10. Study participants
Section 10 of a clinical study report gives the account of the people in the study, from who is screened and randomised to who completes it, why people leave, and which important departures from the protocol occur. ICH E3 divides it into two subsections: the disposition of patients in 10.1, with a listing of discontinued participants in Appendix 16.2.1, and protocol deviations in 10.2, with a listing of the participants concerned in Appendix 16.2.2. The CORE Reference manual moves four further sections into Section 10.
What Section 10 asks for
Section 10 accounts for the people who enter the study: who starts, who finishes, why people leave, and which important departures from the plan occur. The methods in Section 9 establish the intended participant journey. Section 10's participant counts supply context for the efficacy analyses in Section 11 and treatment exposure in Section 12.
Two questions organise a participant's history from their first eligibility assessment to their final visit:
| Everyday question | Subsection |
|---|---|
| Who starts, who finishes, and why do people leave? | 10.1 — Disposition of patients |
| What important departures from the protocol occur? | 10.2 — Protocol deviations |
10.1 — Disposition of patients
Disposition is the accounting of participants as they move through the study. When 250 people are assessed for a trial, some are ineligible, some decline participation, 200 are assigned to treatments, and fewer complete treatment; the reader needs to understand every transition.
These terms describe different stages:
| Term | Plain meaning |
|---|---|
| Screened | Reaches the eligibility assessment described in 9.1 |
| Screen failure | Does not satisfy the conditions for proceeding beyond screening |
| Enrolled | Enters the study under its defined enrolment criteria |
| Randomised | Reaches the assignment stage described in 9.4.3 |
| Treated | Actually receives study treatment |
| Completed treatment | Reaches the end of the planned treatment period |
| Completed the study | Reaches the study’s defined completion point, potentially including follow-up |
| Discontinued | Stops something before its planned completion—the report must specify what |
“Enrolled” needs a clear study-specific definition. It is not automatically interchangeable with “randomised” or “received treatment.”
Stopping treatment versus leaving the study
A participant might stop taking the medicine because of nausea but continue attending visits and providing outcome information. That person discontinues treatment and continues study participation. Another participant might stop treatment and decline any further study participation. A third might become unreachable. Those situations produce different amounts of information and have different implications. E3 explicitly asks the report to make clear whether participants continue to be followed after stopping treatment.
For example:
| Participant | Treatment | Follow-up | What the report should distinguish |
|---|---|---|---|
| A | Stops at Week 4 because of nausea | Attends all remaining visits | Treatment discontinued; study completed |
| B | Stops at Week 4 | Declines further participation | Treatment and participation end early |
| C | Last known treatment at Week 4 | Cannot subsequently be contacted | Lost to follow-up; later status unknown |
Lost to follow-up means that further information cannot be obtained through the study’s follow-up process. It should not be used as a substitute for a known decision to withdraw.
What E3 asks the main report to show
Section 10.1 calls for a clear accounting of everyone who enters, using tables or figures. It asks for:
- Numbers randomised.
- Numbers entering and completing each study phase, or relevant time periods.
- Reasons for all post-randomisation discontinuations.
- Results separated by treatment group and major discontinuation reason.
- Screening numbers and reasons for exclusion when helpful for understanding the population relevant to eventual drug use.
- A clear explanation of follow-up after treatment discontinuation.
E3 names lost to follow-up, adverse event and poor compliance among the major reasons, and says a flow chart is often helpful.
A study “phase” in this list can mean a period such as screening, treatment, or follow-up, not necessarily development Phase I, II, or III.
A simplified treatment-disposition table could look like this:
| Status | Investigational treatment | Placebo |
|---|---|---|
| Randomised | 100 | 100 |
| Received at least one dose | 98 | 99 |
| Completed treatment | 80 | 90 |
| Discontinued treatment early | 18 | 9 |
| — Adverse event | 10 | 2 |
| — Lack of efficacy | 3 | 4 |
| — Participant decision | 3 | 2 |
| — Other reason | 2 | 1 |
The report must also account for the two and one randomised participants who never start treatment. The table’s treatment-discontinuation rows cover only those who start. A separate display clarifies who completes study follow-up, including people who stop treatment.
The arithmetic should reconcile: in the investigational group, 98 start treatment, 80 complete it and 18 discontinue it. But 80 treatment completers does not establish that only 80 complete the study.
The denominator introduced in Section 2 needs to match the participant stage being counted. Tables often use N for the group size and n for the number meeting a condition; the table must define its notation. When percentages are reported, the denominator should be explicit. “10% discontinued because of an AE” could use the randomised or treated population; the reader should not have to guess.
Reasons for leaving
A total number of discontinuations can conceal very different stories, and each pattern below is a reason to investigate, not an automatic conclusion:
- More people leave because of adverse events: possible difficulty tolerating treatment.
