Methods of defining the noninferiority margin: A systematic review

Abstract

Background: Noninferiority trials are conducted to investigate whether the efficacy of the test drug is not worse than the active comparator based on a predefined noninferiority margin. There is no consensus on a preferred method for defining the noninferiority margin, and previous studies showed that the rationale for its choice is often not reported. Objectives: To investigate how the noninferiority margin is defined in the published literature, and whether its reporting has changed over time. Methods: A systematic PubMed search was conducted for all published randomized, double-blind, noninferiority trials from January 1, 1966, to February 6, 2015. The time trend of defining the margin since the first published noninferiority trial was analyzed using Poisson regression analysis. The impact of the 2010 US Food and Drug Administration (FDA) draft guidance for noninferiority clinical trials on the choice of the margin, and the impact of the extension of the Consolidated Standards of Reporting Trials (CONSORT) Statement on the reporting of noninferiority margins were analyzed using generalized estimating equations to account for the clustering of the margins within articles. Results: We included 275 articles, which account for 283 trials and 328 noninferiority margins. The rationale for the choice of the margin was not reported for 191 margins (58.2%). The under-reporting of the rationale for the choice of the margin has not improved neither since the first published noninferiority trial in this review in 2000 (P=0.86), nor since the publication of the extension of the CONSORT Statement in 2006 (P=0.96). The other 137 margins were mainly defined based on the historical evidence of the active comparator (n=56) or subjectively based on expert opinions (n=46). There was a 3.5-fold increase in the use of the fixed-margin method, the recommended method by the FDA to define the margin, after the publication of the FDA draft guidance (from 2.6% to 9%, P=0.04). Conclusions: Margins in noninferiority trials are poorly reported and this has not improved over recent years. Authors, reviewers, and editors need to take notice of reporting this critical information to allow for better judgment of noninferiority trials.