<i>Editorial Comme</i> <i>ntary:</i> The Statistical Fragility Index of Medical Trials Is Low By Design: Critical Evaluation of Confidence Intervals Is Required

The Fragility Index (FI) provides the number of patients whose outcome would need to have changed for the results of a clinical trial to no longer be statistically significant. Although it's a well-intended and easily interpreted metric, its calculation is based on reversing a significant finding and therefore its interpretation is only relevant in the domain of statistical significance. Its interpretation is only relevant in the domain of statistical significance. A well-designed clinical trial includes an a priori sample size calculation that aims to find the bare minimum of patients needed to obtain statistical significance. Such trials are fragile by design! Examining the robustness of clinical trials requires an estimation of uncertainty, rather than a misconstrued, dichotomous focus on statistical significance. Confidence intervals (CIs) provide a range of values that are compatible with a study's data and help determine the precision of results and the compatibility of the data with different hypotheses. The width of the CI speaks to the precision of the results, and the extent to which the values contained within have potential to be clinically important. Finally, one should not assume that a large FI indicates robust findings. Poorly executed trials are prone to bias, leading to large effects, and therefore, small P values, and a large FI. Let's move our future focus from the FI toward the CI.