Work

Design Bias and Clinically Important Treatment Effects in Randomized Trials of Critical Illness