clinical-trial-design-patternslisted

# Clinical Trial Design Patterns ## When to Use This Skill - Selecting appropriate trial design for clinical objectives - Implementing multi-arm or multi-endpoint trials - Designing stratified trials - Planning adaptive designs - Understanding design trade-offs ## Two-Arm Parallel Design ### Standard Design The most common design: randomize patients to treatment or control. ```r # simtrial implementation sim_pw_surv( n = 400, block = c(rep("control", 1), rep("experimental", 1)), # 1:1 enroll_rate = data.frame(rate = 20, duration = 12), fail_rate = fail_rate ) ``` ```r # Mediana implementation DataModel() + OutcomeDist(outcome.dist = "NormalDist") + SampleSize(200) + # Per arm Sample(id = "Control", outcome.par = parameters(mean = 0, sd = 1)) + Sample(id = "Treatment", outcome.par = parameters(mean = 0.5, sd = 1)) ``` ### Unequal Randomization **When to Use:** - Increase exposure to experimental treatment - Ethical considerations - Resource optimization ```r # 2:1 randomization (experimental:control) sim_pw_surv( n = 300, block = c("control", rep("experimental", 2)) ) # Mediana with unequal allocation DataModel() + Sample(id = "Control", sample.size = 100, ...) + Sample(id = "Treatment", sample.size = 200, ...) ``` **Trade-off:** Unequal allocation reduces power for same total N. ## Multi-Arm Designs ### Dose-Finding (Multiple Doses vs Placebo) ```r # Three doses + placebo DataModel() + OutcomeDist(outcome.dist = "NormalDist") + Sam