Estimands in randomised trials: rethinking common wisdoms (October 6, 2021 via “RoyalStatSoc”): 54-min video (37-min presentation & 17-min Q&A)

• Summary: Estimands, defined as a precise description of the intervention effect a researcher wants to estimate (the question one wishes to answer), are separate from the analytical methods (the approach to answering the question). Estimands are made up of five aspects: (1) Treatment condition(s), (2) Patient population, (3) Endpoint, (4) Population-level summary, and (5) Strategies to handle intercurrent events. Estimands make the interpretation of an intervention effect from a trial clearer; this is done through a specification of what the intervention effect represents. Dr. Brennan Kahan discusses how estimands can be used to ensure trials are addressing clinically relevant questions, and how they may make researchers rethink conventional wisdom around which analytical methods are appropriate for a given trial.

Kahan BC, et al. The estimands framework: a primer on the ICH E9(R1) addendum. BMJ. 2024 Jan 23;384:e076316. (12-page paper)

• Summary: Estimands can be used in studies of interventions, including pragmatic clinical trials, to clarify the interpretation of intervention effects. The addendum to the ICH E9 harmonised guideline on statistical principles for clinical trials (ICH E9(R1)) describes a framework for using estimands within a study. This primer provides an overview of the estimands framework (as outlined in the addendum) with the aim of explaining why estimands are beneficial. Clarifications on terminology and practical guidance on using estimands to decide the appropriate analytical design are also provided.

Estimands considerations in cluster randomized trials (HDRN Canada Pragmatic Trials Training Program): 12-min video.

• Summary: Dr. Fan Li discusses estimands considerations in pragmatic cluster randomized trials. The five components of an estimand are overviewed, and the concept of specifying estimands before planning the analysis of the trial is emphasized. It is recommended that interested researchers also review literature related to estimands in traditional clinical trials, as well as the (causal inference) literature describing estimands in the observational research setting. The need for statisticians to help navigate the analytical complexities in pragmatic trials is discussed as well. A related paper can be found here.

Cro S, et al. Starting a conversation about estimands with public partners involved in clinical trials: a co-developed tool. Trials. 2023 Jul 6;24(1):443. (6-page paper)

• Summary: The use of estimands helps to ensure trials are designed and analyzed to answer the questions relevant to patients and the public. This study: (1) co-developed a tool with public partners that helps explain what an estimand is and (2) explores public partner’s perspectives on the importance of discussing estimands during trial design.

Kahan BC, et al. We must let the research question drive study methods. BMJ. 2024 Jan 23;384:q173. (2-page editorial)

• Summary: Briefly emphasizes the point that the research question should drive the analytical methods, and not the other way around. Estimands, which have been discussed and adopted by researchers and regulators around the world, serve to remind trialists of this process.

International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH Harmonised Guideline) - Addendum on Estimands and Sensitivity Analysis in Clinical Trials to the Guideline on Statistical Principles for Clinical Trials - E9(R1) (November 20, 2019): 22-page document.

• Summary: This addendum presents a framework to strengthen the dialogue among those involved in the formulation of randomized trial objectives, design, implementation, analysis, and interpretation. This framework focuses on clarifying specific details of the intervention effect one intends to quantify within a trial.

NIH Pragmatic Trials Collaboratory - Grand Rounds: Estimands in Cluster-Randomized Trials: Choosing Analyses that Answer the Right Question (March 10, 2023): 50-min webinar (31-slide presentation)

• Summary: Dr. Brennan Kahan discusses estimands in the context of a pragmatic cluster randomized trial, where it is emphasized that estimands in cluster trials have additional considerations compared to individually randomized trials.

Hemming K, Taljaard M. Commentary: Estimands in cluster trials: thinking carefully about the target of inference and the consequences for analysis choice. Int J Epidemiol. 2023 Feb 8;52(1):116-118. (3-page commentary)

• Summary: It is argued that the need to consider the estimand of interest before specifying the analytical methods has not received adequate attention in the cluster randomized trials literature. This commentary discusses the Kahan BC, et al. (2023) article titled: "Estimands in cluster-randomized trials: choosing analyses that answer the right question."