Structure and Content of Clinical Study Reports: ICH E3

Under ICH E3, the structure of a clinical study report follows a numbered 16-section outline that walks a regulator from title page through appendices in a fixed order: identification and synopsis, table of contents and abbreviations, ethics, investigators, introduction, objectives, investigational plan, study patients, efficacy, safety, discussion and conclusions, and the appendices holding source documents and patient-level data.¹ The point of the fixed order is practical. A reviewer at the FDA, EMA, or PMDA should be able to open any CSR and find the randomization scheme, the dropout accounting, or the serious adverse event narratives in the same place every time.

What follows is what each section is for and what has to be inside it.

The Synopsis (Section 2)

The synopsis is a standalone summary. ICH E3 says it should usually be limited to about three pages, though the guideline sets no hard upper limit and complex studies routinely run longer.¹ It must carry the title-page identifiers, the objectives, the design, and the key efficacy and safety results with actual numerical values, plus the conclusions. P-values or narrative statements alone are not enough; the numbers behind them belong in the synopsis itself. Cross-references to other sections defeat the purpose, because the synopsis is meant to work on its own for a reviewer who has not yet opened the full report.

Ethics, Investigators, Introduction, and Objectives (Sections 5–8)

Sections 5 and 6 establish that the study was ethically conducted and identify who ran it. The report must confirm that the original protocol and any amendments were reviewed by an Independent Ethics Committee or Institutional Review Board, that the study followed Good Clinical Practice, and that patients gave informed consent.¹ A list of investigators and other key study personnel, with brief CVs, is placed in the appendices.

Section 7 (Introduction) sets the scientific and clinical background. Section 8 (Study Objectives) states the question the trial was designed to answer. Both are short, but they frame everything downstream.

The Investigational Plan (Section 9)

Section 9 lays out the study’s blueprint. It has subsections for the overall design, patient selection, treatments, efficacy and safety variables, data quality assurance, and the statistical methods planned before the study began.

Design and Patient Selection

The report must describe whether the study was randomized, blinded, parallel-group, crossover, or something else, and explain why the chosen control (placebo, active comparator, dose-response) was appropriate. A flow chart showing design and assessment timing is strongly encouraged.¹ Inclusion criteria, exclusion criteria, and the conditions for removing a patient from therapy or assessment each get their own subsection, so the reviewer can judge whether the enrolled population maps onto the eventual prescribing population.

Treatments and Blinding

Every treatment administered must be described with drug identity, dosage form, dose, route, and duration. The randomization procedure and the blinding scheme need explicit documentation: who was blinded, how the blind was maintained, and the circumstances under which it could be broken. Prior and concomitant therapies are documented as well, because they can confound the results.¹

Pre-Specified Statistical Plan and Sample Size

Section 9.7 describes the statistical and analytical plans as they existed before results were known. It identifies which measurements were designated as the basis for comparing treatment and control, whether the primary analysis would adjust for covariates, and how sample size was determined.

Changes in Conduct or Analyses

Section 9.8 is one of the most closely read parts of any CSR. Any change made after the study began, whether a treatment arm was dropped, entry criteria adjusted, sample size revised, or the statistical plan modified, must be described with the timing, the reason, who decided, and what data those decision-makers had access to at the time. ICH E3 draws a sharp line: changes made before the blind was broken carry fewer interpretive concerns than changes made afterward, and the timing relative to unblinding has to be stated clearly.¹

Patient Disposition and Protocol Deviations (Section 10)

Section 10 has to account for every patient who entered the study. Figures or tables in the body of the report show how many were randomized, how many entered and completed each phase, and how many discontinued. Discontinuations are grouped by treatment arm and by reason: adverse event, lost to follow-up, poor compliance, and so on. Whether patients were followed for the full study duration even after stopping the study drug must be made clear.¹

Protocol deviations get a separate subsection (10.2). Individual patients with deviations are listed in the appendices, broken down by center in multicenter studies, because a cluster of deviations at a single site can undermine data reliability there without necessarily tainting the whole study.

Efficacy Evaluation (Section 11)

Section 11 makes the case that the treatment works. It opens by defining precisely which patients were included in each analysis population. Even if the sponsor’s primary analysis uses a reduced subset such as per-protocol completers, ICH E3 expects an additional analysis using all randomized patients with any on-treatment data. Criteria for excluding patients from the efficacy analysis are listed individually in the appendices with reasons.¹

Endpoint Analyses

The report compares treatment groups on all primary and secondary endpoints. “Compare” has a specific meaning here: the analysis should show the size of the difference between treatments as a point estimate, provide confidence intervals, and present hypothesis testing results where used. Tables, figures, and individual patient data listings back the findings.¹

Statistical and Analytical Issues

ICH E3 devotes several subsections to potential analytical pitfalls. The efficacy section has to address:

Covariate adjustments, including why specific demographic or baseline variables were selected and how the adjustment was made.
Missing data and dropouts, including the procedures used to handle missing values and any assumptions behind imputation. Drawing conclusions only from completers is flagged as potentially misleading, especially when dropout rates differ between arms.
Multicenter analyses, including how site data were combined and whether treatment effects were consistent across centers.
Multiple comparisons, including any adjustments for testing multiple endpoints or interim looks at the data.
Subgroup analyses by important demographic variables (age, sex, race, disease severity) in studies large enough to support them, framed as hypothesis-generating rather than as a rescue for a failed primary analysis.