- More leave because of lack of efficacy: possible inadequate benefit.
- People leave because of travel restrictions or site closure: operational disruption.
- People become lost to follow-up: uncertainty about subsequent outcomes.
- People leave because of poor compliance: difficulty following required treatment or procedures.
Differences between groups in withdrawal reasons and in missing values, and clusters around particular time points, should be looked for, and disposition tables are a starting point for recognising these problems.
The adherence methods in 9.4.8 help explain poor-compliance withdrawals. Missing doses does not necessarily end participation.
Attrition and bias
Attrition is the loss of participants over time. If people who experience unpleasant effects disproportionately leave the investigational-treatment group, a comparison of only those who finish may end up evaluating a selected group of people who tolerate the medicine. This is a route to the bias introduced in 9.2: the remaining participants may no longer support the intended comparison. Randomisation helps make groups comparable at the start, but selecting only the people who remain does not necessarily preserve that comparability.
The important questions are therefore:
- How many people leave?
- When do they leave?
- Why do they leave?
- Do those patterns differ between treatments?
- What information remains available afterwards?
The direction of bias depends on the circumstances and analysis; it cannot be inferred from dropout counts alone. Simply discarding participants with incomplete data can both lose useful information and distort conclusions.
Section 10 establishes the participant history, and Section 11.4.2.2 examines how missing observations affect the analysis. Section 11.4.2.2 holds what E3 asks a report to compare between the treatment groups about its dropouts, its caution against conclusions drawn only from completers, and the assumptions it wants stated for the methods used for unavailable observations.
Appendix 16.2.1: discontinued participants
The main text gives the overview. E3 also asks for a listing of all participants discontinued from the study after enrolment, organised by centre and treatment group.
Centre-level organisation follows the study structure described in 9.1. Cumulative dose is the total amount received over time.
| Required information | Why it matters |
|---|---|
| Participant identifier | Connects the record to other study data |
| Specific discontinuation reason | Explains what happened beyond a broad category |
| Treatment and dose | Establishes what the participant receives |
| Cumulative dose, where appropriate | Shows total exposure before discontinuation |
| Treatment duration | Shows how long treatment continued |
| Whether discontinuation involves unblinding | Records whether treatment assignment becomes known |
The blinding procedures in 9.4.6 provide the reference for interpreting any disclosure of treatment assignment at discontinuation.
E3 also says it may be useful to include:
- Age, sex, race, and other critical demographics.
- Concomitant medicines.
- Major response measurements at termination—for example, the principal disease measurement at the final assessment.
E3 points to Annexes IVa and IVb for flow-chart examples and Annex V for a discontinuation-listing example.
10.2 — Protocol deviations
The eligibility, treatment, and assessment rules in Section 9 provide the reference for assessing conduct. A protocol deviation is a departure from that design or those procedures, and the questions and answers document that ICH issued on E3 in 2012 defines it as any change, divergence or departure from the study design or procedures defined in the protocol.
Examples might include:
- Enrolling someone who does not meet an eligibility criterion.
- Giving the wrong dose.
- Missing a required assessment.
- Allowing a prohibited medicine.
A deviation does not automatically mean deliberate misconduct. It describes a departure; the cause and consequences need separate examination.
Important deviations
E3 asks for important deviations. Important protocol deviations are the subset of protocol deviations that may significantly affect the completeness, accuracy or reliability of the study data, or a participant’s rights, safety or well-being: the 2012 questions and answers document, ICH E6(R3) and the TransCelerate template all define them in those terms. ICH E6(R3) says the sponsor should determine the necessary trial-specific criteria for classifying protocol deviations as important.
Importance depends on the study. A small timing departure may be inconsequential for one measurement but critical for another. Failing to obtain a measurement needed for the primary endpoint can be important even without physical harm, and the primary endpoint defined in 9.5.3 helps determine which missed assessments threaten the study’s main answer.
The four categories E3 illustrates
| E3 example | Plain-language explanation | Fictional example |
|---|---|---|
| Entered despite not meeting entry criteria | Someone enters despite not qualifying | The protocol requires confirmed disease, but a participant lacks the required confirmation |
| Meets withdrawal criteria but remains in treatment or assessment | A protocol-defined stopping condition occurs without the required withdrawal | A specified safety threshold is reached and treatment continues contrary to the protocol |
| Received wrong treatment or dose | Treatment administration differs from the protocol | A participant assigned 10 mg receives 20 mg |
| Received an excluded concomitant treatment | A prohibited additional treatment is used | A participant takes a medicine barred because it could affect the study outcome |
Meeting a withdrawal criterion is not itself necessarily a deviation. If the protocol says to stop treatment when a threshold is reached and the team does so, the team has followed the protocol; the example deviation is failing to take the required action. Nor is a dose change permitted by the protocol automatically a deviation: a protocol may allow reductions to manage adverse events.
