Adjust dosing using MIPD on TDMs at specified dose numbers — sample_and_adjust_by

Doses are iteratively adjusted for the specified dose numbers. Estimation is performed by MAP Bayesian estimation, and simulation is used to identify the dose most likely to reach the specified target. See dose_grid_search for dose-finding logic. Optionally, disparate models can be used for simulation and estimation, to allow for intentional model misspecification. Covariates supplied must support both models. True patient parameters (pars_true_i) must match the model used for simulation.

Usage

sample_and_adjust_by_dose(
  regimen_update_design,
  sampling_design,
  target_design,
  regimen,
  covariates = NULL,
  pars_true_i,
  sim_model,
  sim_ruv = NULL,
  verbose = FALSE,
  accumulate_data = TRUE,
  ...
)

Arguments

regimen_update_design: a data.frame with scheme with details on how and when to update the regimen in the MIPD trial. Schemes are created using create_regimen_update_design().
sampling_design: a data.frame with a sampling scheme, created using create_sampling_design().
target_design: target design created using create_target_design()
regimen: PKPDsim regimen object, containing initial dosing regimen.
covariates: named list of PKPDsim covariates.
pars_true_i: PK parameters for the individual. See generate_iiv.
sim_model: model to use for simulating "true" patient response.
sim_ruv: residual variability for sim_model. Named list for proportional (prop) and additive (add) error.
verbose: verbose output?
accumulate_data: if TRUE, will use all available data up until the adjustment timepoint. If set to FALSE, will use only the data since the last adjustment timepoint and the current one.
...: arguments passed on to simulate_fit or dose_optimization_method function.

Value

a named list containing final_regimen (all doses after adjustment), tdms (all collected levels, both true and measured, that is, both with and without residual variability), and additional_info, which varies by dose_optimization_method. See selected function for details.

Details

Trial design or MIPD protocol can vary considerably in complexity. This function is intended to cover a broad range of relatively straight forward designs, and may serve as a template for more complex designs.