Presentation (CTI)

OBJECTIVES: The ICH E9 (R1) addendum introduction aimed at improving clarity in the definition of the treatment effect of interest in clinical research through the estimand framework. Despite the main focus on randomized clinical trials (RCTs), the framework is also intended to be applicable to observational studies (ObS). The estimands adoption in ObS is still limited, despite the greater complexity of clearly measuring treatment effect compared to RCTs. This work provides an explorative overview of the estimand framework current implementation status in ObS conducted by IQVIA Solutions Italy s.r.l. and future developments.
METHODS: We reviewed the Study Protocols (SPs) of 28 multi-country and local studies, with different observational study design, written from Jan 2021 to May 2025. Our analysis focused on estimand attributes use in the primary objective formulation, together with eligibility and exit criteria, primary endpoint, and related statistical methodologies definition. We also examined intercurrent events (ICEs) management strategies.
RESULTS: Sixteen (57.1%) ObS were drug studies, 2 device studies. Estimand methodology was formally implemented in only one of the analyzed SPs, while others included relevant attributes distributed across various SP sections. The estimand attributes were provided in 14 (50.0%) studies, with reference to ICEs missing in 3 (10.7%). In 9 (32.1%) SPs not providing all estimands attributes, some ICEs strategies were mentioned. Overall, the most frequent ICE strategy was the 'while on treatment' in 11 (39.3%) studies, as the most common ICE was treatment withdrawal. In 4 (14.3%) studies the 'irrespective of' approach was considered. Treatment switch was considered in 2 (7.1%) cases.
CONCLUSIONS: Formal use of estimand framework in ObS appears limited. Specific real-world evidence guidelines are being developed by working groups to support estimand implementation in this context.