I want to demonstrate non-inferiority for Group A compared to Group B. The difference (A-B) between the two groups is -4.1, and the 95% confidence interval for the group difference is from -12.1 to 3.8. Since my non-inferiority margin was -15, I concluded that A is non-inferior to B. My question here is, if the lower margin of the 95% confidence interval is not less than -15 but the upper margin is less than 0, how should I draw a conclusion? For instance, if the 95% confidence interval is from -12.1 to -1.5, can I still conclude non-inferiority?
I got good reference for this topic (Article Interpreting the results of noninferiority trials—a review
).To mention the conclusion, we can say that even in the case of the question, it can be considered noninferior. The results were inferior when using the conventional method, but with noninferiority methods, it is considered noninferior becuase the results (lower limit of the 95% confidence interval for the group difference) are superior to the pre-defined non-inferior margin. In other words, noninferior treatments can be conventionally inferior, non-significantly different, or superior to the standard treatment.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS