Outlier in BE [Study Assessment]
Dear All,
Greetings!
We have discussed enough regarding discordant outliers in BE study in the past, but we have struct with typical situation and i would like to know experts opinion on this regard.
we have a done a BE study for GCC submission with two way crossover, here one subject shows less concentration in reference product and the T/R ratio for Cmax 2.43 and AUCt 2.35, because of this subject the study failed in border, if we excluded this as a outlier the 90% CI is within the acceptance limit.
90% CI before excluding particular subject
90% CI After excluding particular subject
It is evident that something happen for this subject during clinical phase but the CRO failed to notify the incident.
GCC just following EMA guidance so as per EMA this is failed study, but out of my curiosity Is there any way to scientifically justify and exclude this subject from statistical calculation.
please let me know your opinion
thanks & Regards
Gopinath
Edit: Category changed; see also this post #1. Tabulators changed to spaces and BBcoded; see also this post #6. [Helmut]
Greetings!
We have discussed enough regarding discordant outliers in BE study in the past, but we have struct with typical situation and i would like to know experts opinion on this regard.
we have a done a BE study for GCC submission with two way crossover, here one subject shows less concentration in reference product and the T/R ratio for Cmax 2.43 and AUCt 2.35, because of this subject the study failed in border, if we excluded this as a outlier the 90% CI is within the acceptance limit.
90% CI before excluding particular subject
Dependent Ratio_ %RefCI_90_Lower CI_90_Upper CV%
Ln(Cmax) 115.46 106.23 125.49 19.47%
Ln(AUClast) 110.25 102.66 118.40 16.64%
90% CI After excluding particular subject
Dependent Ratio_ %RefCI_90_Lower CI_90_Upper CV%
Ln(Cmax) 112.14 104.48 120.36 16.12%
Ln(AUClast) 107.34 101.29 113.75 13.27
It is evident that something happen for this subject during clinical phase but the CRO failed to notify the incident.
GCC just following EMA guidance so as per EMA this is failed study, but out of my curiosity Is there any way to scientifically justify and exclude this subject from statistical calculation.
please let me know your opinion
thanks & Regards

Gopinath
Edit: Category changed; see also this post #1. Tabulators changed to spaces and BBcoded; see also this post #6. [Helmut]
Complete thread:
- Outlier in BEramgopi 2018-03-14 11:29 [Study Assessment]
- Low AUC after reference product Helmut 2018-03-14 16:38
- Low AUC after reference product mittyri 2018-03-15 10:56
- Outlier in BE ramgopi 2018-03-18 07:44
- Outlier in BE ElMaestro 2018-03-18 14:43
- Outlier in BE ramgopi 2018-03-19 10:55
- Outlier in BE ElMaestro 2018-03-18 14:43
- Low AUC after reference product Helmut 2018-03-14 16:38