EMA/2020/46/TDA, Lot 3 - ROC36 Use of causal inference methods commonly used in non-interventional studies to estimate treatment effects in clinical trials

Study type

• Study topic:
  Other
• Study topic, other:
  Causal inference methods in randomized controlled trials
• Study type:
  Not applicable

Study design details

• Study design:
  Review of published literature and regulatory documents, simulation study
• Main study objective:
  (1) To screen the scientific literature and regulatory documents to identify ; (a) randomized clinical trials in which causal inference methods have been used to test and estimate treatment effects, and ; (b) methodological approaches based on causal inference for testing and estimating treatment effects in randomized clinical trials. ; (2) (a) To identify relevant clinical trial scenarios in which the causal pathway from randomized treatment to outcome may be affected by intercurrent events; ; (b) To identify causal inference methods suitable for testing and estimating treatment effects within the estimand framework that appropriately account for these intercurrent events.; (3) To assess the statistical properties (including bias, coverage probabilities, type 1 error rate, and power) of the selected methods in a comprehensive simulation study based on clinical trial data generated under varying scenarios. ;