Clinical Study Report Writing for Trial Sponsors
Pubrica writes clinical study reports to ICH E3 for trial sponsors. Our regulatory medical writers and biostatisticians prepare full, abbreviated and synoptic reports from the protocol, the statistical analysis plan and the locked dataset rather than from top-line tables. Each report states what the trial set out to establish, what was done, what was found and where the study departed from the plan, in the section order that reviewers use to navigate it.
- Written from the locked dataset and the pre-specified analysis plan rather than from a summary presentation, which can lead to statements the data do not support
- Every deviation from the SAP listed in section 9.8, dated and explained, because reviewers examine unexplained reanalyses closely
- Patient narratives for deaths, serious adverse events and discontinuations, written to a consistent template so that the full set reads as one document
- The estimand carried through from the SAP under ICH E9(R1), so the efficacy section answers the question the trial was designed to answer
- Structured for reuse in Module 2.5 and 2.7 and in integrated summaries, because content from the CSR is reused in several later documents
Full, abbreviated or synoptic CSRIncluded
Patient narrativesIncluded
Module 2.5 and 2.7 contentIncluded
Integrated summaries (ISS/ISE)Quoted with the statistical analysis
Database lock and custodyYour data management team
Sponsor sign-off and submissionYour responsibility
Locked data → analysed → drafted → reviewed → report
- Written from the locked dataset
- Traceable to protocol and SAP
- Deviations listed and explained
- Narratives written to one template
- Structured for Module 2.5 and 2.7
Service Scope
Six Deliverables Included in Every Clinical Study Report
Each report is based on one protocol, one analysis plan and one locked dataset, and what was pre-specified determines what the report can conclude.
From the structure checkReviewers navigate a CSR by section number rather than reading it from start to finish, so reorganising the structure for narrative flow costs the reviewer time.
From the efficacy sectionThe treatment group showed a significant improvement in the primary endpoint.Under the treatment policy estimand, with discontinuation and rescue medication handled as pre-specified: 12.4% versus 16.1% (difference 3.7 points, 95% CI 0.9 to 6.5; p=0.01). The same result, answering the question the trial was designed around.
From section 9.8Minor changes were made to the analysis.Three changes, each dated: the covariate added on 14 March, before unblinding, per SAP amendment 2; the subgroup analysis added on 2 August, after unblinding, and reported as exploratory; the sensitivity analysis dropped, with the reason.
From the narrative setSubject identifier, demographics, medical history, study drug exposure, the event with onset and outcome dates, action taken, concomitant medications and the investigator’s causality assessment. Every narrative, same order, every time.
From the safety sectionAdverse events were generally mild and well tolerated.84.2% of the treatment group and 79.6% of controls reported at least one adverse event; the difference was driven by nausea (18.1% versus 9.4%). Eleven subjects discontinued for an adverse event. Tolerability conclusions are presented in the discussion rather than the results.
From the handoverEfficacy and safety sections written so the Module 2.7 summary is an extraction rather than a rewrite, and so a CONSORT 2025 manuscript can be built from the same source without reconciling two versions of the numbers.
All samples are illustrative and anonymised. Hover over a sample to replay it.
Our Services
Clinical Study Report Documents We Prepare
Most sponsors need three or four of these documents from one trial. When ordered together, they are based on a single review of the protocol and the data.
Narrative count affects the quote. Report cost can be estimated reliably from the phase and report type, but narrative count varies widely and is the main source of estimate error. If you provide the number of deaths, serious adverse events and discontinuations at the quote stage, the quote will remain accurate.
Who We Serve
Trial Sponsors We Support
Templates, volumes and destinations vary between sponsors. For every sponsor, the report is written from the locked data and the pre-specified plan.
Compliance and Standards
Regulatory Guidelines and Standards for Clinical Study Reports
Each report is prepared against these sources. Where your sponsor template or regulatory agency has stated a position, we follow it.
Structure and content of clinical study reports (with the 2012 Q&A (R1))
Defines the report structure. E3 has not been revised since 1995, which is correct rather than out of date, and the 2012 questions and answers contain several important practical positions.
Good Clinical Practice (E6(R3), not E6(R2))
Governs how the data in the report were collected, recorded and reported, and sets the source-data expectations behind every number.
Statistical principles with the estimands addendum (E9(R1), 2019)
Defines the estimand framework. The efficacy section must answer the estimand specified in the SAP.
Structured protocol (Step 4, November 2025)
The protocol the report is written against. Where a trial protocol follows M11, the CSR maps directly to it; where it does not, we map the two documents explicitly.
Clinical safety data management (definitions for reporting)
Provides the definitions of serious, unexpected and suspected that the safety section and the narratives must apply consistently.
TransCelerate CSR template (2024 release)
The common industry template adopted by many sponsors. We write to your template if you have one, and to this template if you do not.
Quality management (cert. H3CR9TCU20251212INDQM10N)
Governs how the report is drafted, reviewed, version-controlled and released across review cycles.
Responsible Use of AI
Responsible Use of AI in Regulatory Writing
Regulatory reviewers trace each decision back to evidence and to a responsible individual. Our specialists therefore treat source traceability as more important than speed, and every statement in a document must be supported by a traceable source.
