Part 9 · Module 2
Randomised Controlled Trial Design
Design a randomised controlled trial that aligns eligibility, allocation, intervention delivery, outcomes, analysis and reporting.
Learning outcomes
After completing this lecture, you should be able to:
- Describe the participant pathway of a randomised controlled trial.
- Choose an appropriate randomisation unit and comparator.
- Align primary outcome, estimand and sample-size rationale.
- Plan intention-to-treat and supportive analyses.
- Apply current protocol and reporting guidance.
Overview
A randomised controlled trial assigns eligible participants or clusters to intervention strategies using a chance process and compares outcomes under prespecified follow-up. Randomisation provides a basis for unbiased treatment comparison when allocation is concealed and post-randomisation processes are protected.
The study question, design, participant pathway, measurements and analysis must be specified as one coherent system. A recognised design label is useful only when the operational details preserve the comparison or prediction that the study is intended to support.[1–3]
Why this matters
The credibility of an RCT is not created by the word randomised. Sequence generation, concealment, recruitment timing, intervention delivery, outcome ascertainment, missing data and analysis all contribute. A trial can be randomised yet answer an unclear question if the treatment strategies or intercurrent events are not defined.
Core concepts
Eligibility and randomisation share a time origin
Participants should meet eligibility and consent before allocation. Baseline measurements should precede knowledge of assignment. Follow-up starts at a common time so that early events and withdrawals are not differentially excluded.
Randomisation needs concealment
A valid random sequence is insufficient if recruiters can predict the next assignment. Central randomisation or secure systems protect concealment until irreversible enrolment.
The primary objective drives design
The trial should identify one primary comparison, outcome time and estimand. Multiple co-primary outcomes or time points increase multiplicity and ambiguity unless justified.
Analysis follows assignment
Intention-to-treat analysis preserves the randomised comparison by analysing participants according to assigned strategy. Per-protocol or as-treated analyses can be informative but are vulnerable to post-randomisation selection and require causal assumptions.
Registration and protocol transparency are core
Prospective registration, accessible protocol and statistical analysis plan reduce selective reporting. CONSORT 2025 and SPIRIT 2025 provide current reporting structures for results and protocols.
When to use this approach
Key RCT design choices
- Should individuals, clusters or time periods be randomised?
- What allocation ratio and stratification are justified?
- What is the clinically relevant comparator?
- Which primary outcome and time point answer the decision?
- How will intercurrent events and missing outcomes be handled?
Working through the key design decisions
Decision 1: Should individuals, clusters or time periods be randomised?. This point should be resolved before recruitment, extraction or assignment begins. For randomised controlled trial design, the protocol should state the operational rule, who applies it, when it is applied and what source documents demonstrate compliance. The investigators should also consider how a different choice would change the eligible population, comparison, timing, measurement or interpretation. A pilot review of realistic records or participant pathways can reveal ambiguities before they become irreversible exclusions or post hoc analytical choices. The decision should be reviewed across sites and subgroups because a rule that works in one clinic may operate differently where referral, resources or disease spectrum differ. Any permitted exception should be explicit, dated and reported with its frequency and likely consequence.
Decision 2: What allocation ratio and stratification are justified?. This point should be resolved before recruitment, extraction or assignment begins. For randomised controlled trial design, the protocol should state the operational rule, who applies it, when it is applied and what source documents demonstrate compliance. The investigators should also consider how a different choice would change the eligible population, comparison, timing, measurement or interpretation. A pilot review of realistic records or participant pathways can reveal ambiguities before they become irreversible exclusions or post hoc analytical choices. The decision should be reviewed across sites and subgroups because a rule that works in one clinic may operate differently where referral, resources or disease spectrum differ. Any permitted exception should be explicit, dated and reported with its frequency and likely consequence.
Decision 3: What is the clinically relevant comparator?. This point should be resolved before recruitment, extraction or assignment begins. For randomised controlled trial design, the protocol should state the operational rule, who applies it, when it is applied and what source documents demonstrate compliance. The investigators should also consider how a different choice would change the eligible population, comparison, timing, measurement or interpretation. A pilot review of realistic records or participant pathways can reveal ambiguities before they become irreversible exclusions or post hoc analytical choices. The decision should be reviewed across sites and subgroups because a rule that works in one clinic may operate differently where referral, resources or disease spectrum differ. Any permitted exception should be explicit, dated and reported with its frequency and likely consequence.
Decision 4: Which primary outcome and time point answer the decision?. This point should be resolved before recruitment, extraction or assignment begins. For randomised controlled trial design, the protocol should state the operational rule, who applies it, when it is applied and what source documents demonstrate compliance. The investigators should also consider how a different choice would change the eligible population, comparison, timing, measurement or interpretation. A pilot review of realistic records or participant pathways can reveal ambiguities before they become irreversible exclusions or post hoc analytical choices. The decision should be reviewed across sites and subgroups because a rule that works in one clinic may operate differently where referral, resources or disease spectrum differ. Any permitted exception should be explicit, dated and reported with its frequency and likely consequence.
