03 · What You Need to Know
How to Decide Whether Safety Risks Require Exclusion or Additional Safeguards
Why Is Participant Safety an Eligibility Consideration?
Research procedures can expose participants to physical, psychological, social, informational, or other risks.
These risks may arise from interventions, assessments, data collection procedures, disclosure of sensitive information, or the circumstances of participation.
For example, an immersive learning intervention may produce discomfort for some users, while a study collecting sensitive educational records may create privacy or confidentiality risks.
Eligibility criteria may be necessary when particular participant characteristics make a study procedure inappropriate or unacceptably hazardous.
However, safety is also addressed through study design, consent procedures, monitoring, confidentiality protections, and appropriate support.
Exclusion should therefore be considered within a broader risk-management framework rather than treated as the default response.
What Is the Difference Between a Safety Concern and a Contraindication?
A safety concern is a circumstance suggesting that participation may involve an increased or uncertain risk.
A contraindication is a condition under which a particular intervention or procedure should not be used because of an established or sufficiently supported risk.
For example, a known serious interaction between an investigational medication and another treatment may constitute a contraindication under the applicable protocol.
By contrast, a participant's concern about mild discomfort during a noninvasive procedure may call for additional explanation, monitoring, or modification rather than automatic exclusion.
The distinction depends on the intervention and available evidence.
Researchers should not use contraindication as a generic label for any characteristic that makes participation more complicated.
Which Safety Concerns May Require Different Responses?
| Safety concern |
Possible response |
When exclusion may be necessary |
| Known serious contraindication |
Verify the condition and apply the approved safety criteria. |
When the intervention presents an unacceptable risk. |
| Potentially manageable physical discomfort |
Adjust exposure duration, monitor symptoms, or provide stopping procedures. |
When risk remains unacceptable despite mitigation. |
| Psychological distress |
Use appropriate screening, support, monitoring, and referral arrangements. |
When procedures cannot be conducted within acceptable risk limits. |
| Accessibility requirements |
Provide appropriate accommodations or adapted procedures. |
When essential procedures remain incompatible despite reasonable adaptations. |
| Privacy or confidentiality risks |
Minimize data collection and strengthen access, storage, and disclosure controls. |
When the risk cannot be ethically managed within the study. |
| Uncertain clinical risk |
Seek appropriate specialist assessment, evidence review, and ethics guidance. |
When uncertainty and potential harm make participation unjustifiable. |
| Limited decision-making capacity |
Evaluate approved permission, assent, and additional safeguards. |
When applicable ethical or legal requirements cannot be satisfied. |
These responses are illustrative. The appropriate decision depends on the specific protocol, participant circumstances, and applicable ethical and regulatory requirements.
How Should Researchers Assess Risk Before Deciding on Exclusion?
Risk assessment should consider the nature of potential harm, its likelihood, its possible severity, and the extent to which the risk can be prevented or managed.
Researchers should also consider the expected benefits, scientific value, available alternatives, and the consequences of excluding the affected population.
For example, a study involving a potentially hazardous clinical intervention requires a different assessment from an anonymous questionnaire about routine learning practices.
The same participant characteristic may create substantial risk in one study and little additional risk in another.
Risk should therefore be evaluated in relation to actual procedures rather than inferred solely from a diagnostic label, demographic category, or disability status.
What Does Proportionality Mean in Safety-Based Eligibility?
Proportionality means that the restriction should correspond to the seriousness of the risk it is intended to address.
For example, excluding every participant who has ever experienced mild motion sickness from a virtual reality study may be unnecessarily restrictive if the protocol permits brief exposure, symptom monitoring, and immediate withdrawal.
However, individuals with specific conditions associated with serious adverse responses to the equipment may require exclusion when adequate safeguards are unavailable.
Researchers should avoid using broad categories when more specific criteria can identify the actual risk.
The appropriate restriction should be no broader than necessary to protect participants within the approved study design.
When Are Safeguards Preferable to Exclusion?
Safeguards may be preferable when they can reduce risk to an ethically acceptable level while preserving meaningful participation.
Possible safeguards include modified procedures, shorter exposure periods, monitoring, trained supervision, accessible interfaces, additional consent support, or appropriate referral arrangements.
For example, an educational technology study may offer alternative input methods to participants who cannot use a standard mouse or keyboard.
