Choose-NI-metric: a shiny application for converting and comparing non-inferiority metrics in binary endpoint trials.

backgroundIn non-inferiority (NI) clinical trials with binary outcomes, treatment effects are commonly quantified using the risk difference (RD), relative risk (RR), or odds ratio (OR), each of which can be used to define a non-inferiority margin. Although these measures are mathematically related, results reported under different metrics are often difficult to interpret and compare in practice, particularly for clinicians without extensive statistical training. In addition, recent methodological work has shown that the choice of effect measures can meaningfully influence sample size requirements at the design stage. To address these challenges, we developed Choose-NI-metric, an open-access Shiny application that facilitates translation and comparison across NI metrics within a unified framework.

resultsChoose-NI-metric combines statistical computation with interactive visualizations to support both interpretation and trial design. The application allows users to convert among RD, RR, and OR, and to examine how sample size requirements vary across metrics. Sample size calculations are implemented using both the Wald approach and the method of variance estimation recovery (MOVER). An applied example based on the FRIDA trial illustrates that different choices of effect measure can result in substantial differences in required sample size, highlighting the practical implications of metric selection.

conclusionsChoose-NI-metric offers an accessible platform that connects clinical interpretation with statistical considerations in NI trial design. By enabling direct translation and comparison among RD, RR, and OR, the application supports more transparent interpretation of results across studies and informed selection of effect measures, ultimately contributing to more efficient and cost-effective non-inferiority studies.