In clinical trials, it is essential to clearly define the goal of the study and what exactly is being measured. To address this, the European Medicines Agency introduced the concept of estimands in their 2020 addendum on estimands and sensitivity analysis. This framework helps align the planning, design, conduct, analysis, and interpretation of clinical trials. An estimand provides a precise description of the treatment effect that corresponds to the clinical question posed by the trial.
What Is an Estimand?
An estimand summarizes what outcomes would occur in the same patients under different treatment conditions being compared. It provides clarity by specifying the population, treatment, variable (endpoint), and how intercurrent events — events that happen after treatment initiation — will be handled. Intercurrent events, such as treatment discontinuation or use of additional medication, must be accounted for as they can affect the interpretation or measurement of the treatment effect.
The estimand is complemented by the estimator, a method used to calculate the estimated value of the treatment effect based on clinical trial data, which ultimately results in the estimate, a numerical value representing the treatment effect.
Why Are Estimands Important?
The concept of estimands ensures that the treatment effect being measured is clearly defined and addresses the clinical question at hand. This approach avoids potential confusion caused by intercurrent events, which can obscure the interpretation of trial results if not properly accounted for. For example, in oncology trials, patients may switch treatments, and the way these events are managed in the trial design can affect the trial’s conclusions. By explicitly addressing these events within the estimand framework, the trial can provide a more accurate and relevant assessment of the treatment effect.
Attributes of an Estimand
To define an estimand, several key attributes must be specified:
- Treatment Condition: This includes both the treatment of interest and any alternative treatments for comparison. These could involve individual interventions or more complex regimens.
- Population: The patient population targeted by the trial, which may be the entire population or a specific subgroup based on characteristics such as baseline measurements or specific intercurrent events.
- Variable (Endpoint): The clinical outcome or measurement used to address the trial’s question. This can include whether a patient experienced an intercurrent event, and how that event is factored into the trial's outcome.
- Handling of Intercurrent Events: Strategies are needed to address intercurrent events. These may include treatment policy, hypothetical strategies, or focusing on treatment up to the point of the event. The choice of strategy affects how the trial data are interpreted.
Handling Intercurrent Events
Intercurrent events can significantly affect how the results of a clinical trial are interpreted. These events, such as treatment discontinuation, the introduction of additional medication, or even a patient’s death, occur after the initiation of treatment and can complicate the measurement of treatment effects. Addressing these events is essential when defining an estimand because they may either influence the interpretation of results or make it impossible to measure the planned outcomes. Several strategies can be used to manage intercurrent events, ensuring that the clinical question of interest remains clear.
- Treatment Policy Strategy: In this approach, all intercurrent events are considered part of the treatment process. For example, if a patient discontinues the treatment, the outcome is still measured as if the patient were continuing in the study. This strategy reflects a real-world approach, where treatment changes are part of normal clinical practice.
- Hypothetical Strategy: Here, the effect of the treatment is estimated by imagining a scenario in which the intercurrent event did not occur. For instance, if a patient discontinues treatment due to an adverse effect, the hypothetical strategy assumes the patient continued treatment, allowing the estimand to assess what would have happened under ideal conditions.
- While on Treatment Strategy: This strategy focuses on the period during which the patient remains on the initially assigned treatment. Once the patient experiences an intercurrent event (e.g., treatment discontinuation), data after that point are excluded from the analysis. The effect is measured only while the patient is under the treatment regimen.
- Principal Stratum Strategy: This approach focuses on a specific subgroup of patients defined by whether or not they experience an intercurrent event. For example, the estimand may target only patients who do not switch treatments during the trial, providing insight into treatment effects for this specific group.
The use of estimands ensures that the goals of a clinical trial are clearly defined and that the impact of treatment is measured accurately. By addressing intercurrent events and providing a structured approach to trial design, estimands help align the planning, analysis, and interpretation phases of clinical trials, ultimately leading to better clarity in the assessment of treatment effects.
European Medicines Agency. (2020). ICH E9 (R1) addendum on estimands and sensitivity analysis in clinical trials to the guideline on statistical principles for clinical trials. https://www.ema.europa.eu/en/documents/scientific-guideline/ich-e9-r1-addendum-estimands-and-sensitivity-analysis-clinical-trials-guideline-statistical-principles-clinical-trials-step-5_en.pdf