blk.noninformativeDose: Flag Non-informative Dose Events

In benjaminrich/PCSmisc: Useful functions for longitudinal data set construction and manipulation.

Flag Non-informative Dose Events

Description

This function will flag doses that are non-informative in a PK data set because they either occur after the last observation event or they precede a dose given at steady state with no observation events occurring between the two.

Usage

blk.noninformativeDose(id, dose.ind, ss.ind = NULL, obs.ind = !dose.ind)

Arguments

id	A valid block-format ID.
dose.ind	A logical vector in block-format with respect to id. Indicates rows corresponding to dose events.
ss.ind	A logical vector in block-format with respect to id. Indicates doses given at steady state.
obs.ind	A logical vector in block-format with respect to id. Indicates rows corresponding to observation events.

Details

This function operates on a PK data set in block-format. Rows that corresponding to dose events are indicated by dose.ind, and rows that correspond to doses taken at steady-state are indicated by ss.ind. A steady-state event renders any dose events that precede it with no intervening observation events non-informative. Doses that occur at the end of a block (i.e. after the last observation event) are also considered non-informative. Non-informative dose events can effectively be removed from the data without affecting the analysis of these data.

Value

A logical vector in block-format with respect to id indicating dose events that are non-informative.

Author(s)

Benjamin Rich <[e-mail]>

See Also

• block-format

• blk.isSteadyState

Examples

require(nlme)
data(Phenobarb)
dat <- Phenobarb[1:56,]  # First 4 subjects
attach(dat)

cbind(dat, INTERDOSE.TIME=blk.intereventTime(time, asID(Subject), !is.na(dose)),
           IS.SS=blk.isSteadyState(time, asID(Subject), !is.na(dose), 12, 0.301, 60, dose),
           FLAG=blk.noninformativeDose(asID(Subject), !is.na(dose), blk.isSteadyState(time, asID(Subject), !is.na(dose), 12, 0.301, 60, dose)))

detach(dat)

dat2 <- data.frame(id=rep(1:2, each=10), dose=0, obs=0)
dat2$dose[c(1:3, 7:8, 12:13, 19:20)] <- 1
dat2$obs[c(5:6, 15:18)] <- 1
dat2$flag <- with(dat2, 1*blk.noninformativeDose(asID(id), dose.ind=(dose==1), obs.ind=(obs==1)))
dat2

benjaminrich/PCSmisc documentation built on Feb. 11, 2024, 9:25 p.m.