Search

In non-inferiority trials, acceptable efficacy of an experimental treatment is established by ruling out some defined level of reduced effect relative to an effective active control standard. Serial use of non-inferiority trials may lead to newly approved therapies that provide meaningfully reduced levels of benefit; this phenomenon is called bio-creep. Simulations were designed to facilitate understanding of bio-creep risk when approval of an experimental treatment with efficacy less than some proportion of the effect of the active control treatment would constitute harm, such as when new antibiotics that are meaningfully less effective than the most effective current antibiotic would be used for treatment of Community-Acquired Bacterial Pneumonia. In this setting, risk of approval of insufficiently effective therapies may be great, even when the standard treatment effect satisfies constancy across trials. Modifiable factors contributing to this manifestation of bio-creep included the active control selection method, the non-inferiority margin, and bias in the active control effect estimate. Therefore, when non-inferiority testing is performed, the best available treatment should be used as the standard, and margins should be based on the estimated effect of this control, accounting for the variability and for likely sources of bias in this estimate, and addressing the importance of preservation of some portion of the standard's effect.

Evaluation of non-inferiority is based on ruling out a threshold for what would constitute unacceptable loss of efficacy of an experimental treatment relative to an active comparator "Standard". This threshold, the "non-inferiority margin", is often based on preservation of a percentage of Standard's effect. To obtain an estimate of this effect to be used in the development of the "non-inferiority margin", data are needed from earlier trials comparing Standard to placebo if the non-inferiority trial does not have a placebo arm. This approach often provides a biased over-estimate of Standard's true effect in the setting of the current non-inferiority study. We describe two commonly used non-inferiority margin methods that adjust for this bias, the two-confidence interval (95-95) and the Synthesis margins. However, the added 'variance inflation' adjustment made by 95-95 margin diminishes with increasing information from historical trial(s), and the Synthesis margin is based on a strong assumption that the relative bias is known. We introduce an alternative "Bias-adjusted" margin addressing vulnerabilities of each by attenuating the estimate and by accounting for uncertainty in the true level of bias. Examples and asymptotic estimates of non-inferiority hypothesis rejection rates in the proportional hazards setting are used to compare methods.