sampleN.HVNTID: Sample size estimation for BE decision via FDA method for...

In Detlew/PowerTOST: Power and Sample Size for (Bio)Equivalence Studies

Sample size estimation for BE decision via FDA method for highly variable (HV) narrow therapeutic index drugs (NTIDs)

Description

This function performs the sample size estimation for the BE decision via the FDA’s method for highly variable NTIDs as described in respective guidances based on simulations.
The study designs may be the full replicate design 2x2x4 with 4 periods (TRTR|RTRT) and the 3-period replicate design 2x2x3 with sequences RTR|TRT.

Usage

sampleN.HVNTID(alpha = 0.05, targetpower = 0.8, theta0, theta1, theta2, CV,
               design = c("2x2x4", "2x2x3"), nsims = 1e+05, nstart, imax = 100,
               print = TRUE, details = TRUE, setseed = TRUE)

Arguments

alpha	Type I error probability. Per convention mostly set to 0.05.
targetpower	Power to achieve at least. Must be >0 and <1. Typical values are 0.8 or 0.9.
theta0	‘True’ or assumed T/R ratio. Defaults to 0.95 if not given explicitly.
theta1	Conventional lower ABE limit to be applied in the FDA procedure. Defaults to 0.8 if not given explicitly.
theta2	Conventional upper ABE limit to be applied in the FDA procedure. Defaults to 1.25 if not given explicitly.
CV	Intra-subject coefficient(s) of variation as ratio (not percent). If given as a scalar (length(CV)==1) the same CV of Test and Reference is assumed (homoscedasticity, CVwT==CVwR). If given as a vector (length(CV)==2), i.e., assuming heteroscedasticity, the CV of the Test must be given in CV[1] and the one of the Reference in the CV[2].
design	Design of the study to be planned. "2x2x4" is the full replicate with 2 sequences and 4 periods (TRTR|RTRT). "2x2x3" is the full replicate with 2 sequences and 3 periods (TRT|RTR). Defaults to design="2x2x4".
nsims	Number of simulations to be performed to obtain the empirical power. Defaults to 100,000 = 1e+5.
nstart	Set this to a start value for the sample size if a previous run failed. May be missing.
imax	Maximum number of steps in sample size search. Defaults to 100.
print	If ⁠TRUE⁠ (default) the function prints its results. If ⁠FALSE⁠ only the resulting dataframe will be returned.
details	If set to TRUE, the default, the steps during sample size search are shown. Moreover the details of the method settings are printed.
setseed	Simulations are dependent on the starting point of the (pseudo) random number generator. To avoid differences in power values for different runs a set.seed(123456) is issued if setseed=TRUE, the default.

Details

For deciding BE the study must pass the conventional ABE test and additionally the test that the ratio of sWT/sWR is <= 2.5.

The simulations are done via the distributional properties of the statistical quantities necessary for deciding BE based on these method.
Details can be found in a document ⁠Implementation_scaledABE_sims⁠ located in the /doc sub-directory of the package.

The estimated sample size gives always the total number of subjects (not subject/sequence – like in some other software packages).

Value

Returns a data.frame with the input and sample size results.
The Sample size column contains the total sample size.
The nlast column contains the last n value. May be useful for re-starting.

Warning

For some input constellations the sample size search may be very time consuming and will eventually also fail since the start values chosen may not really reasonable for them.
In case of a failed sample size search you may restart with setting the argument ⁠nstart⁠.

Note

The design recommended by the FDA is the full replicate design "2x2x4".
The sample size estimation is done only for balanced studies since the break down of the total subject number in case of unbalanced sequence groups is not unique. Moreover the formulas used are only valid for balanced designs.
The FDA method is described for the ABE limits 0.8 ... 1.25 only. Setting theta1, theta2 to other values may not be reasonable and is not tested.
The minimum sample size is 6, even if the power is higher than the intended targetpower.

The method is also required by China’s Center of Drug Evaluation.

Author(s)

D. Labes

References

Food and Drug Administration, Office of Generic Drugs (OGD). Draft Guidance on Dabigatran Etexilate Mesylate. Recommended Jun 2012; Revised Sep 2015, Jul 2017. download

Food and Drug Administration, Office of Generic Drugs (OGD). Draft Guidance on Rivaroxaban. Recommended Sep 2015. download

Food and Drug Administration, Office of Generic Drugs (OGD). Draft Guidance on Edoxaban Tosylate. Recommended May 2017; Revised Mar 2020. download

See Also

power.HVNTID
and power.NTIDFDA, sampleN.NTIDFDA for NTIDs with low variability

Examples

# using all defaults but CV
sampleN.HVNTID(CV = 0.3)
# should give
# n=22 with an (empirical) power of 0.829700

# Test formulation with lower variability but same pooled CV
CVs <- CVp2CV(0.3, ratio = 0.25)
sampleN.HVNTID(CV = CVs)
# should give (no distinct difference to example above)
# n=22 with an (empirical) power of 0.837520

Detlew/PowerTOST documentation built on March 26, 2024, 12:18 a.m.