Such adaptations may improve inclusion without changing the central intervention or outcome.
However, safeguards must be evaluated for effectiveness. A proposed accommodation should not be assumed sufficient merely because it is available.
Can Participant Monitoring Replace Exclusion?
Monitoring may reduce risk when early signs of harm can be identified and appropriate responses are available.
For example, a study involving an immersive learning environment might establish procedures for identifying discomfort and stopping participation when symptoms occur.
In clinical research, monitoring may include specified assessments, adverse-event reporting, or predefined stopping rules.
However, monitoring cannot always prevent serious or rapidly occurring harm.
Where an intervention poses an unacceptable risk despite monitoring, exclusion or redesign may remain necessary.
Researchers should distinguish risks that can be managed through observation from risks requiring prevention of exposure.
Should Disability Automatically Lead to Exclusion?
No. Disability alone does not establish that participation is unsafe or scientifically inappropriate.
Researchers should identify whether a particular procedure creates a specific risk or accessibility barrier.
For example, a digital learning platform may be inaccessible to screen-reader users because of its interface design.
That problem may require interface modification or an alternative accessible procedure rather than excluding participants who use assistive technology.
However, adaptations should preserve the intended measurement or be explicitly accounted for in the design.
Broad disability-based exclusions can reduce equitable participation and limit the relevance of findings to the population in which the technology will be used.
Should Pregnancy Be an Automatic Exclusion Criterion?
No universal rule applies across all research.
Pregnancy may be relevant when an intervention presents established or uncertain maternal or fetal risks.
However, automatic exclusion from unrelated or low-risk research may be unjustified.
Researchers should evaluate the intervention, available safety evidence, potential consequences, applicable regulations, and ethics approval.
Where pregnancy-related exclusion is necessary, the criterion should be specific and supported by a documented rationale.
Ethical inclusion requires avoiding both unjustified exposure to harm and unnecessary exclusion from research participation.
Should Older Adults or Children Be Excluded for Safety?
Age may influence intervention risks, physiological responses, consent requirements, or developmental suitability.
However, age alone should not automatically determine exclusion from every study.
For example, a study concerning digital learning among university students may legitimately include older adult learners.
Research involving children may require parental or guardian permission, assent, and additional safeguards under applicable requirements.
Researchers should distinguish age-related scientific or safety considerations from assumptions that particular age groups are inherently unsuitable.
Where age restrictions are imposed, they should reflect the intended population and actual study requirements.
What About Participants With Limited Decision-Making Capacity?
Limited decision-making capacity may require additional protections, but it does not automatically make every individual ineligible for all research.
Depending on the jurisdiction and study, legally authorized permission, participant assent, supported decision-making, or other approved safeguards may be relevant.
Researchers should assess whether participation is ethically permissible and whether the study's procedures are appropriate.
When the necessary protections cannot be established, participation may be inappropriate.
Eligibility should follow applicable ethical and legal requirements rather than a blanket assumption that only independently consenting adults can participate.
Can Psychological or Emotional Risks Justify Exclusion?
Yes, when research procedures pose a substantial risk of distress or other harm that cannot be adequately managed.
For example, interviews concerning traumatic experiences may require carefully designed consent, interviewer training, participant support, and referral procedures.
However, excluding everyone who has experienced psychological distress may remove precisely the people whose perspectives are central to the research.
Researchers should assess the actual procedures, available support, and potential risks rather than relying on broad assumptions about mental health or vulnerability.
Where specific exclusions are necessary, their rationale should be documented and reviewed appropriately.
Are Privacy and Confidentiality Risks Grounds for Exclusion?
Privacy risks should generally prompt appropriate data protection measures before exclusion is considered.
For example, a study collecting sensitive educational records may reduce risk through data minimization, pseudonymization, restricted access, secure storage, and controlled disclosure.
However, some data cannot be collected or used without appropriate authorization or lawful grounds.
When necessary permissions or safeguards cannot be established, participation or data inclusion may be impermissible.
Researchers should distinguish risks arising from participant characteristics from risks created by the study's information-handling procedures.
Data protection should be built into the protocol rather than addressed primarily through excluding individuals with sensitive information.
How Does the Belmont Report Inform Safety-Based Exclusion?
The Belmont Report identifies respect for persons, beneficence, and justice as foundational ethical principles for research involving human subjects.
