Withdrawing a Participant for Their Benefit

Part of Principal Investigator (PI) Training: Medical Care of the Participant.

About 4 minutes. Free, in your language, no account needed.

Why this matters

Discontinuing a participant is one of the few decisions where the investigator's clinical duty and the trial's interests point in opposite directions and everyone can see it. Enrollment was hard. The participant may want to continue. The sponsor counts completers.

Principle 1.5 resolves it before the argument starts: the rights, safety and wellbeing of participants prevail over the interests of science and society. The remaining question is how to do it well.

The distinctions that matter

Three different things get called "withdrawal", and confusing them causes real harm to data and to people.

• Discontinuing investigational product while the participant remains in the trial and continues follow-up. Often the right answer: the exposure stops, the safety follow-up continues.
• Withdrawing from the trial — the participant stops trial participation. Follow-up may still be possible with their agreement, and often should be sought.
• Withdrawing permission to use data already collected. Section 2.9.1 distinguishes this explicitly from withdrawing from the trial. A participant leaving the study does not automatically retract consent for data already gathered, and the two should be discussed separately rather than assumed together.

Say the difference out loud to the participant. "You can stop the treatment and still let us check on you" is an option most people do not know exists, and it is frequently the safest outcome.

When to make the call

Your protocol will list discontinuation criteria. Beyond them, the judgment is yours, and the recurring triggers are:

• A trial-related adverse event that is severe, recurrent, or not resolving.
• A change in the participant's condition that makes continued exposure unfavorable — a new diagnosis, a deterioration, a pregnancy where the protocol requires it.
• An intercurrent illness whose treatment is incompatible with the protocol.
• Persistent non-adherence that makes the data unreliable and means the person is bearing risk without contributing an answer.
• Burden the participant is carrying but not voicing — often visible as repeated late arrivals, missed assessments and canceled visits before it is ever said.

That last one deserves attention. Some participants will not tell you they want out; they will simply become difficult to schedule. Reading the pattern and opening the conversation yourself is a kindness and usually produces a better outcome than attrition.

Doing it properly

Follow the protocol's procedure (§2.9.1) — the required assessments, safety follow-up period, final visit and any taper or washout.

Ask why, and respect the answer (§2.9.2). A reasonable effort should be made to ascertain and address the reasons for premature withdrawal, while fully respecting the participant's rights. "Address" can mean solving a practical problem — transport, timing, a schedule that clashes with work — that would have kept someone in the trial. It never means pressure.

Keep the safety follow-up if you can. Ask, separately and explicitly, whether they will allow follow-up contact for safety purposes. Many will. An untracked participant with an unresolved trial-related event is the worst outcome of the three.

Document the medical reasoning. Not "withdrawn per PI decision" but the clinical basis: what you observed, what you weighed, what you decided. A withdrawal for a participant's benefit is a defensible act and it should be recorded so that it reads as one.

Do not close the door. Section 2.9.3 provides for informing participants of the trial results and, as applicable, the treatment they received, when available and according to their preference. Someone who left early is still owed that.

Key takeaways

• Principle 1.5 settles the conflict: the participant's safety and wellbeing prevail.
• Distinguish stopping the product, leaving the trial, and withdrawing permission to use data already collected — §2.9.1 treats the last as separate.
• §2.9.2: ask why, address practical obstacles where you can, never apply pressure, and seek agreement for safety follow-up separately.
• Record the clinical reasoning, not just the fact — and honor §2.9.3 for participants who left early.