Expected number of observed events in each arm of the target trial
Source:R/simulation_time_to_event.R
time_to_event_expected_events.RdThe number of events is what loss to follow-up, the event time distribution and control-arm heterogeneity change in the time-to-event trial, and what the precision of its log hazard ratio rests on. Uses the same arm parameters and censoring as the data generator.
Usage
time_to_event_expected_events(
case_study_config,
sample_size_per_arm,
control_drift = 0,
dropout_probability = 0,
event_time_distribution = "exponential",
treatment_effect = case_study_config$source$treatment_effect,
treatment_delay = 0
)Arguments
- case_study_config
A time-to-event case study configuration.
- sample_size_per_arm
Number of patients in each arm.
- control_drift
Control-arm heterogeneity, kappa, on the log scale.
- dropout_probability
Probability of loss to follow-up over the maximum follow-up time.
- event_time_distribution
Either "exponential" or "weibull".
- treatment_effect
Target log hazard ratio; the source estimate by default, i.e. a consistent treatment effect. Under a delayed effect, the Cox model's large-sample limit (see
time_to_event_delayed_log_hr()).- treatment_delay
Time before the treatment effect starts, in years.