Generate an Adverse Event table — example

Internal function that simulates adverse events data with a toxicity effect that depends on the treatment arm enrolres$arm. The effect is not simulated according to enrolres$arm3.

Usage

example_ae(
  enrolres,
  seed,
  p_na = 0,
  p_sae = 0.1,
  p_sae_trt = p_sae,
  n_max = 15,
  n_max_trt = n_max,
  w_soc = 1,
  w_soc_trt = 1,
  beta0 = -1,
  beta_trt = 0.4,
  beta_sae = 1,
  ...
)

Arguments

enrolres: the enrolment result table, from example_enrol().
seed: Integer. Random seed for reproducibility (can be NULL).
p_na: proportion of missing values (can be a list with a value for each column)
p_sae, p_sae_trt: proportion of serious AE in control/exp arms
n_max, n_max_trt: maximum number of AE per patient in control/exp arms (binomial with probability 20%)
w_soc, w_soc_trt: log-weights for SOC that should be over-representated in control/exp arms.
beta0, beta_trt, beta_sae: the intercept, treatement coef and SAE coef to be used in the exponential decay model that generates the AE grade.

Value

A tibble with N rows and the following columns:

subjid: the patient identifier. Each patient has a number of AE simulated with a binomial distribution with size n_max or n_max_trt and probability 20%.
aesoc: the CTCAE System Organ Class. Lazily simulated with an uniform probability.
aeterm: the CTCAE High Level Group Term. Lazily simulated with an uniform probability. Four examples of HLGT per SOC are provided.
aegr: the CTCAE grade, ranging from 1 to 5. The probability is simulated using an exponential decay rate for grades 1 to 4 (probs = exp(rate * c(1:4))), with probability for grade 5 being the one for grade 4 divided by 7. The probability vector is then normalized to sum to 1. The rate is calculated as rate = beta0 + beta_trt*(arm!="Control") + beta_sae*(sae=="Yes").
aerel: the causality of the AE. Lazily simulated with an uniform probability.
sae: Indicator of Serious AE. Simulated using p_sae and p_sae_trt.