Node-RADS: a systematic review and meta-analysis of diagnostic performance, category-wise malignancy rates, and inter-observer reliability

Question Node-RADS is designed for structured reporting of the possibility of cancer involvement of lymph nodes, but the evidence supporting its application has not been summarized. Findings Node-RADS presented diagnostic performance with AUC of 0.92, and malignancy rates for categories 1-5 ranged from 4% to 100%, while the inter-observer reliability was unclear. Clinical relevance Node-RADS is a useful tool for structured reporting of the possibility of cancer involvement of lymph nodes with high diagnostic performance and appropriate malignancy rate for each category, but unclear inter-observer reliability may hinder its implementation in clinical practice.