Function for creating sampling designs. — create_sampling

Sampling times can usually be pre-specified if the dosing schedule is fixed. In that case a scheme can be specified with static timepoints e.g. time = c(1, 7.5, 25, 31.5) for taking peak-trough samples at dose 1 and 4 for an 8-hour regimen. However, when the dosing interval is adaptive, the peak and trough sampling times cannot stay static and have to be adapted on-the-fly during simulations. In that case, it is possible to use when to specify sampling at peak, trough or dose concentrations. In that case, the time is not allowed. If not exact peak and trough concentrations should be sample, use the offset parameter to sample e.g. an hour post-peak, or shortly before true troughs.

Usage

create_sampling_design(
  time = NULL,
  when = NULL,
  offset = NULL,
  scatter = NULL,
  at = NULL,
  lloq = 0,
  anchor = c("dose", "day")
)

Arguments

time: a vector of numeric values. If no at values are specified, these will be used as the fixed sampling times in the simulated trial. If at values are supplied, the sampling times will be calculated adaptively during the trial. The at determine which dose or day is used as reference, and time will be relative to the specified at anchor.
when: character vector of same length as time (or single value) determining how to interpret the provided sampling time. If NULL will use the dose time as offset (default). Other options are: cmax or peak, which will use the end of infusion as the base for the time, or cmin or trough, which will use the time of next dose as the offset, or middle or cmid which will use the middle between the anchored dose and the next, or random which takes a random time point between the anchored dose and the next.
offset: offset from standardized PK moments specified in when, e.g. c(1, -1) with when = c("peak", "trough") to sample 1 hour after peak and 1 hour before trough.
scatter: optional random variation in time, specified as the standard deviation, e.g. scatter = 0.1 to allow for variation in sampling time with an SD of 0.1 hours. Only relevant for sampling times, not for regimen_update designs or target designs. Random variation does not protect for peaks or troughs becoming sampled during infusion or in previous / next dose. So value for scatter should be chosen appropriately and probably used in conjunction with approriate offset values.
at: numeric vector of the dose or day number to "anchor" the sampling times to. Vector needs to be of same length as time. If anchor is set to day, then the first dose in that day is used. If later doses in the day are preferred, the anchor can also be specified fractionally, e.g. 1.5 will use the time of the first dose in the second half of the 1st day.
lloq: lower limit of quantification for TDMs
anchor: either day or dose. Single value required, i.e. anchor types cannot be mixed.

Examples

# to sample at 1.5/8.5/23.5 hours
create_sampling_design(time = c(1.5, 11.5, 23.5))
#> $lloq
#> [1] 0
#> 
#> $scheme
#>   base offset at anchor
#> 1 dose    1.5  1   dose
#> 2 dose   11.5  1   dose
#> 3 dose   23.5  1   dose
#> 

# to sample at the peak (end of infusion) and trough (start of
# next dose) of each dose
create_sampling_design(
  when = c("peak", "trough")
)
#> $lloq
#> [1] 0
#> 
#> $scheme
#>     base offset at anchor scatter
#> 1   peak      0  1   dose       0
#> 2 trough      0  1   dose       0
#> 

# to sample 30 min post-peak and 30-min pre-trough for the first dose
create_sampling_design(
  when = c("peak", "trough"),
  offset = c(0.5, -0.5),
  at = c(1, 1)
)
#> $lloq
#> [1] 0
#> 
#> $scheme
#>     base offset at anchor scatter
#> 1   peak    0.5  1   dose       0
#> 2 trough   -0.5  1   dose       0
#> 

# The same as above, but now for the first and third dose
create_sampling_design(
  when = c("peak", "trough", "peak", "trough"),
  offset = c(0.5, -0.5, 0.5, -0.5),
  at = c(1, 1, 3, 3),
  anchor = "dose"
)
#> $lloq
#> [1] 0
#> 
#> $scheme
#>     base offset at anchor scatter
#> 1   peak    0.5  1   dose       0
#> 2 trough   -0.5  1   dose       0
#> 3   peak    0.5  3   dose       0
#> 4 trough   -0.5  3   dose       0
#>