficient at intermediate risk levels.

Conclusions:

• LIRADS v2014 provides acceptable inter-observer reproducibility for clinical practice, particularly for definitive malignant and benign categories.
• Reproducibility is insufficient for intermediate-risk categories, highlighting a need for refinement.
• Enhanced training, numerical probability assignments, and clearer rules for ancillary features could further improve LIRADS utility.