Research With Vulnerable Participants: What Extra Protections Are Needed?

What changes when your research involves participants who may be vulnerable?

Suppose your study involves children, patients receiving care, people with limited financial resources, employees recruited by their employer, migrants with precarious legal status, or adults whose decision-making capacity may be impaired. You may already know that these participants can require additional ethical protection. The harder question is what that protection should actually look like.

There is no single package of safeguards that applies to everyone described as vulnerable. A participant might face difficulty understanding a complex consent process, pressure arising from a dependent relationship, unusually serious consequences if confidential information is disclosed, limited freedom to refuse, or several of these concerns at once. Another person belonging to the same population may face a different combination of risks.

This matters because protection can fail in two directions. Too little protection may expose people to coercion, exploitation, avoidable harm, or loss of privacy. Too much protection can become paternalistic and exclude people from research that concerns them, leaving important evidence gaps in the process.

Extra protection should respond to the source of vulnerability

Protection begins by asking what could make participation less safe or less voluntary

Vulnerability is not simply a label attached to certain people

Older approaches to research ethics often identified particular populations as vulnerable. Contemporary guidance is more cautious about treating vulnerability as a permanent label. The 2024 World Medical Association Declaration of Helsinki recognizes that individuals, groups, and communities may experience greater vulnerability because of factors that are fixed or because of circumstances that are contextual and dynamic.

CIOMS similarly describes vulnerable persons as those who may have an increased likelihood of being wronged or incurring additional harm in research. Its guidance emphasizes characteristics and circumstances that create vulnerability rather than assuming that everyone within a broad demographic category requires identical protection.

This distinction changes the researcher's task. Instead of stopping at the question, “Are these participants vulnerable?”, ask what vulnerability means in the circumstances of this particular research.

Group membership can still be ethically and legally important. Some jurisdictions establish specific requirements for particular populations. But identifying a population should usually begin the analysis rather than finish it.

First identify the source of vulnerability, then choose the safeguard

Different vulnerabilities call for different protections. A person who has difficulty understanding information does not necessarily need the same safeguard as someone who understands the study perfectly but fears retaliation for refusing. Likewise, a participant whose principal risk is involuntary disclosure of a stigmatized identity faces a different problem from someone being offered an incentive in circumstances of severe financial need.

These are possible safeguards, not a universal prescription. Their appropriateness depends on the research, participant population, governing rules, and ethics review.

Extra protection does not automatically mean substitute consent

One particularly important distinction concerns vulnerability and decision-making capacity. Being vulnerable does not itself establish that a person lacks the capacity to make a research decision. Someone may be vulnerable because of poverty, dependence on healthcare, stigma, institutional authority, or privacy risks while remaining fully capable of understanding the study and deciding whether to participate.

Conversely, impaired decision-making capacity may create a specific need for additional measures. The appropriate response can depend on the degree and duration of impairment, the complexity and risk of the decision, applicable law, and institutional requirements. The ethical questions surrounding vulnerability and informed-consent capacity therefore need to be kept separate.

Recruitment itself may create pressure

Protection begins before a consent form is signed. Consider who identifies potential participants, who approaches them, where recruitment occurs, what they are told, and what relationship exists between recruiter and participant.

The Belmont Report distinguishes coercion from undue influence and emphasizes that voluntariness requires conditions free from both. Pressure becomes particularly important where authority or dependency is involved. A student recruited directly by a professor, an employee approached by a supervisor, or a patient invited by someone involved in their care may wonder whether refusal could affect grades, employment, treatment, or access to services, even when the researcher never intended such pressure.

The ethical problem, then, may not lie in the wording of the consent form at all. It may lie in who is holding it.

When participation occurs within healthcare, employment, education, or other dependent relationships, recruitment procedures should be examined for actual and perceived pressure. Depending on the context, separating the authority figure from recruitment, consent, or knowledge of who participated may provide a more meaningful safeguard.

Payment deserves contextual analysis, not an automatic ceiling

Compensation can reimburse expenses, compensate participants for time or inconvenience, or provide an incentive to participate. Payment is not inherently unethical. The concern is whether an offer could improperly influence a person's judgment about participation, particularly in circumstances where available alternatives are severely constrained.

