Can Excluding Vulnerable Participants From Research Also Be Unethical?

If a population is vulnerable, isn't excluding them the safest option?

Suppose including a population requires additional consent procedures, stronger confidentiality protections, accessible study materials, more complicated recruitment, closer monitoring, or additional regulatory review. Excluding that population can appear to solve several ethical problems at once.

Sometimes exclusion is justified. A study may expose particular participants to risks that cannot be reduced adequately, an intervention may be inappropriate for them, or the research question may genuinely concern another population.

But exclusion has consequences too. If children, older adults, people with disabilities, pregnant people, economically disadvantaged communities, migrants, prisoners, or people with impaired decision-making capacity are routinely removed whenever their participation complicates a protocol, researchers may eventually know the least about precisely the populations for whom evidence is hardest to generate.

The ethical question is therefore not simply whether inclusion carries risk. It is whether exclusion carries costs that also deserve justification.

Yes, exclusion can be unethical when it is unnecessary or unfair

Exclusion is a research decision with consequences, not the absence of one

Research ethics protects people from both exploitation and unfair exclusion

Much of human-subjects ethics understandably focuses on preventing people from being exposed to inappropriate risk. The Belmont Report's principle of justice adds another dimension: the burdens and benefits of research should be distributed fairly, and participant selection should not systematically burden some groups or favor others without justification.

This means that ethical participant selection cannot be reduced to identifying whom it would be safest to leave out.

A population may need additional protections during research without needing exclusion from research altogether.

Exclusion can create evidence that does not fit the people who will use it

Imagine that a treatment will routinely be prescribed to adults over 70, yet the pivotal research largely excludes them. Researchers may know considerably less about effectiveness, adverse effects, interactions, dosing, or practical use in a major portion of the eventual patient population.

The problem is not limited to age. Similar evidence gaps can arise when research excludes people with comorbidities, disabilities, pregnancy, impaired capacity, language differences, or other characteristics common in real-world populations.

FDA's December 2025 final guidance on enhancing participation in clinical trials recommends broader eligibility and enrollment approaches that can produce study populations more reflective of people likely to use a drug if approved. It specifically discusses demographic and non-demographic characteristics including age, location, organ dysfunction, comorbid conditions, disabilities, and other baseline characteristics.

Underrepresentation can become a safety and effectiveness problem

Representation is not merely about making a demographic table look balanced. The scientifically relevant question is whether characteristics that vary across the intended population could affect outcomes.

If an intervention behaves differently in a population that was systematically excluded, clinicians and patients may later make decisions with less direct evidence.

This is one reason NIH's Inclusion Across the Lifespan policy requires participants of all ages in NIH-supported human-subjects research unless there are scientific or ethical reasons for exclusion. NIH states that the purpose is to make knowledge applicable to people affected by the diseases or conditions being studied.

That policy applies to NIH-supported research rather than establishing a universal rule for every study, but its rationale illustrates the scientific consequences of unnecessary exclusion.

Protection by exclusion can become circular

There is a subtle problem when lack of evidence becomes the reason for continued exclusion.

This does not mean uncertainty should be ignored. It means researchers should recognize when past exclusion has helped create the uncertainty now being used to justify future exclusion.

Exclusion can deny access to potential direct benefits

Some research offers no prospect of direct benefit, while other studies may provide access to potentially beneficial interventions, additional clinical monitoring, or other research-related opportunities.

No participant has an automatic entitlement to enrollment in every study. Eligibility must remain scientifically and ethically defensible. But systematically excluding a population can also systematically deny that population opportunities available to others.

This concern becomes particularly important when the excluded population experiences the condition under investigation and the study offers a prospect of direct benefit.

Exclusion can be unfair even when researchers have protective intentions

Researchers may exclude a population because obtaining appropriate consent is difficult, translation is expensive, accessible facilities are unavailable, recruitment takes longer, or additional ethics review is required.

Those are real operational challenges. They are not automatically ethical justifications.

NIH's current policy on women and members of racial and ethnic minority groups requires inclusion in NIH-funded clinical research unless a clear and compelling rationale establishes that inclusion is inappropriate with respect to participant health or the purpose of the research. The policy specifically states that cost is generally not an acceptable reason for exclusion.

Again, this is a specific funding policy rather than a universal standard. The broader lesson is useful: inconvenience to the research enterprise should not quietly masquerade as participant protection.

Exclusion criteria should correspond to the actual concern

A broad exclusion criterion often uses an easily measured characteristic as a proxy for something else.

Sometimes the answer will still support exclusion. The ethical improvement lies in making the exclusion correspond to the real risk rather than an imprecise population label.

Some exclusions are clearly justified

An argument for inclusion should not become an assumption that every population belongs in every study.

Exclusion may be appropriate when an intervention is physiologically inappropriate for a population, the study question genuinely concerns a defined group, risks cannot be reduced adequately, legally required protections cannot be satisfied, participation would interfere with necessary treatment, or including a population would not contribute scientifically meaningful information.

Protection remains a legitimate reason for exclusion. The requirement is justification, not inclusion at any cost.

Vulnerability can sometimes be reduced without exclusion

Before removing a population, researchers can ask whether the relevant concern is modifiable.

Can recruitment be separated from an authority figure? Can sensitive identifiers be removed? Can consent information be translated? Can transportation requirements be reduced? Can capacity be assessed individually? Can an accessible alternative be provided? Can additional monitoring reduce risk?

These are applications of the principle that research design can create or reduce vulnerability.

Inclusion itself must still be fair

There is an important counterpoint. Researchers should not recruit a population merely to satisfy a representation target when the study is irrelevant to them, participation exposes them to unnecessary burden, or the design cannot answer meaningful questions about that population.

Ethical inclusion is not demographic decoration. It should follow the scientific purpose of the study and a fair distribution of research burdens and potential benefits.

The difficult question is therefore not “How diverse is the sample?” in isolation. It is “Who needs to be represented for this research to answer its question fairly and usefully?”

An exclusion criterion can look protective until you ask what it is protecting against

The lesson is not that everyone must be included. It is that the reason for exclusion should survive the question, “Why this person, rather than the actual risk we are concerned about?”

Exclusion can look ethically conservative while quietly creating another form of unfairness

Treat every exclusion criterion as a claim that should be defensible

This leads directly to the more difficult boundary between legitimate protection and unfair exclusion from research. The answer rarely comes from asking whether inclusion is perfectly safe. Research seldom offers that luxury. The question is whether the remaining risk and the consequences of exclusion have both been considered honestly.

Before excluding a population, make the reason explicit

Frequently asked questions about excluding vulnerable participants

Keeping someone out of research is also an ethical decision

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.