Non-Inferiority (Study)

A non-inferiority study aims to demonstrate that the effectiveness of one technology is not clinically inferior to a comparator technology by more than a predefined margin. This margin, known as the non-inferiority margin (M2 or ‘preserved fraction’), represents the percentage of the effect size of the comparator technology against a placebo (M1) that is considered clinically acceptable to be maintained.

Key aspects of non-inferiority studies include:

– Non-Inferiority Margin (M2): The clinically acceptable degree of inferiority compared to the comparator technology. This is defined as a fraction of the effect size of the comparator against a placebo (M1).

– Pre-Study Establishment of M1 and M2: These margins need to be determined before the study begins to calculate the required sample size with sufficient power to support a conclusion of non-inferiority.

– Study Design and Execution: Special attention must be paid to deriving M1 and M2 accurately and ensuring the study design and execution maintain high quality.

– Analysis Approach: Unlike superiority studies, intention-to-treat (ITT) analyses in non-inferiority studies might not be conservative, and per-protocol analyses may be necessary. Additionally, statistical hypothesis testing is typically one-sided at the 2.5% significance level.

Non-inferiority studies have become increasingly common, especially when placebo comparators are unethical or when healthcare payers are interested in establishing the comparative effectiveness of new technologies. These studies are crucial for demonstrating that a new intervention is at least as effective as an existing standard treatment within an acceptable margin, thus providing valuable information for healthcare decision-making.