Blog Post
Back to Blog

Clinical Study Report (CSR): What It Looks Like and What Goes In It

An annotated sample synopsis, the ICH E3 section map, public disclosure and redaction rules, and the writing workflow for sponsors.
Get Matched Free
Insights
Colby Flood
LinkedInx

ICH E3 has set the structure of the clinical study report since it reached its final version on 30 November 1995 [1]. After database lock the CSR is on the critical path to submission, yet most guides stop at a list of headings. Because disclosure rules mean a CSR may be published, it is best drafted for a stranger, and the annotated sample synopsis, section map, redaction rules and workflow below show how. Refero does not run recruitment campaigns. We vet and match the clinical research marketing agencies that do, so what follows is written for the buyer, from what those vendors need and see.

What a Clinical Study Report Is and When a Trial Needs One

A clinical study report (CSR) is the integrated full report of one study's design, conduct, analysis and results, written to the structure in ICH E3 and accepted by the regulators of the ICH regions [1]. The FDA publishes the guideline as its own guidance [2]. A full report is expected for studies that support a marketing application, while ICH E3 lets uncontrolled studies, aborted studies and studies not designed to establish efficacy use an abbreviated report with summarised data or some sections deleted, although a controlled safety study should be reported in full [1].

The CSR is easy to confuse with two neighbours. The protocol is the plan written before the study, and it returns as an appendix of the CSR, while a journal article is a short, selective account of the same study. The CSR sits between them as the complete record of what happened, detailed enough that an authority can replicate the critical analyses [1].

What a Clinical Study Report Looks Like: Section Map and Annotated Sample

A finished CSR is a long document with a short front page: the synopsis comes first, and ICH E3 usually limits it to three pages that must carry numerical data on results, not just text or p-values [1]. The sample below fills in the ICH E3 synopsis fields for a fictional phase 2 study and notes what a reviewer checks in each. The values are invented.

Synopsis fieldFictional entryWhat a reviewer checks
Title of studyPhase 2 Examplumab study, plaque psoriasisMatches protocol and registry title
Investigators and centres14 centres in 4 countriesAgrees with appendix investigator list
ObjectivesPrimary: skin clearance, week 16Same wording as protocol and SAP
DesignRandomized 1:1, double-blind, 16 weeksDeviations from protocol described
Population117 adults analysed, 120 plannedCounts reconcile with disposition table
EndpointsClear or almost clear skin, week 16Prespecified, none added after unblinding
ResultsClearance 41% against 12% on placeboNumbers as well as p-values
ConclusionPrimary objective metNo claim beyond efficacy and safety sections

An invented synopsis for illustration; the fields follow the ICH E3 synopsis format [1].

Behind the synopsis, the report follows the 16 numbered sections of ICH E3, and the table below maps the ones a reviewer meets most, with who typically drafts each [1].

Three terms need defining because the section map uses them. Tables, listings and figures (TLFs) are the statistical outputs the biostatistician produces, and the report places summary tables in the text and the rest in section 14. Patient narratives are short accounts of each death, each other serious adverse event and the adverse events of special interest, and ICH E3 lets them sit in the text or in section 14.3.3 depending on their number [1]. The appendices in section 16 hold the protocol and amendments, a sample case report form, investigator information, statistical documentation and patient data listings [1].

ICH E3 sectionWhat it holdsTypical drafter
1 to 3: Title page, synopsis, contentsSignatory, GCP statement, synopsisMedical writer
5: EthicsIEC or IRB review, consentRegulatory and clinical lead
6: Investigators and administrative structureWho ran the study, whereMedical writer
7 to 9: Introduction, objectives, investigational planDesign, population, treatments, endpoints, statisticsWriter with biostatistician
10: Study patientsDisposition, protocol deviationsWriter, biostatistician
11: Efficacy evaluationData sets, results, statistical issuesBiostatistician; clinical lead reviews
12: Safety evaluationExposure, adverse events, deaths, labsWriter; safety physician reviews
13: Discussion and conclusionsOverall benefit-risk readingClinical lead and writer
14 to 16: TLFs, references, appendicesOutputs outside text, narratives, protocol, listingsBiostatistics; writer assembles

Section numbers and contents follow ICH E3 [1]; the drafter column is a typical split and varies by sponsor.

