Defining estimands in anaesthesia clinical trials: a scoping review and framework

Intercurrent events, such as treatment switching and rescue medication, commonly occur in anaesthesia trials, and can affect the occurrence or interpretation of primary outcomes. The estimand framework provides strategies for handling these events, aligning trial objectives with outcomes.

We conducted a scoping review of RCTs with ≥100 participants published from January 2020 to June 2024 in ten high-impact journals, including patients receiving anaesthesia before surgery. We examined whether estimands were reported and identified intercurrent events. A structured framework was developed by mapping intercurrent events to intervention-outcome combinations investigated across all included anaesthesia trials.

Among 139 included trials, none explicitly used the estimand framework, and an estimand could be inferred in only 27 (19%). The primary outcome was defined in all trials, while other attributes were often not inferable: target population (44 [32%]), treatment condition (21 [15%]), summary measure (22 [16%]) and intercurrent event handling (84 [60%]). The most common intercurrent events were not receiving the randomised intervention (64 [46%]), treatment discontinuation (38 [27%]), and treatment switching (27 [19%]). Anaesthesia-related intercurrent events included surgery cancellation (37 [27%]) and non-protocolised anaesthesia management (17 [12%]). Findings informed the development of a structured, interactive web-based tool linking intercurrent events to intervention-outcome combinations, refined by expert review.

Use of estimands in anaesthesia trials is limited. Our tool supports structured identification and handling of intercurrent events in anaesthesia trials. Greater collaboration between statisticians and clinicians, with consistent application of the estimand framework, is needed to improve clarity and interpretability of future anaesthesia research.