Handling intercurrent events and missing data in non-inferiority trials using the estimand framework: A tuberculosis case study

Using a tuberculosis clinical trial as a case study, this paper proposes a primary estimand, and an additional estimand suitable for non-inferiority studies.