Can a Participant Withdraw From Only Part of a Study?

Does Withdrawal Have to Be All or Nothing?

A participant no longer wants the study intervention but is happy to complete follow-up questionnaires. Another wants to stop attending clinic visits but would permit researchers to obtain routine outcome information. Someone else is comfortable with surveys but no longer wants to provide biological samples.

Must researchers interpret each of these decisions as withdrawal from the entire study?

Not necessarily. Participation can contain several components, and participants may sometimes change which components they are willing to continue. But partial withdrawal does not mean a participant can always redesign the study around their preferred combination of procedures. The approved protocol, scientific validity, safety, and applicable requirements still matter.

Partial Withdrawal May Be Possible

“Withdrawal” Can Describe Several Different Decisions

Start by Asking What the Participant Wants to Stop

The word withdraw is often treated as though it has one obvious meaning. In practice, a participant may be referring to treatment, visits, interviews, specimens, access to records, future contact, or every aspect of research participation.

OHRP specifically recognizes this distinction. Its withdrawal guidance explains that some participants want to discontinue only the primary interventional component while allowing other activities described in the approved protocol and consent to continue. OHRP recommends clarifying whether withdrawal concerns all components or only the primary intervention.

Different Components Can Sometimes Be Separated

A clinical trial may include an investigational intervention, clinic visits, laboratory testing, questionnaires, safety follow-up, medical-record review, and later contact. An observational study might include interviews, surveys, passive data collection, and access to administrative records.

Those activities need not always stand or fall together. OHRP states that when a participant withdraws only from the primary interventional component, other research activities for which the participant previously consented may continue. Examples include follow-up interviews, examinations, blood tests, imaging, or obtaining identifiable information from records, provided those activities remain within the consented and approved research.

A Participant Can Refuse a Procedure Without Automatically Losing Every Choice

A participant's refusal of a research procedure should be respected. Researchers should not perform it merely because the procedure appeared in the consent form signed at enrollment.

What happens next is a separate question. If the procedure is optional or scientifically separable, participation may continue without it. If it is essential to the intervention, safety monitoring, eligibility, or integrity of the study, continuing in the remaining study may no longer be possible.

Informed consent regulations anticipate this possibility by allowing consent materials, when appropriate, to describe circumstances under which investigators may terminate participation without the participant's consent and the consequences and procedures associated with withdrawal.

The Right to Refuse Is Not the Same as a Right to Choose Any Study Configuration

This distinction prevents an important misunderstanding. A participant controls whether they undergo a research procedure. They do not necessarily control whether a scientifically defined study can continue to enroll them after that procedure is refused.

For example, if a safety laboratory test is essential to determining whether an investigational treatment can be administered safely, a participant may refuse the blood draw, but the investigator may consequently be unable to continue giving the investigational treatment.

Respecting refusal and following the protocol are therefore compatible. The participant can say no, and the research team can explain honestly what that choice means for further participation.

Stopping Treatment and Leaving the Trial Are Not Always the Same Thing

This distinction is particularly important in clinical trials. ICH E6(R3) separately recognizes stopping investigational treatment, stopping trial visits, and completely withdrawing from a trial. It directs investigators to follow the protocol and sponsor instructions for appropriate follow-up when participation ends or changes.

FDA guidance similarly states that a participant may withdraw from the interventional portion of a study while agreeing to continued collection of clinical outcome information. When such limited continued participation was not covered by the original informed consent, FDA states that informed consent for the continued follow-up must be obtained.

This is why researchers should avoid treating “stopped treatment” and “lost to all follow-up” as synonyms.

Clarification Should Be Neutral

Researchers have legitimate reasons to understand exactly what a participant wishes to stop. Follow-up data can be important for assessing safety and reducing bias, particularly in clinical trials. OHRP explicitly recognizes the importance of continued follow-up information in some trials.

But scientific importance does not justify pressure. Researchers may explain why follow-up matters and describe available options, while making clear that the participant controls whether to agree to research activities for which their consent is required.

Partial Withdrawal Can Change Over Time

A participant might initially stop treatment but agree to follow-up, then later decide that they no longer want follow-up either. Their earlier willingness to remain partially involved does not permanently commit them to that arrangement.

The broader principle that participants can change their minds after giving consent therefore continues to apply after a partial withdrawal.

Already Collected Data Remain a Separate Issue

Partial withdrawal primarily tells researchers which future activities the participant no longer accepts. It does not by itself determine whether information already collected through those activities must be deleted.

OHRP allows retention and analysis of already collected data within the approved protocol, while FDA requires already-accrued data from FDA-regulated clinical trials to remain in the trial database.

Questions about whether participants can also withdraw permission for continued use of their data therefore require a separate analysis.

Plan for Different Levels of Withdrawal Before Anyone Withdraws

OHRP recommends that investigators anticipate withdrawal in protocols and consent materials and explain what withdrawal means. Doing this well can prevent confusion later.

For a complex study, the consent process can distinguish mandatory from optional procedures and explain whether someone who stops the intervention may remain in safety follow-up, whether record-based follow-up is contemplated, and what happens to previously collected information.

Good planning makes it less likely that a participant's simple “I don't want that anymore” will trigger an improvised negotiation at the study visit.

Stopping the Intervention but Continuing Follow-Up

The participant has stopped one important component without necessarily disappearing from the study altogether. That distinction can protect autonomy while preserving ethically authorized follow-up.

Common Mistakes About Partial Withdrawal

Translate “I Want to Stop” Into the Participant's Actual Choice

When a participant wants to reduce their involvement, map that preference against the components of the study rather than forcing it immediately into a binary stay-or-leave category.

When a Participant Wants to Stop Only Part of the Study

Questions About Partial Withdrawal From Research

Withdrawal Does Not Always Have to Mean Leaving Everything

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.