Core Concepts in Pharmacoepidemiology: New-User Designs.

In this article, we review the history and key reasons for new-user comparisons in pharmacoepidemiology, highlighting the target trial framework as a unifying framework. We describe three distinct pharmacoepidemiological new-user study designs: (1) new-user versus non-user; (2) active comparator new-user; (i.e., ACNU) and (3) prevalent new-user (i.e., PNU) designs, and discuss how each relates to key issues of defining time zero, choosing appropriate comparator groups, and potential sources of bias they do and do not account for. We use a clinical example of hormone replacement therapy and the risk of coronary heart disease to illustrate practical considerations surrounding the application of the three designs presented.