Write a controlled clinical study report using ICH E3

What should a controlled clinical study report contain?

Integrate the study’s clinical and statistical account: objectives, design, ethical conduct, actual participation and treatment, analyses, efficacy and safety results, limitations and supporting documents. Make every major conclusion traceable to its endpoint, population and output. Use E3 flexibly to communicate the evidence without omitting relevant information.

Before you begin

Controlled efficacy or safety studies organized within Module 5. A report’s location does not establish that its design is adequate or its evidence sufficient for approval.

What you will prepare: A coherent individual-study CSR with reconciled populations, prespecified analyses, safety interpretation and accessible companion evidence.

Fix the evidence set before drafting the conclusion

Collect the final protocol and amendments, statistical analysis plan and changes, trial conduct records, data-cutoff/lock information, validated outputs and the relevant medical review. Establish the claimed indication, control type and role of the study. M4E organizes efficacy/safety reports by design and control, with specific treatment of studies relevant to more than one indication; the study's phase label alone is not the placement rule.

Build a working map from each study objective to its endpoint, analysis population, statistical method and report result. Include negative and inconclusive findings. An individual CSR should integrate the clinical and statistical account, not simply concatenate separate departmental documents.

Establish report identity, ethics and study oversight

Reconcile the title page and report identifiers with the protocol, investigational product, sponsor, study dates and final report version. Identify the investigators and administrative responsibilities that explain how the study was organized. Obtain the controlled investigator/site list and relevant oversight documentation instead of copying names from an earlier study.

For the ethics account in E3 sections 5–6, inspect the actual IEC/IRB review records for the study and amendments, participant information and consent documents, and the documented process for obtaining consent. Describe the ethical conduct and applicable GCP basis from evidence reviewed by the responsible clinical/quality team. Do not generate a compliance statement solely because a template contains one.

If an approval, consent version or oversight record cannot be reconciled, identify the affected site, period or amendment and assign investigation to its owner before finalizing the statement. Keep relevant supporting documents available through the package and applicable appendix references. A writer can describe confirmed conduct; writing the statement cannot establish that the conduct occurred.

Explain the difference between the plan and what happened

Describe design choices and control selection, participant selection, treatment, endpoint assessment and statistical methods. Then account for participant disposition, important protocol deviations, exclusions and changes affecting interpretation. Identify when consequential analysis decisions were made relative to access to study data.

Define each analysis population and use those definitions consistently in tables and text. Present the primary analysis with its effect estimate and uncertainty, then distinguish supportive, sensitivity and exploratory analyses. Address missing data and multiplicity through the study-specific plan and actual analyses; do not repair an unfavorable prespecified analysis with an unlabeled favorable subset.

Write the safety account from the exposed population

Reconcile treatment exposure, adverse-event summaries, serious events, deaths, discontinuations and relevant laboratory or other observations. Preserve the event definitions, denominators and time windows used. Narratives should explain clinically important sequences with evidence and medical assessment; event counts alone cannot describe timing, treatment or outcome.

Fictional exercise: the synopsis reports the randomized count as the safety denominator, while safety tables use participants who received treatment. Trace both definitions and correct the mismatch through the author and analyst. A numerically consistent percentage is still misleading when its population is wrong.

Make the report easy to verify without claiming acceptance

Write a standalone synopsis with design, population, key findings and material limitations consistent with the main report. E3 allows useful adaptation; its synopsis examples are not universal page quotas. Provide the reviewer-needed appendices and cross-references through the agreed electronic package, not merely a statement that documents exist in the trial master file.

Before release, have reviewers trace the main conclusion to the primary result, population and underlying protocol/analysis decisions, and compare it with the clinical efficacy and safety summaries. A complete CSR is a report of the evidence, not a declaration that FDA has accepted the trial or the benefit-risk conclusion.

Build a conclusion-to-evidence map for the CSR review

Review the key conclusion before polishing the executive wording. This map exposes disagreements between the synopsis, main report and statistical outputs.

Synopsis exercise: the primary analysis is inconclusive while an exploratory subgroup is favorable. Keep their roles explicit in the synopsis and discussion. Do not promote the subgroup to the primary conclusion by omitting its exploratory status or the overall result. Have the statistician and clinical author agree the supported interpretation.

Reconcile the final CSR with the efficacy summary, safety summary and appendix/data package. A consistent sentence across documents is useful only when it remains faithful to the underlying evidence.

Frequently asked questions

Sources and revisions

Requirements, source recommendations and editorial preparation advice have different roles. Review the scope and revision of the source you use.

Related guides