Clinical trial design begins well before the protocol is written. Between a promising scientific idea and a protocol robust enough to be submitted to the competent authorities and ethics committees, many decisions need to be made: what question are we really trying to answer? In which patients? With which comparator? What primary endpoint? How many participants will be needed?
These choices are closely interconnected. A lack of precision at the start of the project can have major consequences for study feasibility, the quality of the data generated and, ultimately, the trial’s ability to answer its objective.
Here are 7 key steps for turning a research idea into a clinical trial protocol.
1. Clinical Trial Design: Start with a Precise Scientific Question
The first step in clinical trial design is to turn a general idea into a clear research question.
“Assess the efficacy of a new treatment” is not yet a sufficiently precise objective. The effect to be demonstrated, the target population and the therapeutic strategy used for comparison must all be clearly defined.
This assessment should take into account, in particular:
- the available preclinical data;
- the results of any previous clinical studies;
- the safety data already available;
- current knowledge of the disease and its progression;
- currently available treatments;
- the remaining unmet medical need.
ICH E8(R1) notably emphasizes the need to clearly define study objectives and to identify, from the design stage onwards, the factors that are critical to trial quality.ICH E8(R1)
The more precise the scientific question, the more coherent the subsequent methodological choices can be.
2. Clinical Trial Design: define the Primary and Secondary Objectives
For clinical trial design, once the research question has been established, it must be translated into measurable objectives.
The primary objective addresses the key question the study is intended to answer. Secondary objectives can explore additional dimensions such as complementary efficacy outcomes, tolerability, quality of life, pharmacokinetics or different patient subgroups.
This hierarchy is essential.
Multiplying objectives or endpoints without clear justification can complicate interpretation of the results and unnecessarily increase the burden of the study.
In confirmatory trials, the concept of the estimand, developed in ICH E9(R1), should also be considered. The aim is to specify exactly which treatment effect the study seeks to estimate, particularly by anticipating events that may occur after randomization, such as treatment discontinuation, use of rescue medication or a change in treatment.ICH E9(R1).
The estimand therefore helps maintain consistency between the clinical objective, the data collected and the statistical analysis.
3. Clinical Trial Design: define the Study Population
To whom should the trial results apply?
Defining the population is one of the major decisions in clinical trial design.
For clinical trial design, inclusion and exclusion criteria should enable recruitment of a population that is sufficiently homogeneous to answer the scientific question, without becoming so restrictive that recruitment is impractical or the results are no longer representative of future patients.
Several parameters may need to be considered:
- age;
- diagnosis and disease stage;
- severity;
- previous treatments;
- comorbidities;
- biological parameters;
- specific risk factors.
A balance must therefore be found between scientific validity, participant safety, representativeness and operational feasibility.
This consideration is particularly important for vulnerable populations or groups that are underrepresented in clinical research.
4. Choose the Trial Design and Comparator
The design should be chosen to fit the research question, not the other way around.
Randomized or non-randomized? Parallel-group or crossover? Open-label, single-blind or double-blind? Active comparator, placebo or standard of care?
Each choice must be justified.
In many confirmatory trials, randomization helps limit selection bias and balance prognostic factors between groups. Blinding can help reduce certain biases related to assessment or patient management.
Other designs may be appropriate depending on the context, including adaptive trials, basket or umbrella studies, platform trials, or designs incorporating certain decentralized elements.
ICH E6(R3) explicitly requires the protocol to describe the type and design of the trial, as well as the measures intended to minimize bias.ICH E6(R3)
5. Define the Primary Endpoint
The primary endpoint is the main measure used to determine whether the trial objective has been met.
It should be selected very early in the design process.
It should be:
- clinically relevant;
- sufficiently sensitive to detect the expected effect;
- reliably measurable;
- appropriate for the duration of the study;
- consistent with the scientific question.
Depending on the disease, it may be a clinical event, a biological parameter, a validated score, a time-to-event measure or a composite endpoint.
Secondary endpoints complement this assessment, but they should not distract from the primary question.
An inappropriate choice of primary endpoint can compromise the interpretation of an otherwise well-conducted trial.
6. Build the Statistical Strategy and Calculate the Sample Size
Statistics do not come into play only after the data have been collected.
They are an integral part of trial design.
The required number of participants depends in particular on:
- the expected effect;
- the variability of the endpoint;
- the accepted statistical error rate;
- the target statistical power;
- the anticipated rate of study withdrawals or missing data.
An insufficient sample size may prevent a genuine effect from being detected. Conversely, recruiting far more participants than necessary may unnecessarily expose more people to research and substantially increase the cost of the study.
ICH E8(R1) notably provides that the methods used to analyze primary and secondary endpoints, any interim analyses and the justification for the sample size should be defined in the protocol.
The statistical strategy should therefore be developed alongside the clinical design rather than added at the end of the process.
7. Turn Scientific Choices into a Feasible Protocol
A study can be scientifically excellent on paper yet impossible to conduct in practice.
Before finalizing the protocol, the project must be tested against the realities faced by investigational sites and patients.
Is the number of visits acceptable? Are the planned assessments available at the sites? Will the inclusion criteria realistically allow recruitment? Is the duration of participation compatible with the disease? Is data collection proportionate to the study objectives?
This step also helps identify critical-to-quality factors and risks that could affect participant safety or the reliability of the results.
The final protocol brings together the scientific, medical, statistical and operational choices: rationale, objectives, design, population, treatments, endpoints, visit schedule, safety monitoring, statistical methods and overall study organization.
In the European Union, the protocol forms part of Part I of the clinical trial application submitted through CTIS, in accordance with Regulation (EU) No 536/2014.
A Good Protocol Is Designed Before It Is Written
Clinical trial design is therefore, above all, an exercise in consistency.
The scientific question determines the objective; the objective determines the population, design and endpoints; these, in turn, shape the statistical strategy and the operational organization of the study.
Current guidelines, particularly ICH E8(R1), ICH E9(R1) and ICH E6(R3), reinforce this approach by encouraging quality to be built into the study from the outset (quality by design).
A robust protocol is therefore not simply a well-written regulatory document. It is the outcome of a series of decisions made early enough to ensure that the trial is scientifically relevant, ethically acceptable, statistically robust and operationally feasible.
For sponsors, clinical research teams and professionals involved in drug development, mastering these steps is essential to ensure a robust transition from scientific idea to clinical trial.