Decision 5: How will intercurrent events and missing outcomes be handled?. This point should be resolved before recruitment, extraction or assignment begins. For randomised controlled trial design, the protocol should state the operational rule, who applies it, when it is applied and what source documents demonstrate compliance. The investigators should also consider how a different choice would change the eligible population, comparison, timing, measurement or interpretation. A pilot review of realistic records or participant pathways can reveal ambiguities before they become irreversible exclusions or post hoc analytical choices. The decision should be reviewed across sites and subgroups because a rule that works in one clinic may operate differently where referral, resources or disease spectrum differ. Any permitted exception should be explicit, dated and reported with its frequency and likely consequence.
The approach should be chosen because its structure answers the research question, not because the available dataset happens to resemble a familiar table. The investigator should state the target population, time origin, comparison, outcome and intended measure before collecting or extracting data.
Design and participant pathway
Randomised controlled trial
Core structure: Use a concealed chance process to assign eligible units to prespecified strategies and compare outcomes under concurrent follow-up.
Best suited for: Causal evaluation of feasible interventions where assignment is ethical and contamination manageable.
Principal threats: Inadequate concealment, post-randomisation exclusions, non-adherence, differential care, missing outcomes, selective reporting and multiplicity.
Parallel two-group trials are common, but the design may use unequal allocation, multiple arms, adaptive features or cluster assignment. Complexity should serve the question. Stratification or minimisation can improve balance on a few strong prognostic factors, but every feature must be implemented securely and incorporated into analysis.
Protocol specification
Decisions to prespecify
- Write the primary objective and estimand.
- Define eligibility and consent before allocation.
- Specify sequence generation and concealment separately.
- Describe interventions using a replication framework.
- Prespecify primary, secondary and harm outcomes.
- Plan analysis populations, missing data and multiplicity.
Protocol item 1: Write the primary objective and estimand. This item requires a reproducible definition rather than a broad intention. The study manual should identify the responsible role, data field or procedural record, permissible time window, quality check and response to deviations. Where judgement is unavoidable, assessors should receive examples and escalation rules, and a sample of decisions should be reviewed for consistency. The analysis dataset should retain enough provenance to distinguish planned exclusions, measurement failure, participant withdrawal and genuinely absent information.
Protocol item 2: Define eligibility and consent before allocation. This item requires a reproducible definition rather than a broad intention. The study manual should identify the responsible role, data field or procedural record, permissible time window, quality check and response to deviations. Where judgement is unavoidable, assessors should receive examples and escalation rules, and a sample of decisions should be reviewed for consistency. The analysis dataset should retain enough provenance to distinguish planned exclusions, measurement failure, participant withdrawal and genuinely absent information.
Protocol item 3: Specify sequence generation and concealment separately. This item requires a reproducible definition rather than a broad intention. The study manual should identify the responsible role, data field or procedural record, permissible time window, quality check and response to deviations. Where judgement is unavoidable, assessors should receive examples and escalation rules, and a sample of decisions should be reviewed for consistency. The analysis dataset should retain enough provenance to distinguish planned exclusions, measurement failure, participant withdrawal and genuinely absent information.
Protocol item 4: Describe interventions using a replication framework. This item requires a reproducible definition rather than a broad intention. The study manual should identify the responsible role, data field or procedural record, permissible time window, quality check and response to deviations. Where judgement is unavoidable, assessors should receive examples and escalation rules, and a sample of decisions should be reviewed for consistency. The analysis dataset should retain enough provenance to distinguish planned exclusions, measurement failure, participant withdrawal and genuinely absent information.
Protocol item 5: Prespecify primary, secondary and harm outcomes. This item requires a reproducible definition rather than a broad intention. The study manual should identify the responsible role, data field or procedural record, permissible time window, quality check and response to deviations. Where judgement is unavoidable, assessors should receive examples and escalation rules, and a sample of decisions should be reviewed for consistency. The analysis dataset should retain enough provenance to distinguish planned exclusions, measurement failure, participant withdrawal and genuinely absent information.
Protocol item 6: Plan analysis populations, missing data and multiplicity. This item requires a reproducible definition rather than a broad intention. The study manual should identify the responsible role, data field or procedural record, permissible time window, quality check and response to deviations. Where judgement is unavoidable, assessors should receive examples and escalation rules, and a sample of decisions should be reviewed for consistency. The analysis dataset should retain enough provenance to distinguish planned exclusions, measurement failure, participant withdrawal and genuinely absent information.