Why We Do Not Publish AI Accuracy Figures
Accuracy figures for AI tools are usually reported by the companies that sell them.
For this reason, Pubrica does not quote an accuracy figure of its own, as it would carry the same conflict of interest. We tell you which tool was used on your project, at which stage, and what a specialist changed afterwards.
Methodology
How Our Clinical Study Report Writing Process Works
Before working with the tables, our writers read the protocol, the SAP and all amendments, so that the report describes the trial and not only its outputs.
Source reading and gap check
We read the protocol and amendments, the SAP and its amendments, the TLFs or their shells and the deviation log before drafting, and raise any gap between what was planned and what exists at the start.
Report shell and content plan
We agree the report structure against ICH E3 and your template, confirm the TLF mapping and fix the narrative scope and count, so that the workload is clear from the start.
Drafting
Methods, results, safety and discussion are written from the locked outputs, with the estimand carried through and every departure from the plan recorded as it is identified.
Narratives
Narratives for deaths, serious adverse events and discontinuations are written to one template and reconciled against the safety database and listings, so that narratives and tables agree.
Review cycles and finalisation
We consolidate your medical, statistical and regulatory comments into a single set per cycle, resolve them and finalise the report with indexed appendices.
Request a Free Draft of CSR Section 9.8
We draft the changes-in-conduct section, covering every departure from the protocol and the analysis plan with dates and explanations, from your documents before the rest of the report is written. Reviewers use this section to judge how closely to read the others, and it is a frequent source of errors in first drafts. The draft is free and takes five business days.
Please send the SAP amendments as well as the SAP. A change made before unblinding is interpreted very differently from one made after it, and the dates establish which applies.
Our Experts
Our Clinical Study Report Writing Team
The same named team of regulatory specialists works on your project from scoping to delivery, and your project coordinator introduces each member before work begins.
Writes the document against the governing template and regulatory guidance, keeps every statement traceable to its source data, and maintains the version history and comment log through each review round.
Reviews the clinical and scientific content for accuracy and confirms that each conclusion is supported by the data. The expert is matched to your therapeutic or device area.
Checks that statistical methods, tables and conclusions are consistent with the analysis plan, and reviews how the data are presented before the document reaches regulatory review.
Performs an independent check of the final document against source data, the template and the applicable regulatory checklist. Findings are logged and resolved before delivery.
Sample Work
Sample CSR Section 9.8: First Draft and Final Version
Step through the changes one at a time. The same three protocol changes are written in two ways: the first draft invites reviewer questions, and the final version answers them.
As drafted. Changes in conduct, from a first-pass report.
- “Minor changes were made to the analysis”
- No dates
- No indication of blinding status at the time
- Subgroup result presented alongside the primary
- Dropped sensitivity analysis not mentioned
- “Adverse events were generally mild and well tolerated”
As issued. Same three changes, dated and positioned.
- Three changes, each stated separately with its rationale
- Covariate added 14 March — before unblinding, SAP amendment 2
- Subgroup added 2 August — after unblinding, reported as exploratory
- Exploratory results kept out of the primary efficacy section
- Sensitivity analysis dropped, with the reason recorded
- 84.2% versus 79.6% reported at least one event — tolerability argued in the discussion
This example is illustrative; the indication, dates and values have been changed. It reflects the pattern that most first-pass reports follow.
Benefits of a Well-Prepared Clinical Study Report
Common Deficiencies in Clinical Study Reports
Each of these is a pattern we see in first-pass reports, and each can lead to an additional review cycle or a question at submission.
Packages
Four Clinical Study Report Service Options, Quoted per Project
Cost depends on the trial phase, number of subjects, narrative count, report type and the amount of existing documentation. Each project is therefore quoted individually.
Patient narratives
Sponsors who need the narrative set written or rewritten
Quoted per projectPer narrative · on complexity and source quality
Delivery By volume
- ✓Deaths, serious adverse events and discontinuations
- ✓One template applied across the set
- ✓Reconciled against the safety database and listings
- ✓Consistent order, tense and level of detail
- ✓Delivered as a set with a reconciliation log
Narratives are the most commonly outsourced part of a CSR, and the part where consistency matters most.
View a sample Request a quoteSynoptic or abbreviated report
Supportive and early-phase studies
Quoted per projectOn phase, size and report type
Delivery By scope
- ✓Synopsis-level or abbreviated E3 structure
- ✓Written from the protocol, SAP and locked outputs
- ✓Deviations captured in section 9.8
- ✓Two review cycles
- ✓Abbreviation justified in the report
Where a full report is not warranted, the justification for the abbreviation is stated in the document.
View a sample Request a quoteFull clinical study report
Pivotal and registrational studies
Quoted per projectOn phase, subjects, narratives and TLF volume
Delivery By scope
- ✓Complete ICH E3 report with indexed appendices
- ✓Efficacy reported against the pre-specified estimand
- ✓Full safety section with exposure and laboratory evaluation
- ✓Patient narratives included in scope
- ✓Section 9.8 drafted against protocol and SAP amendments
- ✓Structured for reuse in Module 2.5 and 2.7
- ✓Three review cycles consolidated per cycle
This is our most common engagement. Please send the SAP amendments with the SAP, because section 9.8 depends on their dates.
