From: Machine learning to help researchers evaluate biases in clinical trials: a prospective, randomized user study
Overall
(n = 328)
RSG
(n = 82)
AC
BPP
BOA
Judgements changed
27 (9%)
7 (9%)
6 (7%)