α and no ω [Power / Sample Size]
❝ I am suffering trying to understand the problems and perspectives of alpha-adjustment.
You are not alone.

❝ Wherever I look I see TIE inflation...
❝ For HVD:
❝ – FDA RSABE seemed to have problems for CV around 30%
Only ≤30% (see there).
❝ For well-known adaptive designs (Potvin, Xu...) TIE also behaves badly...
For most of them, not at all. Ones with a slight (!) inflation can be handled be more adjustment. Xu is fine.
❝ TIE may be also inflated by evaluating different methods for PK metrics (when one of them turns to be overpowered - as in the case of adaptive design or different CI for two Cmax and AUC).
Theoretically IUT should protect us. Benjamin and Detlew are working on a new function
power.2TOST.sim()
and updated sampleN.2TOST()
in PowerTOST
. As expected the overall TIE is much lower than the single ones.❝ I suspect TIE will be also inflated in studies of NTDs also (by using FDA approach or by using different confidence limits for two metrics).
Might well be.
❝ Now you talk about higher-order design and dose-proportional studies...
That’s an endless story.
❝ So... Are there any ways to deal with it excepting the bright idea of iteratively adjusted alpha?
THX for calling it bright! For reference-scaling a new method* by the two Lászlós control the TIE much better than RSABE/ABEL – but only for full replicate designs. Yet another good reason to avoid the bloody partial replicate. Detlew already implemented a new function
power.RSABE2L.sdsims()
in PowerTOST
. You need the development version on GitHub to run this code:library(PowerTOST)
CV <- 0.3
n <- 24
design <- "2x2x4"
U.EMA <- scABEL(CV, regulator="EMA")[["upper"]]
U.FDA <- scABEL(CV, regulator="FDA")[["upper"]]
st <- proc.time()[[3]]
TIE <- data.frame(method=c("EMA ABEL", "EMA ABEL adj", "FDA RSABE",
"FDA RSABE adj", "2L FDA", "2L Hyslop",
"2L exact"),
TIE=rep(NA, 7), stringsAsFactors=FALSE)
res <- scABEL.ad(CV=CV, theta0=U.EMA, n=n, design=design,
regulator="EMA", print=FALSE, sdsims=TRUE, nsims=1e6)
TIE[1, 2] <- res[["TIE.unadj"]]
TIE[2, 2] <- res[["TIE.adj"]]
res <- scABEL.ad(CV=CV, theta0=U.FDA, n=n, design=design,
regulator="FDA", print=FALSE, nsims=1e6)
TIE[3, 2] <- res[["TIE.unadj"]]
TIE[4, 2] <- res[["TIE.adj"]]
if (as.character(packageVersion("PowerTOST")) >= "1.4.6.9000") {
TIE[5, 2] <- power.RSABE2L.sds(CV=CV, theta0=U.FDA, n=n, design=design,
SABE_test="hyslop", nsims=1e6)
TIE[6, 2] <- power.RSABE2L.sds(CV=CV, theta0=U.FDA, n=n, design=design,
SABE_test="fda", nsims=1e6)
TIE[7, 2] <- power.RSABE2L.sds(CV=CV, theta0=U.FDA, n=n, design=design,
SABE_test="exact", nsims=1e6)
}
rt <- proc.time()[[3]]-st
print(TIE, row.names=FALSE); cat("Runtime:", round(rt/60, 1), "minutes\n")
method TIE
EMA ABEL 0.080639
EMA ABEL adj 0.050000
FDA RSABE 0.133510
FDA RSABE adj 0.049999
2L FDA 0.062536
2L Hyslop 0.063513
2L exact 0.066088
Runtime: 4.6 minutes
❝ While looking through the literature I've found only some suggestions of modificating ABE (with new procedures or nonlinear CI limits) or alternatives for ABE (like GSD or Two-stage designs)...
❝
- Knahl SIE, Lang B, Fleischer F, Kieser M. A comparison of group sequential and fixed sample size designs for bioequivalence trials with highly variable drugs. Eur J Clin Pharmacol. 2018;74(5):549–59. doi:10.1007/s00228-018-2415-7.
❝
- Molins E., Cobo E., Ocaña J. Two-stage designs versus European scaled average designs in bioequivalence studies for highly variable drugs: Which to choose? Stat Med. 2017;36(30):4777–4788. doi:10.1002/sim.7452.
One of the authors of [1] is user Ben in the Forum and a co-author of
PowerTOST
. Incidentally I reviewed both papers. Demanding but great fun at the end. 
- Tóthfalusi L, Endrényi L. Algorithms for evaluating reference scaled average bioequivalence: Power, bias, and consumer risk. Stat Med. 2017;36(27):4378–90. doi:10.1002/sim.7440.
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png)
Helmut Schütz
![[image]](https://static.bebac.at/img/CC by.png)
The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
Complete thread:
- Sample size for 4-period 4-sequence crossover BE study Bryony Simmons 2018-02-01 12:16 [Power / Sample Size]
- function sampleN.TOST of package PowerTOST Helmut 2018-02-01 13:01
- Alpha adjustment in higher order crossover d_labes 2018-02-01 14:01
- Deficiencies Helmut 2018-02-01 15:48
- Deficiencies nobody 2018-02-01 17:10
- Deficiencies d_labes 2018-02-01 18:57
- Deficiencies Relaxation 2018-02-02 11:12
- Deficiencies nobody 2018-02-02 12:49
- Deficiencies Helmut 2018-02-02 16:14
- Deficiencies Relaxation 2018-02-02 19:41
- alpha... where is omega? Astea 2018-02-02 21:45
- α and no ωHelmut 2018-02-02 23:39
- TIE for NTIDs d_labes 2018-02-04 12:40
- TIE for NTIDs Astea 2018-02-04 20:04
- TIE for NTIDs Helmut 2018-02-05 01:01
- TIE for NTIDs d_labes 2018-02-05 16:40
- TIE for NTIDs Helmut 2018-02-05 17:49
- TIE for NTIDs d_labes 2018-02-05 22:17
- TIE for NTIDs Helmut 2018-02-06 12:34
- TIE for NTIDs d_labes 2018-02-05 22:17
- TIE for NTIDs Helmut 2018-02-05 17:49
- TIE for NTIDs d_labes 2018-02-05 16:40
- TIE for NTIDs d_labes 2018-02-05 16:35
- bow TIE for NTIDs Astea 2018-02-05 17:52
- bow TIE for NTIDs Helmut 2018-02-05 18:10
- 111.11 for NTIDs Astea 2018-02-05 19:27
- 111.11 for NTIDs Helmut 2018-02-06 00:12
- 111.11 for NTIDs Astea 2018-02-05 19:27
- bow TIE for NTIDs d_labes 2018-02-05 22:33
- bow TIE for NTIDs Helmut 2018-02-05 18:10
- bow TIE for NTIDs Astea 2018-02-05 17:52
- TIE for NTIDs Helmut 2018-02-05 01:01
- TIE for NTIDs Astea 2018-02-04 20:04
- TIE for NTIDs d_labes 2018-02-04 12:40
- α and no ωHelmut 2018-02-02 23:39
- alpha... where is omega? Astea 2018-02-02 21:45
- Deficiencies Relaxation 2018-02-02 19:41
- Deficiencies Relaxation 2018-02-02 11:12
- Deficiencies Helmut 2018-02-01 15:48
- Alpha adjustment in higher order crossover d_labes 2018-02-01 14:01
- function sampleN.TOST of package PowerTOST Helmut 2018-02-01 13:01