A protocol should make these decisions before outcome-dependent choices become tempting. It should also distinguish the design population from the analysis population, document exclusions after enrolment or sampling, and explain how missing or indeterminate observations will be handled. Any deviation should be dated, justified and retained in the study record.[2–4]
Measurements, timing and analysis interface
Primary outcomes should be valid, reliable and measured equally across groups. Patient-reported outcomes require scheduled administration and plans for proxy or translated versions. Objective record linkage can reduce assessor bias but must capture care outside the site. Baseline imbalances should be described without significance testing and handled according to prespecified adjustment.
The analysis should reflect how participants or clusters entered the study, how observations are related, and how the effect or performance measure was defined. Statistical adjustment cannot repair a comparison that was invalid at the design stage. Conversely, a sound design can still be undermined by an analysis that ignores matching, clustering, censoring, sampling fractions or prespecified thresholds.
Worked example
Constructing the primary trial question
The trial defines the population as women with gestational diabetes discharged from participating wards, the strategies as offer of navigation versus current care, the endpoint as verified glucose testing within six months and the population-level summary as the risk difference. Follow-up continues regardless of reminder engagement. Death before six months is incorporated using a prespecified estimand strategy, and missing linkage is addressed with sensitivity analyses.
The example should be interpreted in the population and time frame created by the design. The numerical result does not by itself establish causation, transportability or clinical usefulness; those conclusions require additional assumptions and evidence.
Appraising methodological quality
Questions for critical appraisal
- Randomisation: Was the sequence genuinely unpredictable?
- Concealment: Could enrolment decisions be influenced by upcoming allocation?
- Interventions: Were both strategies described and delivered as planned?
- Outcomes: Was follow-up complete and comparable?
- Reporting: Do registration, protocol, analysis plan and publication agree?
A reporting guideline can improve transparency, but it is not a substitute for sound design. Appraisal should ask whether the reported methods could generate an unbiased, clinically meaningful answer—not merely whether every checklist item has been mentioned.[3–6]
Suggested reporting
Reporting should identify the design in the title or abstract when appropriate, describe the source and flow of participants, define all key variables, report missing and indeterminate observations, and present estimates with uncertainty. The direction and denominator of every comparison must be explicit.
Ethical and governance considerations
Covariate adjustment in randomised trialsMMed extension
Prespecified adjustment for strong baseline prognostic factors can improve precision and address stratified randomisation. Variables measured after allocation should not be treated as ordinary baseline covariates. Adjusted and unadjusted estimates should target the same estimand and use appropriate standard errors.
Common mistakes
Knowledge check
Why analyse by assigned group?
Answer and reasoning
It preserves the comparison created by randomisation and estimates the effect of assignment under the stated estimand.
Can an RCT have unequal allocation?
Answer and reasoning
Yes, when justified by cost, safety, recruitment or information considerations, although equal allocation is often most statistically efficient.
Why register prospectively?
Answer and reasoning
To make planned outcomes and analyses visible before results are known and reduce selective reporting.
Does concealment occur after allocation?
Answer and reasoning
No. Allocation concealment protects the assignment process before and at enrolment; blinding concerns knowledge after assignment.
Key take-home messages
Key take-home messages
- Randomise only after eligibility and consent are secured.
- Separate sequence generation from concealment.
- Define one clear primary estimand and outcome.
- Preserve assignment in primary analysis.
- Use registration, protocol and reporting guidance as a connected transparency system.
References
- Hopewell S, Chan AW, Collins GS, et al. CONSORT 2025 statement: updated guideline for reporting randomised trials. BMJ. 2025;389:e081123. https://doi.org/10.1136/bmj-2024-081123
- Butcher NJ, Monsour A, Mew EJ, et al. SPIRIT 2025 statement: updated guideline for protocols of randomised trials. BMJ. 2025;389:e081477. https://doi.org/10.1136/bmj-2024-081477
- International Council for Harmonisation. ICH harmonised guideline E6(R3): Guideline for Good Clinical Practice. Geneva: ICH; 2025. https://database.ich.org/sites/default/files/ICH_E6%28R3%29_Step4_FinalGuideline_2025_0106.pdf
- International Council for Harmonisation. ICH E9(R1) addendum on estimands and sensitivity analysis in clinical trials. Geneva: ICH; 2019. https://database.ich.org/sites/default/files/E9-R1_Step4_Guideline_2019_1203.pdf
- Moher D, Hopewell S, Schulz KF, et al. CONSORT 2010 explanation and elaboration. BMJ. 2010;340:c869. https://doi.org/10.1136/bmj.c869
- Hoffmann TC, Glasziou PP, Boutron I, et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ. 2014;348:g1687. https://doi.org/10.1136/bmj.g1687
These materials support learning and research planning. They require adaptation to the study context and do not replace methodological, statistical, ethical, or regulatory consultation.