Clinical Study Report (CSR): Writing in Accordance with ICH E3

The Clinical Study Report (CSR) is the reference document that provides a comprehensive overview of the methodology, efficacy, and safety results of a clinical trial. Its structure is defined by the international ICH E3 guideline. It is the document that regulatory authorities review when evaluating a drug, and it often serves as the basis for a marketing authorization application. At Statitec, a biometrics company within the MultiHealth Group, we assist sponsors in drafting the report, from the statistical tables through to the signed version.

What is a clinical study report?

The CSR is a comprehensive report: it brings together in a single document the context of the study, what was planned, what was actually done, and what was observed. It is not limited to positive results. It must describe all patients enrolled, deviations from the protocol, and all safety events.

According to ICH E3, the CSR is primarily intended as a report for regulatory authorities (ANSM, EMA, FDA). It is also used by ethics committees, investigators, and the sponsor’s internal teams. In the European Union, Regulation (EU) No. 536/2014 requires that a summary of the results be published on the CTIS portal within one year of the trial’s completion; this deadline is reduced to six months for certain pediatric trials. When the trial is intended to support a marketing authorization application, the CSR must also be submitted within 30 days of the marketing authorization being granted, the conclusion of the procedure, or the applicant’s withdrawal of the application.

The ICH E3 guideline establishes the structure of the CSR

Published in 1995 and supplemented by a Q&A document in 2012, ICH E3 outlines a 16-section framework. It is recognized by European, U.S., and Japanese authorities, which allows for the submission of a single report across multiple regions. The table below summarizes the main sections.

ICH Section E3	Expected Content
1–4. Title page, synopsis, table of contents, abbreviations	Study identification, abstract of no more than 3 pages covering the objectives, methodology, results, and conclusions
5 and 6. Ethics and Organization	Ethics Committee Opinion, Consent, List of Investigators and Service Providers
7–9. Introduction, Objectives, Course Outline	Rational, evaluation criteria, study design, randomization, sample size calculation, protocol amendments
10. Study Population	Patient flow, study dropouts, deviations from the protocol, baseline characteristics
11. Effectiveness	Analysis populations, results for primary and secondary endpoints, sensitivity analyses
12. Safety	Exposure, adverse events, serious events, deaths, biological data
13. Discussion and Conclusions	Interpretation of Results, Limitations, Risk-Benefit Ratio
14–16. Tables, References, Appendices	TLF, protocol and amendments, CRF template, patient lists, signatures

ICH E3 is a guide, not a rigid framework. The text provides for adapting the protocol to the nature of the study, provided that the choices are justified. The synopsis deserves special attention: it is often the only section that a busy reviewer reads in its entirety.

The CSR is based on the statistical analysis plan and the TLFs

A CSR is not written from scratch. It is derived directly from the protocol and the statistical analysis plan (SAP), which are finalized before the data freeze. Once the database has been locked by the data management team, the biostatisticians produce the TLFs (tables, listings, figures) that form the basis for sections 10 through 12.

That is why we have the medical writer and the biostatistician work together from the very beginning of the SAP drafting process. Table templates are validated in advance, which prevents the last-minute discovery that a key result is not presented as required by the text.

The CSR must also be consistent with the investigator’s brochure, which is updated based on its findings, and its discussion is often based on a systematic review of the literature.

The CSR drafting process consists of 5 steps

Preparation. Kickoff meeting, selection of the document template, review schedule, and list of reviewers. The descriptive sections (ethics, organization, study plan) can be drafted before the baseline is finalized, based on the protocol.
Integration of Results. Upon receipt of the TLFs, the writer drafts the efficacy and safety sections and updates the synopsis.
Multidisciplinary review. A physician, biostatistician, project manager, and representatives from pharmacovigilance and regulatory affairs review the first draft. Comments are consolidated in a single round to save time.
Quality control. Line-by-line verification of consistency between the text, tables, and code listings, followed by editorial proofreading. If a first draft was generated using agenerative AI tool, this verification becomes essential.
Approval. The final version is signed by the sponsor’s medical director and, when required by the regulatory authority, by the principal investigator or coordinating investigator, and is then included in the regulatory dossier.

This process builds on the quality framework described in ICH E6(R3): a reliable report begins with a well-designed protocol, as we explain in our article on the master protocol.

Common mistakes that delay validation

Inconsistencies between the text and the tables: a percentage rounded differently or a number of people that doesn’t match is enough to prompt a question from the evaluator.
Deviations from the protocol that were not discussed: their potential impact on the results must be explained, not just listed.
Incomplete safety section: Serious adverse events must be described individually and be consistent with the pharmacovigilance data.
Downplaying negative results: A trial that fails to meet its primary endpoint must state this clearly. The authorities always notice it.
Outdated synopsis: It is often written first and then overlooked during the final revisions.

A condensed CSR and a layman's summary address different needs

Not all trials require a full CSR. For a study that was terminated early or is not essential to the dossier, an abridged CSR may suffice, provided that all safety data are retained. The European regulation also requires a summary of the results written for the general public—one that a patient without medical training can understand. This is a writing task in its own right, requiring the results to be conveyed accurately without distorting them.

Entrust MultiHealth Group with drafting your CSR

Medical writers, biostatisticians, and data managers work together within the MultiHealth Group as part of a single organization, which streamlines the back-and-forth between data and text. Learn more about our expertise in medical and scientific writing, or contact our teams to discuss your project.

Frequently Asked Questions About the Clinical Study Report

How long does it take to write a clinical study report?

It generally takes 6 to 12 weeks from the database freeze to the final, signed version of a CSR. This timeframe depends on the complexity of the study, the number of review cycles, and the availability of the TLFs. Preparing the descriptive sections before the database freeze can significantly reduce this timeframe.

Who signs the clinical study report?

According to ICH E3, the CSR must bear the signature of the sponsor’s medical director. The signature of the principal investigator or coordinating investigator is required when mandated by the competent authority. The signatures are included in the appendices to the report. The sponsor remains responsible for the content submitted to the authorities.

Is a CSR required for a trial that was terminated prematurely?

Yes. A trial that is terminated before completion must be documented in a report—at a minimum, in the form of an abbreviated CSR—that explains the reasons for termination and presents all safety data collected. A summary of the results is also required in CTIS for trials subject to the European regulation.

What is the difference between a CSR and a scientific publication?

The CSR is a comprehensive regulatory document—often several hundred pages long, including its appendices—intended for regulatory authorities. A scientific publication is a short article intended for the medical community that highlights the main findings. The publication is based on the CSR but does not replace it.

Is the clinical trial report made public?

In Europe, the summary of results and the lay summary are published in CTIS. The CSR for trials intended to support a marketing authorization application is also submitted there and made available after personal data and confidential commercial information have been removed. The EMA also publishes certain clinical data as part of its transparency policy.

What is the equivalent of CSR for a medical device?

For clinical investigations of medical devices, the ISO 14155:2026 standard requires a clinical investigation report and describes its expected content. Its structure is similar to that of the CSR, with particular emphasis on the device’s performance and incidents related to its use.

Need assistance?

Do you need to draft the clinical study report for your trial? Our Statitec experts will guide you through the process, from the preparation of the TLFs to the final signed version: drafting in accordance with ICH E3, ensuring consistency between the text and tables, and providing a synopsis and summary of the results for the CTIS.

Contact our experts