How to understand what is guilty? [Study Assessment]
Hi BE-Proff,
Yes that can happen.
In practice, you often can't directly tell what caused the problem.
A few rules of thumb:
1. If the CI is completely outside the acceptance range (e.g. 54.66% - 77.19%) then you can say you have shown bioinequivalence. In this case it is statistically a formulation problem. At least.
2. If the CI is not within the acceptance range but the point estimate is "close" to 1.0, then your trial was most likely underpowered, and you would in this case be justified in repeating it.
3. If your CI is not within the acceptance range and point estimate is far from one, then there is really no good way of telling what has happened or how to remedy it. Perhaps you can repeat it, perhaps you shouldn't, it will always be a difficult discussion.
It is pt. 3 that often happens, unfortunately.
Look also at predose concentrations, subjects lacking AUCinf values, protocol deviations, failed analytical runs. These may indicate a practical issue. I believe there are people much more experienced than I who can write entire books about what to do when a study fails and when to repeat it.
I hope the few points above are helpful.
Let's hear your numbers and some background of your failure, please.
❝ Saying, I have conducted a BE-study which doesn't show bioequivalence.
Yes that can happen.
❝ Is it possible to understand what was the reason of this failure - poor formulation or incorrect time points in the protocol?
In practice, you often can't directly tell what caused the problem.
A few rules of thumb:
1. If the CI is completely outside the acceptance range (e.g. 54.66% - 77.19%) then you can say you have shown bioinequivalence. In this case it is statistically a formulation problem. At least.
2. If the CI is not within the acceptance range but the point estimate is "close" to 1.0, then your trial was most likely underpowered, and you would in this case be justified in repeating it.
3. If your CI is not within the acceptance range and point estimate is far from one, then there is really no good way of telling what has happened or how to remedy it. Perhaps you can repeat it, perhaps you shouldn't, it will always be a difficult discussion.
It is pt. 3 that often happens, unfortunately.
Look also at predose concentrations, subjects lacking AUCinf values, protocol deviations, failed analytical runs. These may indicate a practical issue. I believe there are people much more experienced than I who can write entire books about what to do when a study fails and when to repeat it.
I hope the few points above are helpful.
Let's hear your numbers and some background of your failure, please.
—
Pass or fail!
ElMaestro
Pass or fail!
ElMaestro
Complete thread:
- How to understand what is guilty? BE-proff 2016-07-30 18:21 [Study Assessment]
- How to understand what is guilty?ElMaestro 2016-07-30 23:15
- How to understand what is guilty? BE-proff 2016-08-02 17:06
- How to understand what is guilty? DavidManteigas 2016-08-08 11:18
- Post-hoc power is useless! d_labes 2016-08-09 08:31
- Post-hoc power is useless! ElMaestro 2016-08-09 13:00
- Post-hoc power is useless! d_labes 2016-08-09 14:30
- Post-hoc power is useless! ElMaestro 2016-08-09 16:17
- Post-hoc power is useless! d_labes 2016-08-09 14:30
- Post-hoc power is useless! DavidManteigas 2016-08-09 13:51
- Post-hoc power is useless! d_labes 2016-08-09 14:11
- Post-hoc power is useless! ElMaestro 2016-08-09 13:00
- Post-hoc power is useless! d_labes 2016-08-09 08:31
- How to understand what is guilty?ElMaestro 2016-07-30 23:15