intercurrentevent

An intercurrent event is an event that occurs after treatment initiation and affects the interpretation or measurement of a trial’s outcome. In clinical trials, common intercurrent events include treatment discontinuation, addition of rescue or supplementary therapy, withdrawal of consent, death, and loss to follow-up. These events can occur in any arm and may change what the study is measuring, or prevent observation of the originally planned outcome, making explicit handling essential for valid interpretation.

To address this, modern trial design often uses the estimand framework, as formalized in the ICH E9(R1) addendum. An estimand specifies the target population, the variable to be used as the endpoint, the population summary, and, crucially, how intercurrent events are handled. Several strategies are used to address intercurrent events, depending on the clinical question.

Strategies include treatment policy (analyzing outcomes regardless of intercurrent events), hypothetical (estimating the outcome if the event had not occurred), while-on-treatment (using data up to the intercurrent event), a composite approach (treating the intercurrent event as part of the endpoint), and principal stratification (estimating effects within subgroups defined by the potential occurrence of the event).

The choice of estimand shapes interpretation and regulatory decisions and is central to causal inference; intercurrent events clarify what a trial result means for real-world use. By making the handling of post-randomization events explicit, researchers can align trial conclusions with the intended clinical question and the practical realities of patient care.