Any changes to the planned statistical approach made after breaking the blind must be clearly identified.¹

Safety Evaluation (Section 12)

Section 12 characterizes the risks of the treatment. It is typically the longest section of a CSR because regulators need to see safety data from several angles.

Extent of Exposure

Before adverse events, the report documents how much drug patients actually received: the number exposed, the duration (as median and in categorized bands such as one day or less, up to one week, up to one month, and longer), and the doses used. Where available, drug concentration data help correlate exposure with adverse events or lab abnormalities.¹

Adverse Events

All adverse events that occur after treatment begins are presented in summary tables grouped by body system, with the number and rate of occurrence in each treatment group. Events may be further split by severity (mild, moderate, severe) and by the investigator’s causality assessment. Patient-level listings in the appendices carry demographics, event term, duration, severity, seriousness, action taken, outcome, and timing relative to the last dose.¹

Serious adverse events and deaths get more. Each is listed individually and accompanied by a narrative describing the event, its severity, the investigator’s causality assessment, and the outcome. Narratives supply the context that summary tables cannot.

Laboratory Findings, Vital Signs, and ECG

Safety data extend beyond adverse event reports. The CSR summarizes laboratory evaluations covering hematology, clinical chemistry, and urinalysis, with attention to values that shifted from normal to abnormal during the study. Vital signs and ECG results round out the physiological picture. Presentation by demographic subgroup helps flag populations at heightened risk.

Appendices (Section 16)

Section 16 is the evidentiary backbone. It holds the documents and data listings that let a regulator audit the study independently, organized into four groups.¹

Study Information (16.1)

16.1.1 Protocol and amendments — the final protocol and every amendment.
16.1.2 Sample case report forms — unique pages only.
16.1.3 IEC/IRB list and representative consent forms.
16.1.4 Investigator list and brief CVs.
16.1.5 Signatures of the principal or coordinating investigator, or the sponsor’s responsible medical officer.
16.1.6 Batch assignments, if more than one manufacturing batch was used.
16.1.7 Randomization scheme and codes.
16.1.8 Audit certificates, if available.
16.1.9 Documentation of the statistical methods used.
16.1.10 Inter-laboratory standardization and quality assurance documentation, when applicable.
16.1.11–16.1.12 Publications based on or referenced in the study.

Patient Data Listings (16.2)

These appendices carry individual-level data for defined groups: discontinued patients, protocol deviations, patients excluded from the efficacy analysis, demographic data, compliance and drug concentration data, individual efficacy responses, and per-patient adverse event listings. ICH E3 stresses that listings should be “readily usable by the reviewer,” so clarity outranks compactness.¹

Case Report Forms and Archival Listings (16.3–16.4)

Section 16.3 contains actual case report forms for deaths, serious adverse events, and withdrawals due to adverse events. Section 16.4 holds complete individual patient data listings, sometimes called archival listings, which supply the full dataset for regulatory archival purposes.

Where the CSR Sits in the Submission

A CSR is not submitted alone. It lives inside Module 5 of the Common Technical Document, which organizes clinical study reports by category: biopharmaceutic studies, pharmacokinetic studies, pharmacodynamic studies, efficacy and safety studies, and post-marketing experience.² Each study report is submitted as multiple documents — synopsis, main body, and appendices per ICH E3 — and each is identified in the Module 5 table of contents. That granularity lets sponsors replace specific documents as new information arrives without resubmitting the whole module.

U.S. Requirements Layered on Top: 21 CFR 314.50

ICH E3 gives you the internationally harmonized structure, but a U.S. New Drug Application must also satisfy the content requirements of 21 CFR 314.50. The regulation calls for a clinical data section that includes a description and analysis of each controlled and uncontrolled clinical study, an integrated summary of effectiveness, and an integrated summary of safety.³

It adds a few requirements ICH E3 does not spell out on its own. Effectiveness data must be presented by gender, age, and racial subgroups and must identify any dose modifications needed for specific subgroups such as patients with renal impairment or varying disease severity. Safety data carry the same subgroup presentation requirement. Interim analyses must be noted with a projected completion date.³

Write for Eventual Public Disclosure

CSRs are increasingly public documents. The European Medicines Agency publishes the clinical data submitted to support marketing authorization applications, including individual study reports, the study protocol, sample case report forms, and statistical methods documentation, typically within 60 days after the European Commission’s decision for standard marketing authorizations.⁴

Before publication, reports are anonymized to prevent identification of trial participants, with personal data redactions appearing as labeled boxes. Companies can request redaction of commercially confidential information, but the EMA takes the position that clinical data generally does not qualify as commercially confidential except in limited circumstances.⁴ Draft the CSR so that anonymization can be applied cleanly without hollowing out the report.

1
International Council for Harmonisation. ICH E3 – Structure and Content of Clinical Study Reports
2
European Medicines Agency. CTD for the Registration of Pharmaceuticals for Human Use – Organisation of CTD
3
eCFR. 21 CFR 314.50 – Filing of an Application
4
European Medicines Agency. Clinical Data Publication