Clinical research is a dynamic field where quantification matters. Appropriate choices in outcomes, endpoints, and estimands enhance a study's rigor, transparency, and clinical relevance. At the DCRI, our leading biostatistical experts have joined to form O2E, the Program for Outcomes, Endpoints, and Estimands. The program aims to fundamentally enhance our impact on science and our global reputation for excellence in these three areas. O2E investigates outcomes that matter to clinicians, patients, and caretakers, endpoints optimized for study success, and transparent, purpose-driven estimands. We work closely with clinical researchers, sponsors, and our peers to understand their unique challenges and forge evidence-based solutions.
How do we enable O2E that matters?
- Transform traditional approaches
- Excel in methodology
- Educate the research community
The Three Pillars of O2E
Outcomes
Outcomes refer to measurements taken on subjects for evaluation. They can be clinical, biological, or related to a patient's well-being. If they are measured after a medical intervention or treatment, outcomes help us understand the impact of a treatment on patients' health and quality of life. They guide decision-making and inform healthcare practices. Defining their measurement can be straightforward (survival status, disease progression) or challenging (patient-reported outcomes).
DCRI faculty contributing O2E outcomes expertise: 42
Endpoints
Endpoints are specific outcomes or combinations of outcomes used to assess treatment efficacy and safety in clinical studies. They form the basis for statistical analysis, which determines whether a therapy successfully achieves its intended effect. Different measurements, such as composite, time-to-event, categorical, or continuous data, can be used to evaluate the treatment effect. Recently, O2E research has been focusing on new methods for the use of hierarchical endpoints in the design and analysis of clinical trials.
DCRI faculty contributing O2E endpoints expertise: 44
Estimands
Estimands have received substantial attention since the 2019 International Council for Harmonization (ICH) E9 Guidelines addressed intercurrent events in the estimand framework. This guidance challenges researchers to be precise about what is being estimated and how that is relevant to the scientific purpose. Estimands are defined more broadly, including attributes such as population characteristics, endpoint definition, handling of intercurrent events, and measure summaries.
DCRI faculty contributing O2E estimands framework expertise: 36
Sample Size and Power Calculator for Design with Hierarchical Endpoints
Using the O2E-developed calculator, determine your study's sample size or power based on hierarchical endpoints. Add multiple endpoints sequentially, adjust requirements and randomization ratio, and get a sense of the sample size with given power or a sense of power with given sample size. The calculated magnitude of the overall win measure and probability of ties are also displayed.
