Scaled (widened) BE Acceptance Limits
The (widened) scaled BE acceptance limits are calculated according to the regulatory settings of EMA, HC, FDA or via user defined regulatory settings. utf-8
scABEL(CV, regulator)
CV: Coefficient of variation (of the Reference) as ratio.
regulator: Regulatory body settings for the widening of the BE acceptance limits.
May be given as character from the choices "EMA", "HC" (Health Canada), "GCC" (Gulf Cooperation Council), "FDA" or as an object of class 'regSet' (see reg_const).
Defaults to regulator="EMA" if missing.
The former regulator="ANVISA" is no longer allowed. The ANVISA recommends since 2016 the EMA s regulatory settings.
The former regulator="USER" is no longer accepted but can be handled now via function reg_const() to define an object with class 'regSet'.
Returns a vector of lenghth 2 if one CV is given or a matrix if CV is given as vector with named components lower and upper of the scaled acceptance limits.
The widened BE acceptance limits are calculated by the formula
[L, U] = exp(-/+ r_const * sWR)
with r_const the regulatory constant and sWR the standard deviation of the within subjects variability of the Reference.
regulator="EMA" or regulator="HC"
r_const = 0.76 (~ log(1.25)/sqrt(log(0.30^2+1)))
regulator="GCC"
r_const = 0.97997... (= log(1/0.75)/sqrt(log(0.30^2+1)))
regulator="FDA"
r_const = 0.89257... (= log(1.25)/0.25)
If the CVwR of the Reference is < CVswitch=0.3 the conventional ABE limits apply (mixed procedure).
In case of regulator="EMA" a cap is placed on the widened limits if CVwR>0.5, i.e., the widened limits are held at the value calculated for CVwR=0.5.
In case of regulator="HC" the capping is done such that the acceptance limits are 0.6666 ... 1.5 at maximum, i.e., CVcap=0.57382. Literally it is given by Health Canada rounded to three significant digits as 57.4%.
The scaled acceptance limits (coined implied limits by Davit et al.) are not directly used in the BE evaluation for HVDP(s) recommended by the FDA. They are included here for comparative purposes. Moreover, there are controversies where to locate the implied limits and whether the so-called desired consumer-risk model should be used.
Davit BM, Chen ML, Conner DP, Haidar SH, Kim S, Lee CH, Lionberger RA, Makhlouf FT, Nwakama PE, Patel DT, Schuirmann DJ, Yu LX. Implementation of a Reference-Scaled Average Bioequivalence Approach for Highly Variable Generic Drug Products by the US Food and Drug Administration. AAPS J. 2012;14(4):915--24. tools:::Rd_expr_doi("10.1208/s12248-012-9406-x")
Health Canada, Therapeutic Products Directorate. Guidance Document. Comparative Bioavailability Standards: Formulations Used for Systemic Effects. 2018/06/08. ISBN: 978-0-660-25514-9
D. Labes
power.scABEL, sampleN.scABEL, reg_const
scABEL(CV = 0.3, regulator = "EMA") # should give the conventional (unscaled) BE limits: # lower upper # 0.80 1.25 scABEL(CV = 0.5, regulator = "EMA") # should give the (maximum) expanded limits: # lower upper # 0.6983678 1.4319102 # define old ANVISA settings via reg_const() rc <- reg_const("USER", r_const = 0.76, CVswitch = 0.4, CVcap = 0.5) rc$name <- "ANVISAold" scABEL(CV = 0.4, regulator = rc) # should give the conventional (not expanded) limits: # lower upper # 0.80 1.25 scABEL(CV = 0.55, regulator = "HC") # should give the widened limits: # lower upper # 0.6765789 1.4780241 scABEL(CV = 0.55, regulator = "GCC") # should give the widened limits: # lower upper # 0.750000 1.333333 scABEL(CV = 0.55, regulator = "FDA") # should give the 'implied' limits: # lower upper # 0.6320032 1.5822705