Clinical trial design for unmet clinical needs: a spotlight on sepsis

<b>Introduction</b>: Despite considerable advances in our understanding of how sepsis develops and multiple clinical trials of potential therapies, no new pharmacologic agent has been consistently shown to improve survival. <b>Areas covered</b>: We reviewed relevant publications identified through PubMed and from the authors' knowledge of this field. We discuss the main reasons why clinical trials on new therapeutic interventions have failed in the past, including heterogeneity of study populations and choice of outcome measures. We discuss how changes in study design and in patient selection could help improve identification of effective agents in the future. <b>Expert opinion</b>: The search for new sepsis therapies must continue but lessons must be learned from previous clinical trials so that the same mistakes are not repeated. Rather than grouping all patients with sepsis together, we should study only those most likely to benefit from the intervention. Better characterization of patients will be facilitated using modern 'omics technology and analysis of the increasingly large quantities of clinical data available, enabling more personalized patient selection for trial inclusion. New clinical trial design and inclusion of other endpoints in addition to mortality will also aid our search for the elusive positive clinical trial and effective interventions for sepsis.