View a sample Request a quoteProgramme-level writing
Sponsors taking a programme toward submission
Quoted per projectOn study count and submission destination
Delivery By arrangement
- ✓Multiple CSRs written to one house style and template
- ✓Integrated Summary of Safety and of Efficacy
- ✓Module 2.5 clinical overview and 2.7 clinical summary
- ✓Consistency maintained across reports and summaries
- ✓Publication set built from the same source
- ✓A single terminology and numbers base across the programme
Across a programme, the costliest error is two documents reporting different values for the same result. A single writing team and source base prevents this.
View a sample Request a quoteSend us the protocol and the SAP and we will draft section 9.8 free of charge, listing every departure from the plan with dates and blinding status. The draft takes five business days, and this section influences how closely reviewers read the rest of the report.
Quality Assurance
Our Service Commitments and Limitations
Our revision commitment, its limits and the remedy are set out before work begins.
Our revision commitment
Revision Support Within the Agreed Scope
Every statement in the report is traceable to the protocol, the analysis plan or the locked dataset, and the narratives are reconciled with the listings before delivery. Within the agreed scope, we correct any internal inconsistency in the report’s numbers at no additional cost.
01
Traceable to source
Every number in the text is traceable to a table, listing or figure, and every method statement is traceable to the protocol or the SAP.
02
Narratives reconciled
Each narrative is checked against the safety database and the subject listings, and a reconciliation log is delivered with the set.
03
Deviations listed and explained
Departures from the protocol and the analysis plan are recorded in section 9.8 with dates and blinding status, rather than included in the methods without comment.
04
Review cycles included
Two or three review cycles are included, depending on the engagement, with comments consolidated into one set per cycle at no extra charge.
Client Stories
FAQs
Clinical Study Report FAQs
If your question is not answered here, ask an expert. No quote is required.
The protocol with all amendments, the statistical analysis plan with its amendments, the TLFs or their shells, and the deviation log. If narratives are in scope, the counts of deaths, serious adverse events and discontinuations.
The amendments are as important as the original documents. A change made before unblinding is interpreted very differently from the same change made after it, and only the dates establish which applies.
Yes, and this is usually the best sequence. Methods, background and much of the structure can be drafted from the protocol and the SAP while the data are being cleaned.
Anything that depends on the results cannot be drafted early. Results written before database lock will need their conclusions revisited, and reviewers can usually tell.
It depends on the role the study plays in the submission and on what your agency expects. Pivotal and registrational studies get a full E3 report; supportive and early-phase studies often do not.
Whichever format you choose, the justification for abbreviating should be stated in the document, because an abbreviated report without a stated basis will raise questions.
Per narrative, on complexity and on the quality of the source material. A narrative built from a clean, complete listing costs materially less than one assembled from scattered sources.
Narrative count is the largest source of estimate error in CSR work. If you give us the counts at the quote stage, the quote will remain accurate.
Yes. We write to your template and style guide where you have one, and to the TransCelerate CSR template where you do not.
If you have neither, it is best to agree a house style before the first report, because a style that develops inconsistently across several reports is costly to correct.
Yes. Writing them alongside the CSRs costs less and reduces risk compared with writing them afterwards. Where the underlying reports were written for reuse, the summaries are largely extracted.
When they are written separately by a different team, the two documents often describe the same analysis in incompatible terms, which then has to be reconciled under submission pressure.
Yes. We offer analysis of your locked dataset, TLF production and SAP authoring with the estimand defined under E9(R1), and ordering these with the writing removes a handover between teams.
We do not manage the clinical database. Database build, validation, lock and custody remain with your data management function or CRO, and we write from the locked data they provide.
Yes. They are different documents: device clinical investigations are reported under ISO 14155:2026 rather than ICH E3, and the report informs the clinical evaluation rather than a CTD module.
Please note that ISO 14155:2020 has been withdrawn and was replaced by the fourth edition on 23 March 2026, so a device report template written before that date should be checked.
Related Services
Related Clinical and Regulatory Writing Services
The CSR links the trial to the regulatory submission. These related services cover the documents before and after it.
1BEFORE THE REPORTthe documents it reports against
Clinical trial supportProtocol, SAP, ICF and disclosure. BiostatisticsAnalysis and TLFs from your locked dataset.2AFTER THE REPORTdocuments that reuse its content
Regulatory writingCTD modules, submissions and responses. Clinical evaluation reportFor devices, the evidence file the report informs.3PUBLICATIONmanuscripts from the same source
Original research articleThe manuscript, to CONSORT 2025. Manuscript editingA full copyedit plus formatting to one journal.Insights
Further Reading on Clinical Study Reports
Request a Free Draft of CSR Section 9.8
We list every departure from the plan with its date and blinding status. Reviewers use this section to judge how closely to read the rest of the report. The draft is free and takes five business days.