ICH E3 is the structure regulators expect, and no single mandated form exists beyond it. A common template is available as an alternative to building the document from the guideline: TransCelerate, a non-profit industry consortium, publishes a common CSR template [3].

Clinical Study Reports at a Glance

A structure written in the 1990s now meets disclosure programmes that started two decades or more after it.

  • ICH E3 reached its final version on 30 November 1995 [1].
  • ICH E3 usually limits the synopsis to 3 pages, with numerical results [1].
  • The EU's CTIS went live with a searchable public website on 31 January 2022 [4].
  • EMA began publishing clinical data under Policy 0070 in October 2016 [5].
  • FDA ran a clinical data summary pilot from 2018 that ended in 2020 [6].

Your Clinical Study Report Will Probably Be Public: Plan Redaction Early

A CSR is likely to be published in redacted form when the trial ran in the EU or the report sits inside a marketing application to EMA, so the safest drafting assumption is that a stranger will read it. For trials under the EU Clinical Trials Regulation, the public CTIS website has carried trial information since it went live on 31 January 2022 [4], and revised transparency rules applied from 18 June 2024 [4]. The regulation itself provides for a clinical study report to be made public when a trial supports a marketing authorisation [7]. EMA Policy 0070 separately covers clinical reports submitted for centrally authorised medicines, including individual study reports [5]. In the US, FDA piloted releasing portions of CSRs, and that programme is reported to have ended in 2020 [6][8].

Redaction removes two categories of content. Personal data is anything that identifies a person, such as a patient identifier or date of birth in a narrative, and EMA marks it with light blue boxes labelled PPD. Commercially confidential information (CCI) is content a company can justify as confidential, such as a proprietary manufacturing detail in an investigational product section, and EMA marks it with black boxes labelled CCI [5]. Because ICH E3 asks each narrative to include a patient identifier, age and sex, and medical history [1], narratives are where personal data concentrates, so writing them in a form that can be pseudonymised from the start costs far less than reopening a signed report. Keeping confidential material in clearly separable places saves the same rework. This is not legal advice; confirm with regulatory counsel and the IRB of record.

Readers who only want to read existing CSRs have three routes. EMA runs a clinical data website where released reports can be viewed with an account [5], the CTIS public search lists trials and their documents [4], and some sponsors keep libraries of CSR synopses, although most US reports are not public beyond FDA's pilot [6].

How to Write a Clinical Study Report: Workflow, Roles and Timeline

The writing starts from a locked database and a finalised statistical analysis plan, and it moves through TLF delivery, a first draft, review rounds, QC and sign-off, after which the report is filed in eCTD Module 5 of the submission.

The biostatistician and writer agree TLF shells before lock, so tables arrive in the layout the report needs. Once the database locks, the biostatistician runs the analyses and produces the TLFs, and the writer drafts around them. The first draft goes to the clinical lead, safety and regulatory reviewers, and each review round returns comments that the writer reconciles. A QC pass then checks every number in the text against the source tables, because the same result appears in the synopsis, the text and the tables, and all three must match. After sign-off by the sponsor's responsible person, the report is placed in the dossier.

Each role owns a stage. The medical writer leads the document, the biostatistician owns the analyses and the statistical sections, the clinical lead owns the interpretation, the safety physician reviews section 12 and the narratives, regulatory owns submission fit, and QA checks the process. For a study with remote or home-visit data, the writer also needs to know how those data reach the TLFs, and the guide to how decentralized trials work covers the designs involved.

No regulator or journal source gives a standard duration, so plan the CSR backward from the submission or disclosure date and treat any figure as the sponsor's own estimate. Late TLFs, extra review rounds and a large number of narratives typically stretch the timeline.

Two recent changes sit alongside the CSR: ICH E6(R3), adopted on 6 January 2025, is the revised ICH good clinical practice guideline, the usual reference for the GCP statement that ICH E3 asks for on the title page [9], and the EU Clinical Trials Regulation also requires a lay summary of results [7].

