Top Things to Know: Methodological Standards for Follow-up Extension in Cardiovascular Trials
Updated: August 03, 2026
Prepared by Alexander C. Gregg, MD General Surgery Resident and Research Fellow Department of Cardiothoracic Surgery, Weill Cornell Medicine
- An extension follow-up study reports outcomes for the entire original randomized cohort beyond the follow-up duration of the primary endpoint. It excludes investigations such as post hoc analyses and subgroup-only extensions and may yield findings that differ from the initial report and have greater implications for clinical practice.
- In extension follow-up studies the primary endpoint is assessed after the initial trial period and may differ from that of the parent trial; endpoint selection aligned with the clinical question and prespecification helps minimize bias.
- The duration of extension follow-up should balance disease biology, expected temporal dynamics of treatment effects and practical trial constraints, with the goal of adequately capturing both efficacy and safety outcomes.
- Loss to follow-up can be minimized through proactive patient engagement, clear articulation of long-term goals and flexible visit options, which may require additional (re)consent. Follow-up completeness is reported by randomized group, with reasons for missingness clearly specified.
- For the primary analysis, the intention-to-treat population serves as the anchor estimate to preserve the benefits of initial randomization, while estimand-aligned per-protocol and/or as-treated analyses provide supportive context with censoring rules clearly defined.
- Endpoint definitions and adjudication processes remain most interpretable when they are consistent across trial phases, with any mapping to external datasets documented and accompanied by corresponding robustness analyses.
- Proportional hazards assumptions are often formally assessed in extension follow-up studies, and if violated, prespecified alternative approaches, such as time-varying effects, flexible hazard models, milestone analyses or restricted mean survival time, can be utilized.
- Cause-specific hazard models are preferred for estimating etiologic effects, while subdistribution models or cumulative incidence functions estimate absolute risk.
- Ensuring long-term data continuity in extension follow-up studies requires the integration of trial data with external sources (e.g. registries, electronic health records (EHR), insurance claims data), standardization of data definitions and coding across trial phases, and the establishment of secure data governance, transfer and data harmonization processes.
- Long-term follow-up is often planned during the initial trial design stage, including resources for study personnel, longitudinal data management, and participant engagement; cost-efficient approaches may include registry-based designs or EHR linkage.
Citation
Gaudino M, Sandner S, Bagiella E, Beckman JA, Belley-Cote EP, Flather M, Gregg AC, Hameed I, Pocock S, Redfors B, Rockhold F; on behalf of the Ameri¬can Heart Association Leadership Committee of the Council on Clinical Cardiology; Council on Cardiovascular and Stroke Nursing; and Council on Lifelong Congenital Heart Disease and Heart Health in the Young. Methodologic standards for follow-up extension in cardiovascular trials: a scientific statement from the American Heart Association. Circulation. Published online August 3, 2026. doi: 10.1161/CIR.0000000000001458