The Belmont Report notes that inducements ordinarily considered acceptable may operate differently when a prospective participant is especially vulnerable. HHS guidance likewise advises ethics committees to consider incentives in relation to the circumstances of the participant population.

For researchers, this means that the amount alone does not answer the ethical question. The payment structure, study risks, local economic context, consequences of withdrawal, and relationship between financial need and possible undue influence may all matter.

Privacy protections may need to go beyond ordinary confidentiality procedures

For some participants, the principal danger is not a research procedure but what could happen if participation or collected information became known. Disclosure might expose someone to stigma, discrimination, family conflict, employment consequences, legal jeopardy, or other social harms.

Protection may therefore require reconsidering what data are collected in the first place. Researchers can ask whether direct identifiers are genuinely necessary, who requires access, how participants will be contacted, whether recruitment itself reveals sensitive information, how data will be stored or transferred, and whether quotations or combinations of demographic details could make someone recognizable in a publication.

These questions become especially consequential when stigma or legal status makes disclosure unusually risky. Merely replacing names with participant codes may not resolve every confidentiality problem.

The study design itself can be part of the vulnerability

It is tempting to think of vulnerability as something participants bring into a study. Sometimes the research creates or amplifies it.

A survey might unnecessarily request identifiable information about illegal behavior. Recruitment might occur in front of peers. A study involving employees might allow supervisors to know who declined. An interview could require participants to disclose traumatic experiences without a defensible research need. A mobile application might collect location information that is irrelevant to the research question.

In such cases, adding another paragraph to the consent form does little to address the underlying problem. Researchers should consider whether they can redesign the study so that it creates less vulnerability.

Additional safeguards can extend throughout the study

Protection is not confined to recruitment and consent. Depending on the research, safeguards may be needed during data collection, follow-up, analysis, publication, and withdrawal.

Examples include monitoring participants for distress or adverse events, establishing procedures for responding to unexpected problems, repeating consent discussions when circumstances change, limiting staff access to sensitive data, allowing participants to skip questions when methodologically and ethically appropriate, and planning carefully for incidental or sensitive findings.

The level of protection should be proportionate to the actual concern. A minimal-risk anonymous questionnaire and an invasive intervention involving participants with impaired decision-making capacity clearly do not require identical safeguards simply because both samples could be described as vulnerable.

Protection also includes fair access to research

There is another side to participant protection that can be easy to miss: exclusion can cause harm too.

The 2024 Declaration of Helsinki states that when individuals, groups, or communities in situations of vulnerability have distinctive health needs, excluding them from medical research can perpetuate or worsen disparities. It therefore calls for the harms of exclusion to be considered alongside the harms of inclusion.

This does not mean that every vulnerable population should be included in every study. Rather, researchers need an ethical justification for both inclusion and exclusion. Automatically removing a population because ethics review would be more complicated can leave the very people affected by a problem without evidence relevant to them.

The challenge is therefore not simply to protect vulnerable participants from research. It is also to determine when they should be fairly included in research with appropriate safeguards.

What tailored protection looks like in an actual study design

Notice what was not necessary merely because the employees occupied a potentially vulnerable position: declaring them incapable of consent, excluding them from the study, or automatically requiring someone else to decide for them. The protections addressed the mechanism of vulnerability rather than the label.

Common mistakes when trying to protect vulnerable participants

Design safeguards around the vulnerability, not around the label

When preparing a protocol, avoid treating “vulnerable participants” as a box that can be checked and forgotten. A more useful approach is to trace the ethical problem from its source to the protection intended to address it.

For some populations, these general principles operate alongside specific legal or regulatory requirements. For example, the U.S. Department of Health and Human Services regulations contain additional provisions for certain research involving pregnant women, human fetuses and neonates, prisoners, and children. Other jurisdictions and institutions may structure their requirements differently. Researchers should therefore verify the rules governing the particular study rather than assuming that one international guideline or one country's regulation applies everywhere.

Population-specific issues can also require more detailed treatment. Research involving children may involve additional protections such as parental permission and child assent, while research involving adults whose capacity is impaired raises distinct questions about assessment, authorization, and respect for the person's own wishes.

Before recruiting participants who may be vulnerable, check these safeguards

Questions researchers ask about vulnerable participants

Good protection is specific enough to solve the actual ethical problem

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.