de.sim: Simulate a '3+3/PC' dose-titration study

In dcnorris/DTAT: Dose Titration Algorithm Tuning: a framework for dose individualization

Description

Please see the reference cited below.

Usage

de.sim(CV = 0.7, mean_mtd = 1, start.dose = 0.25, dunit = "mg",
  dose.jump = 0.4, N = 24, periods = N/3 + 2,
  testing = is.null(sys.call(-1)), once = testing, ...)

Arguments

CV	Coefficient of variation for MTDi in the population
mean_mtd	Mean MTDi in the population
start.dose	Lowest dose in geometric sequence of doses to trial
dunit	A string specifying the dose unit
dose.jump	The fractional increase between dose levels
N	Number of simulated subjects to enroll
periods	Number of DLT assessment periods to simulate
testing	This parameter helps differentiate function behavior (type of value returned) appropriately to interactive vs programmed invocation
once	When TRUE, does a single simulation returning de object
...	Additional arguments passed to function step

Value

When once=TRUE, a list of data frames summarizing the events of a dose-titration study up to a given DLT assessment period. Thus, de[[3]] is a data frame with columns as follows:

• id Integer: participants are numbered in order of enrollment, starting from 1

• period Integer: DLT assessment periods of the study, numbered from 1

• dose Integer: dose level received by id during period

• dlt Logical: did participant id experience a DLT during period?

For periods after dose escalation stops, a 'stop.esc' attribute will be attached to subsequent period elements. Thus, if escalation stops at the end of period 7, attr(de[[8]],'stop.esc') will equal 7, and this will be true for all subsequent periods as well.

When testing=FALSE, a simple summary result is returned, suitable for aggregating an ensemble of simulations.

Author(s)

David C. Norris

References

Norris DC. Precautionary Coherence Unravels Dose Escalation Designs. bioRxiv. December 2017:240846. doi:10.1101/240846. https://www.biorxiv.org/content/early/2017/12/29/240846

See Also

step, which de.sim calls recursively

Examples

# Perform a regression test for package:DTAT
de.compare <- de.sim(testing=TRUE)
# Strip down to basics from v0.2-1
attributes(de.compare) <- NULL
class(de.compare) <- "list"
data(de.bioRxiv.240846)
stopifnot(identical(de.compare, de.bioRxiv.240846))

dcnorris/DTAT documentation built on May 7, 2019, 10:45 p.m.