how we can know before conducting the experiment that it will be non-inferiority or superiority trial? How we can know that the drug /treatment is superior or just equal ?
generally at the outset you will have an idea of an expected outcome of test agent vs the comparator. In other situations standard of care outcomes may be already very good when looking at traditional endpoints. as such the test agent may not be superior to the comparator and as such a non-inferiority study is performed.
I would suggest that while writing you protocol you must decide beforehand about the objective of the clinical trial which will help you to decide whether to have a non-inferiority or superiority trial.
Control data and the data about the test will also help in estimating the superiority or non-inferiority study design.
Please refer to US FDA guidance on these type of trials by visiting the suggested link http://www.fda.gov/downloads/Drugs/Guidances/UCM202140.pdf
Hope this is of help.
Dear Heba
I agree with Elmer that you have asked two questions. The first is how to decide whether your study should be designed around non-inferiority/superiority and the second is whether your results will show as equivalence.
As highlighted in the FDA document suggested by Shahid Karim, there are pitfalls in designing the non-inferiority/superiority study model, especially if there is a possibility of needing to switch from one to the other. However, in designing your study you should always be aware that the switch may be necessary, so the first thing to consider is adequate powering in order to achieve a non-biased outcome. For both approaches it is safest to consider both ITT and PP analyses and to then compare the outcomes of those analyses for bias and to ensure that the bounds of non-inferiority/superiority are clearly set before you start your study.
Since you do not know whether your test material is likely to be non-inferior or superior, unless you have performed some preliminary testing, you need to actually design a study that will identify superiority of one of the products under test (I have encountered questions from the IRB/ethics committee as to how I could be confident that Product A might be superior to Product B without incorporating bias) so you must be open to either product being superior at the commencement. Since in general terms the analysis for both non-inferiority and superiority are one-sided tests they are effectively the same. The clearer the outcome of your intervention in yes/no terms the easier this becomes.
If you do not have superiority of either product you presumably, provided you have sufficient data and efficacy with the pre-determined bounds, you may be able to identify equivalence. However, the numbers of subjects required are normally considerably greater by a factor of 2-3 depending upon the clarity of your outcome. In this case I would suggest the following document:
http://www.fda.gov/downloads/Drugs/Guidances/ucm070244.pdf
I hope this helps
Regards, Ian
- Badges
- Science topic
- Similar topics
- Chemistry
- Pharmacology
- Medical Pharmacology
- Clinical Pharmacology