Abstract:Background: Safety analyses play a pivotal role in drug development, ensuring the protection of patients while advancing innovative pharmaceuticals to market. A single study generally does not have sufficient sample size to evaluate all important safety events with reasonable precision and may not cover the full target population for the investigational treatment. Integrated analyses (pooled or meta-analysis) over several studies may be helpful in that regard. But without a structured conscious workflow accompanied with appropriate statistical methods for the integrated analysis, this can easily take a route compromising the interpretation. Methods: In this article we apply the ICH estimand framework to clinical trial integration and summarize respective critical statistical assumptions to ensure the integrated analyses are interpretable. Results: The estimand framework is valuable for developing principles for a deeper understanding of the critical statistical aspects of planning an integrated safety analysis. Our principles address the clinical question of interest, estimand and estimation. Special focus was given to the criteria for inclusion and exclusion of the component studies in the integrated analysis, and to integration of estimates pertaining to signal detection. Conclusion: Performing an integrated analysis and its preparatory steps calls for a good understanding of the clinical question of interest and its estimand, care and sound practice, to enable interpretation and avoid introducing unnecessary bias. It is valuable to use the estimand framework not only for efficacy evaluations, but also for safety evaluations in clinical trials and for integrated safety analyses.

Implementation of ICH E9 (R1): a few points learned during the COVID-19 pandemic

Assessing the Impact of COVID-19 on the Objective and Analysis of Oncology Clinical Trials -- Application of the Estimand Framework

Choosing Estimands in Clinical Trials: Putting the ICH E9(R1) Into Practice

Defining estimands using a mix of strategies to handle intercurrent events in clinical trials

Application of the estimand framework for an emulated trial using reference based multiple imputation to investigate informative censoring

Estimating hypothetical estimands with causal inference and missing data estimators in a diabetes trial

The Role of Occlusion: Potential Extension of the ICH E9 (R1) Addendum on Estimands and Sensitivity Analysis for Time-to-Event Oncology Studies

Time‐to‐event estimands and loss to follow‐up in oncology in light of the estimands guidance

Statistical methods for handling missing data to align with treatment policy strategy

Treatment Effect Quantification for Time-to-event Endpoints -- Estimands, Analysis Strategies, and beyond

Statistical Issues and Recommendations for Clinical Trials Conducted During the COVID-19 Pandemic

Decentralized Clinical Trials: Scientific Considerations Through the Lens of the Estimand Framework

Estimand Framework and Statistical Considerations for Integrated Analysis of Clinical Trial Safety Data

Time-to-event estimands and loss to follow-up in oncology in light of the estimands framework

Inference for Cumulative Incidences and Treatment Effects in Randomized Controlled Trials with Time-to-Event Outcomes under ICH E9 (R1)

Defining estimands in clinical trials: A unified procedure

Clinical Trials Impacted by the COVID-19 Pandemic: Adaptive Designs to the Rescue?

Bayesian and Frequentist Approaches to Rescuing Disrupted Trials: A Report from the NISS Ingram Olkin Forum Series on Unplanned Clinical Trial Disruptions

Conditional and Unconditional Treatment Effects in Randomized Clinical Trials: Estimands, Estimation, and Interpretation

Statistical methods for clinical trials interrupted by the severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) pandemic: A review

Implementation of tripartite estimands using adherence causal estimators under the causal inference framework