A Bayesian approach to pilot-pivotal trials for bioequivalence assessment

Table 4 Summary statistics of the predictive distribution of and the prediction about probability of declaring bioequivalence in pivotal trial, given the Pantoprazole tablet pilot trial example

Comparator	Mean	Standard deviation	Percentile
Comparator	Mean	Standard deviation	5.0%	50%	95.0%
C\(_{\text {max}}\)
\(\theta _2\)	0.1721	3.0455	-0.1131	0.1543	0.4507
exp (\(\theta _2\))	1.1878	21.0205	0.8931	1.1668	1.5694
pred.prob.be¹	0.681	0.466	0	1.000	1.000
pred.prob.be²	0.831	0.375	0	1.000	1.000
AUC
\(\theta _2\)	0.0772	3.0454	-0.1332	0.0510	0.2776
exp (\(\theta _2\))	1.0803	21.0184	0.8753	1.0523	1.3200
pred.prob.be¹	0.900	0.299	0	1.000	1.000
pred.prob.be²	0.905	0.293	0	1.000	1.000

¹ Pred.prob.be \(^1\) is the probability of making a Go decision to implement a pivotal trial, based on the unadjusted method, i.e., the proposed robust Bayesian hierarchical model with \(\gamma\) = 1
² Pred.prob.be \(^2\) is the probability of making a Go decision to implement a pivotal trial, based on the proposed robust Bayesian hierarchical model with \(\gamma\) defined in Eq. (7)

ISSN: 1471-2288