11.2 Adherence, Discontinuation and Participant Choice

Adherence, Discontinuation and Participant Choice

Review the actual prescribed regimen, dispensing, returns, medical holds, participant report and relevant clinical events. A low tablet count can reflect missed doses, lost supplies, an incorrect denominator or recording error. A participant may misunderstand instructions or stop because of an unreported adverse effect. Ask respectfully and involve the investigator, avoiding an accusation based solely on a percentage.

For a fixed-dose example, a participant receives 28 tablets for one tablet daily over 28 days. Four tablets remain and there were no authorised holds. The simple return-based estimate is (28 − 4)/28 × 100 = 85.7%, to one decimal place. It estimates unreturned tablets, not observed ingestion. If two tablets were lost, the count cannot establish that 24 were swallowed. Record the uncertainty and apply the protocol's evaluation process.

Do not assume a universal acceptable adherence range. The clinical importance depends on the product, schedule and study question. Missing one dose of a narrow-margin treatment may matter differently from a missed optional questionnaire. A protocol threshold defines an operational action; it does not replace medical assessment or prove dishonesty.

Participant responsibilities should be understandable

The consent process and study instructions should explain relevant visits, product use, diaries, reporting of symptoms, prohibited treatments and contact routes. Provide information suited to the participant's language, abilities and technology access. Check understanding and practical barriers. A participant cannot reliably follow a diary schedule if they were never shown how to use the application or what to do during a connection failure.

Documenting discontinuation

Capture the relevant date, activity stopped, reason if provided, investigator assessment, product disposition, safety status and permitted follow-up. Distinguish treatment discontinuation from withdrawal of further contact. A participant is not obliged to explain withdrawal. Staff may make a reasonable respectful effort to understand concerns and offer permissible alternatives, but should not coerce a reconsideration.

Suppose someone stops dosing because of nausea but agrees to telephone safety follow-up and the planned final assessment. The record should preserve those distinctions. The site should not classify the person as lost to follow-up, erase earlier adverse events or automatically cancel every future activity. If the person later declines all contact, update the scope accurately and follow the applicable process.

For a protocol-mandated safety stop, verify that qualified staff reviewed the triggering information and implemented the appropriate care and notifications. A stopping decision may require prompt sponsor communication and ethics/authority reporting depending on the circumstances. Do not assume that following a prewritten stopping rule is itself a protocol deviation. Failing to follow that rule can be a significant concern.

Data and safety after stopping

The handling of already collected information should be described in the protocol/consent process and follow applicable requirements. Withdrawal does not automatically require deletion of all existing research data, and a desire for complete data does not authorise new collection without permission. Preserve the decision record and seek responsible guidance when scope is uncertain.

Safety follow-up should respect the protocol, applicable requirements and participant choice. Track outstanding clinical information, reasonable contact efforts and the point at which further attempts are no longer permitted or appropriate. Do not mark an event resolved merely because the participant is unreachable. Retain an honest status such as ongoing or unknown when supported by the facts.

At close-out, reconcile participant status across the source, EDC, product records and tracking systems. Explain legitimate differences, such as a product stop date preceding the final follow-up visit. These distinctions support both participant protection and interpretation of the treatment effect, especially when intercurrent events and missing data affect the planned analysis.

Identify the interest before deciding how to manage it

A conflict of interest arises when a secondary interest could influence, or appear to influence, professional judgement about participant welfare or research integrity. Financial interests are one example, but professional advancement, family relationships, employment hierarchies and institutional recruitment pressure can also matter. A conflict is not proof that misconduct occurred; an undisclosed or unmanaged conflict can undermine independent decisions.

Official sources

Checked October 18, 2026: ICH E6(R3), Principles and Annex 1, ICH E2A, WMA Declaration of Helsinki, 2024.