Sepsis is a clinical entity with complex pathophysiological processes, and it is the leading cause of death in critically ill patients. Unfortunately, a majority of clinical trials aiming to reduce sepsis mortality have come to negative results. It is of great importance to consider the limitation in the designing and conducting of these trials: whether the premise is scientifically sound, the study population is properly selected to test the efficacy of intervention, the confounding factors besides intervention are fully considered, and whether the statistical conclusion is robust and meaningful.