Beneficence requires attention to potential benefits and harms, while justice concerns fair selection and distribution of research burdens and benefits.
These principles can create competing considerations when researchers contemplate exclusion.
Protecting participants from unacceptable risks may require restrictions, but unnecessary exclusion can undermine equitable access and produce evidence that does not adequately represent relevant populations.
The appropriate response is a reasoned assessment rather than assuming that maximal exclusion is always the most ethical option.
What Is the Role of an Institutional Review Board or Research Ethics Committee?
Research ethics committees or institutional review boards evaluate ethical aspects of proposed studies according to their applicable mandates and regulatory frameworks.
They may examine risk-benefit considerations, participant selection, consent arrangements, safeguards, and the justification for excluding particular groups.
Researchers should present a clear rationale for safety-based eligibility restrictions and proposed alternatives.
Approval does not eliminate the research team's ongoing responsibility to monitor risks and follow the approved protocol.
Substantive changes to eligibility or safety procedures may require review or amendment before implementation.
Can Safety Exclusions Introduce Selection Bias?
Yes, depending on the research question and selection mechanism.
For example, a clinical intervention study that excludes patients with common comorbidities may produce a population that differs from patients who would receive the intervention in routine practice.
If those comorbidities influence treatment effects, the study's findings may not generalize directly to the broader clinical population.
However, exclusion does not automatically create bias in the estimate for the properly defined eligible population.
Researchers should distinguish the validity of inference within the eligible study population from applicability to excluded groups.
Can Safety Exclusions Change the Estimand?
Yes. An estimand specifies the quantity the study seeks to estimate, including the relevant population.
If a study initially targets all eligible patients with a condition but subsequently excludes those with specified comorbidities, the treatment-effect population changes.
Similarly, excluding students who require accessibility support from an educational intervention changes the population represented by the findings.
Such changes may be justified, but they should be recognized as substantive methodological decisions.
Researchers should not describe effects estimated in a restricted population as though they necessarily apply to the original broader population.
Should Participants Be Excluded After an Adverse Event?
An adverse event may require stopping an intervention, withdrawing a participant from particular procedures, or implementing additional safety measures.
However, discontinuing treatment is not necessarily the same as removing the participant from the study's analysis population.
For example, in a randomized trial, participants who discontinue treatment because of adverse events may remain relevant to an assignment-based treatment-effect estimand.
Researchers should distinguish treatment discontinuation, study withdrawal, outcome follow-up, and analytical exclusion.
The appropriate handling depends on the estimand, consent, protocol, and ethical requirements.
Automatically deleting participants with adverse events from the analysis could distort the evaluation of intervention benefits and harms.
How Should Safety Criteria Be Operationalized?
Safety-related exclusion criteria should specify the relevant condition, assessment method, timing, and decision process.
For example, "participants with serious health problems will be excluded" is too vague for a protocol requiring consistent clinical screening.
A more defensible criterion would identify the specific contraindication or risk condition using an appropriate clinical or procedural definition.
Where professional judgment is necessary, the protocol should describe the responsible assessor and relevant standards.
Uncertain or borderline cases should be handled through prespecified procedures rather than arbitrary decisions by individual recruiters.
How Can Researchers Audit Proposed Safety Exclusions?
For each proposed criterion, identify the risk it is intended to prevent and the evidence supporting that concern.
Then determine whether the risk can be reduced through safeguards, alternative procedures, or monitoring.
Consider whether the criterion excludes a broad category when a more specific restriction would be sufficient.
Finally, assess the consequences for equitable participation and the population to which findings will apply.
This audit should occur before recruitment and be revisited when new safety evidence becomes available.
Researchers should document the rationale and relevant ethics review decisions.
How Should Safety-Based Exclusion Criteria Be Reported?
Methods sections should identify safety-related exclusions and explain why they were necessary.
Describe the assessment procedures, relevant thresholds, and safeguards considered or implemented.
Where appropriate, report the number of potential participants excluded for safety reasons without disclosing identifiable or sensitive information.
Distinguish baseline safety eligibility from adverse events and treatment discontinuation after enrollment.
In the discussion, explain how safety restrictions affect the applicability of findings to excluded populations.
Transparent reporting allows readers to evaluate whether exclusions were proportionate and scientifically justified.