The examples are not exhaustive. E3 also covers important departures involving overall conduct, participant management, and assessment. The 2012 questions and answers document allows study-specific categories, with substantial changes or additions explained to the reviewer.
Presentation of deviations
E3 asks for:
- A description of all important deviations.
- A summary in the main text, organised into categories and appropriately by centre.
- A participant-level listing in Appendix 16.2.2, broken down by centre in multicentre trials.
A practical summary might look like this:
| Important deviation category | Investigational treatment: participants affected | Placebo: participants affected |
|---|---|---|
| Key eligibility criterion not met | 2 | 1 |
| Incorrect dose | 3 | 0 |
| Prohibited concomitant treatment | 4 | 2 |
| Required primary-endpoint assessment missed | 5 | 3 |
That is an illustrative format, not an exact E3 template; centre-level information also needs appropriate presentation.
Participants and occurrences are different counts. One person might have three dosing deviations and a missed assessment. The report should make clear whether a number represents people or events. Category totals may overlap.
Centre-level summaries help the reviewer distinguish a widespread issue from one concentrated at a particular site. They support investigation of the pattern; they do not by themselves establish its cause.
Deviations and exclusion from analysis
The following are separate decisions:
| Question | Relevant CSR location |
|---|---|
| Does a participant leave, and why? | 10.1 |
| Does an important deviation occur? | 10.2 |
| Which participants and observations enter each efficacy analysis? | 11.1 |
| How does the analysis handle dropouts and missing data? | 11.4.2.2 |
For example, someone who receives an incorrect dose can still have valuable safety data and may remain in an efficacy analysis, depending on the defined analysis rules.
The 2012 questions and answers document distinguishes withdrawal due to a deviation from inclusion in study analysis. E3 asks, in 11.1, for precise definitions of the analysis populations and explanations of exclusions.
“Protocol violation” versus “protocol deviation”
The term protocol violation appears in older material, including E3’s example disposition chart, where it is one of the listed reasons for withdrawal.
The two terms do not establish a universal minor-versus-major classification; the report’s definitions matter.
The 2012 questions and answers document defines a protocol deviation and an important protocol deviation as given above, and says of the third term that protocol violation and important protocol deviation are sometimes used interchangeably to refer to a significant departure from protocol requirements. To avoid confusion over terminology, sponsors are encouraged to replace the phrase "protocol violation" in E3 Annex IVa with "protocol deviation" or another descriptor, provided the information presented stays generally consistent with that definition of protocol violation.
Disposition and deviations together
A participant can:
- Complete the study with an important deviation.
- Leave the study without a deviation.
- Have a deviation that leads to withdrawal.
- Stop treatment under the protocol and continue follow-up.
The amendments described in 9.8 establish which procedures apply. Following an applicable amended procedure does not automatically constitute a deviation.
Pandemic disruption gives a practical example of why causes should remain visible: it means distinguishing pandemic-related discontinuations, missed visits and dosing disruptions, and examining whether the treatment groups differ in their exposure to disruption. It is an example, not an additional universal Section 10 requirement.
Where the information lives
| Location | Content |
|---|---|
| 10.1 | Participant flow, completion, discontinuation reasons, and continued follow-up |
| 16.2.1 | Individual participants discontinued after enrolment |
| 10.2 | Important protocol deviations, summarised by category and centre |
| 16.2.2 | Individual participants with those deviations |
| 11.1 and 16.2.3 | Analysis populations and exclusions from efficacy analyses |
| 11.4.2.2 | Statistical handling of dropouts and missing data |
Section 10 in CORE Reference
The CORE Reference manual enlarges Section 10. Its reading is that three of E3's Section 11 subsections — data sets analysed, demographic and other baseline characteristics, and measurements of treatment compliance — are inappropriately located at the start of the efficacy evaluation, and it takes a fourth, extent of exposure, out of the safety evaluation, where E3 has it as Section 12.1. The manual moves all four into Section 10, as 10.3, 10.4, 10.5 and 10.6.
The manual gives a reason for each move. Data sets analysed describes the participant population and so applies to more than the efficacy evaluation. Demography, medical history and prior medications relate to the safety population as well as the efficacy one, and so should not be limited to Section 11. Treatment compliance relates to every participant, not only to those with efficacy data. And extent of exposure relates to every participant, not only to those with safety data.
At Section 10, the manual draws attention to protected personal data and to the risk of de-anonymisation from aggregated data. It asks the writer to consider whether discussing any individual participant’s information in the text is necessary, and to consider data presentations that keep the data’s meaning in context and conform to current minimum standards for de-identifying data, standards that continue to evolve.