Prospective study of 110 locally advanced or metastatic genitourinary cancer cases: Claude 3.5 Sonnet (mSCS 0.731) narrowly missed and ChatGPT-4 (0.660) clearly failed non-inferiority to tumour boards (0.849; margin 0.699).
Studyn = 110 case scenarios
Rinderknecht et al. Benchmarking Large Language Models Against Multidisciplinary Tumor Boards in Urological Oncology: Results from the Blinded, Prospective CONCORDIA Study