How to Choose and Brief a Clinical Study Report Writing Partner

The sponsor owns the report even when a vendor drafts it, so the sponsor's brief and checks decide quality more than the vendor's pitch. For wider outsourcing choices, the outsourced clinical trials guide shows what stays with the sponsor, and the recruitment benchmarks and strategies hub covers the enrollment work that comes before any report exists.

Ask for a sample synopsis and a redacted CSR before you commit

Use this when comparing medical writing vendors or a CRO's writing team. Ask for a synopsis and a redacted report they have produced, and check the synopsis against the ICH E3 synopsis fields and the report against the section map, looking for missing sections or a results field with only p-values.

Put the template and disclosure plan in the brief before database lock

Use this before lock. Specify ICH E3 or the TransCelerate template, whether CTIS or Policy 0070 will apply, and the redaction approach, including how narratives are written and where confidential material is kept. The TLF shells and narrative format are set before lock, so a vendor briefed later has to rework both.

Check that text, tables and figures agree at the QC round

Use this at QC, before sign-off. Ask for a source-to-text check of every number in the synopsis and body. Some vendors offer AI drafting tools for first drafts; such tools can speed a draft but cannot replace the numerical check, so ask who verifies the output.

If you would rather not run that evaluation alone, Refero screens healthcare marketing and patient recruitment agencies against five published criteria and introduces up to three that fit your brief. It is free for buyers, and there is no obligation to hire anyone we introduce. Tell us what you need.

Frequently Asked Questions

Are clinical study reports publicly available?

Many are, in redacted form. Trials under the EU Clinical Trials Regulation publish through CTIS [4], EMA publishes clinical reports from marketing applications under Policy 0070 [5], and some sponsors keep synopsis libraries. Most US CSRs are not public beyond FDA's pilot [6].

Who writes a clinical study report?

Medical writers lead, with biostatisticians, clinical leads, safety and regulatory staff reviewing their sections. Sponsors often outsource the drafting to a CRO or medical writing vendor and keep sign-off, since the sponsor's responsible person signs the report [1].

How long is a clinical study report?

It varies with design and size. The body and the appendices make up the report, and the appendices hold the protocol, sample case report form and patient data listings [1], so large late-phase reports with full listings run very long, and no standard page count applies.

What is the difference between a clinical study report and a protocol?

The protocol is the plan written before the study. The CSR reports what happened and the results after the study ends, and the protocol and its amendments become an appendix of the CSR [1].

Is there a clinical study report template?

ICH E3 is the structure regulators expect, and no single form is mandated [1]. TransCelerate, a non-profit industry consortium, publishes a common CSR template [3].

Sources

  1. International Council for Harmonisation. E3 Structure and Content of Clinical Study Reports, Step 4, 1995. Link
  2. US Food and Drug Administration. E3 Structure and Content of Clinical Study Reports, Guidance for Industry, 1996. Link
  3. TransCelerate BioPharma. Clinical content and reuse assets: common clinical study report template, undated. Link
  4. European Medicines Agency. Clinical Trials Information System (CTIS), public website and transparency rules, 2024. Link
  5. European Medicines Agency. Clinical data publication (Policy 0070), 2025. Link
  6. Clinical Study Reports: a systematic review with thematic synthesis, Part 1. History, contents and structure, definitions, and terminology. PMC, 2025. Link
  7. European Union. Regulation (EU) No 536/2014 on clinical trials on medicinal products for human use, Article 37, 2014. Link
  8. US Food and Drug Administration. FDA continues to support transparency and collaboration in drug approval process as the clinical data summary pilot concludes, 2020 (archived copy). Link
  9. International Council for Harmonisation. ICH E6(R3) Guideline for Good Clinical Practice, Step 4, 2025. Link
Written by
Colby Flood
Founder and Editor, Refero
HIPAA Awareness for Business Associates certified (valid to September 2028). Registered submitter on WCG eReview Manager and Advarra CIRBI.
Colby Flood is the founder and editor of Refero. He has worked in healthcare and performance marketing since 2019, and set Refero up as a vetted index after watching healthcare buyers pick agencies from ad placements and logo walls. He writes and edits everything published here against the standards set out in the editorial policy.

More from Refero

Get Matched
Thinkwise consulting website template

Looking for a healthcare marketing agency?

Tell us what you need. We screen the vetted network and introduce up to three agencies that fit your brief, free.