pkm: Posterior concentration estimates

In hlweeks/pkpredict: Compartment Model Implementation for Pharmacokinetic Analysis

Posterior concentration estimates

Description

Fits a two-compartment model to obtain posterior estimates of concentration of drug over time.

Usage

pkm(
  formula,
  data,
  subset,
  ivt,
  pars = c(getOption("pkpredict.pip.default.prior")$log_pk_mean,
    getOption("pkpredict.pip.default.prior")$log_err_mean),
  alp = 0.05,
  cod = 12,
  thres = 64,
  timeint = c(0, max(sapply(ivt, function(x) x$end)) + cod),
  mcmc = FALSE,
  nreps = 5000,
  nburnin = 2000,
  nthin = 10,
  seed = NULL,
  shiny = FALSE,
  ...
)

Arguments

formula	A formula where the left side is the measured concentration of drug and the right side is the times of concentration measurements
data	Data frame with concentration data (time of measurement in hours and concentration in mcg/ml)
subset	Subset of the dat data frame to use
ivt	List with containing start of infusion times, end of infusion times, and rate of infusion at each dose
pars	Vector of (prior) log-pharmacokinetic parameters of length 5: (lv_1, lk_10, lk_12, lk_21, lerr)
alp	Value of alpha to use for generating pointwise (1 - alp)% confidence bands
cod	Length of time after end of last dose to consider
thres	Threshold for effective treatment (mcg/ml)
timeint	time interval over which to compute estimate
mcmc	logical: should estimate of time above threshold be computed using MCMC (false = laplace approximation)
nreps	number of MCMC iterations to perform (including burn in)
nburnin	number of burn in replications to perform
nthin	mcmc thinning interval
seed	seed for replicating MCMC results
shiny	is this being used within shiny_pkm
...	additional arguments (e.g., 'mu', 'sig', 'ler_mean', 'ler_sdev' for changing the PK parameter prior mean, variance-covariance matrix and error prior mean and standard deviation, respectively)

Details

Measurements must be entered in particular units: mcg/ml for concentrations, g/h in rate of infusion, hours for times.

Value

posterior estimates

Examples

ivt_d <- list(list(begin=0.0, end=0.5, k_R=6),
              list(begin=8.0, end=8.5, k_R=6),
              list(begin=16.0, end=16.5, k_R=6))
dat_d <- data.frame(time_h = c(1,4,40), conc_mcg_ml = c(82.7,80.4,60))

pkm(conc_mcg_ml ~ time_h, data = dat_d, ivt = ivt_d)

hlweeks/pkpredict documentation built on Oct. 29, 2023, 6:08 a.m.