non-inferiority

Most drug trials ask whether a treatment beats a placebo. That question cannot be asked when an effective treatment already exists and withholding it would harm the people taking part, so the new drug is compared against the existing one instead, which regulators call an active control. The goal of such a trial, in the US Food and Drug Administration's own words, is not to show that the new drug is equivalent, but that it is not materially worse than the control.

How much worse is too much is decided before the trial starts, as a figure called the non-inferiority margin. If the result rules out a gap larger than that margin, the new treatment passes. The reading matters: passing means the new option is close enough to be worth having, not that it is better, and a treatment can be judged non-inferior while scoring slightly lower on the raw numbers. The FDA also cautions that the conclusion rests on something the trial itself does not measure, namely that the older drug worked as well in this trial as it has in the past.