cir (version 2.5.1)

DRshrink: Shrinkage fix to mitigate bias in observed rates, under adaptive dose-finding designs

Description

Adaptive dose-finding designs induce a bias on observed rates away from the target dose. This is well-known in other adaptive-design fields, but has been overlooked by the dose-finding research community. Flournoy and Oron (2020) examine the bias in the dose-finding context, and suggest a simple shrinkage fix that reduces both bias and variance. The fix is analogous to the empirical-logit fix for zero counts in binary data, but instead of adding 0.5 to each cell, target is added to the 1's at each dose, and 1-target to the 0's. The shrinkage is applied to the raw observation, so CIR or IR are carried out on the shrunk data.

Usage

DRshrink(y, x = NULL, wt0 = NULL, target, swt = 1, nmin = 2, ...)

Arguments

y

can be either of the following: y values (response rates), a 2-column matrix with positive/negative response counts by dose, a DRtrace object or a doseResponse object.

x

dose levels (if not included in y).

wt0

weights (if not included in y).

target

the balance point (between 0 and 1) around which the design concentrates allocations.

swt

the weight of the shrinkage. Default 1 (a single observation)

nmin

the minimum n at each dose, for the shrinkage to be applied. Default 2.

...

parameters passed on to doseResponse()

Author

Assaf P. Oron <assaf.oron.at.gmail.com>

References

Flournoy N and Oron AP, 2020. Bias Induced by Adaptive Dose-Finding Designs. Journal of Applied Statistics 47, 2431-2442.

Examples

## Summary of raw data from the notorious Neuenschwander et al. (Stat. Med., 2008) trial
## Note the use of the 'cohort' argument to specify the cohort order
neundatDose = doseResponse(x=c(1,2.5,5,10,20,25), y = c(rep(0,4),2/9,1), wt = c(3,4,5,4,9,2) )

neundatDose

# Compare to this:
DRshrink(neundatDose, target = 0.3)