03 · What You Need to Know
Making Consistent and Defensible Decisions About Borderline Eligibility
What Is Borderline Eligibility?
Borderline eligibility occurs when a prospective participant's characteristics are close to an eligibility boundary or cannot be classified confidently using the available information.
For example, a screening score may fall near a cutoff, a clinical classification may be uncertain, or an experience requirement may be open to interpretation.
However, being close to a threshold is not the same as being uncertain about eligibility.
If a protocol validly requires a score below 40 and the measurement is sufficiently reliable, a score of 40 does not satisfy the criterion.
Uncertainty arises when the measurement, definition, or available evidence does not support a confident decision.
Researchers should distinguish borderline values from genuinely ambiguous eligibility determinations.
Why Is Borderline Eligibility a Methodological Problem?
Eligibility determines which individuals can contribute evidence to the study.
Inconsistent decisions may cause participants with similar characteristics to be treated differently.
For example, one recruiter might admit students scoring slightly above a cutoff while another applies the threshold strictly.
Such differences can change the study population and create site-specific or assessor-specific selection patterns.
When decisions depend on perceived motivation, expected performance, or recruitment needs, the resulting selection may also introduce bias.
Consistent eligibility procedures are therefore important for methodological transparency and fair participant selection.
What Types of Borderline Eligibility Can Occur?
| Situation |
Example |
Appropriate consideration |
| Near-threshold measurement |
A digital literacy score is close to the eligibility cutoff. |
Measurement precision and prespecified reassessment rules |
| Conflicting evidence |
Self-report differs from an administrative record. |
Source definitions, timing, and verification |
| Incomplete documentation |
A diagnosis date or enrollment record is missing. |
Additional verification and unresolved eligibility procedures |
| Ambiguous criterion |
The protocol requires substantial professional experience. |
Operational definition and consistent interpretation |
| Changing status |
A student transfers between programs during recruitment. |
Eligibility reference date and timing rules |
| Clinical uncertainty |
Assessments provide conflicting severity classifications. |
Appropriate clinical review or adjudication |
| Mixed eligibility characteristics |
A participant satisfies most but not all required conditions. |
Whether every criterion is mandatory under the protocol |
These situations require different responses. Researchers should identify the source of uncertainty before deciding whether reassessment or clarification is justified.
Should Someone Be Included Because They Almost Meet a Criterion?
Not automatically. If a criterion is valid, clearly defined, and measured appropriately, being close to its boundary does not establish eligibility.
For example, if the protocol specifies a maximum age or a score below a particular threshold, the rule should be applied as written.
However, researchers should consider whether the threshold itself is scientifically justified and whether the measurement is sufficiently precise for classification.
If the criterion is arbitrary or poorly defined, the appropriate response is to review the protocol rather than grant informal exceptions to selected participants.
Consistency should not be confused with the assumption that every threshold is scientifically meaningful.
What If the Cutoff Is Arbitrary?
An arbitrary cutoff may create an artificial distinction between individuals with very similar underlying characteristics.
For example, defining low digital literacy using an unsupported score threshold may exclude individuals whose competence is practically indistinguishable from those admitted.
Researchers should evaluate whether the cutoff is supported by validation evidence, established standards, intervention requirements, or another defensible rationale.
If the threshold lacks justification, it may need revision before recruitment or through an appropriate protocol amendment.
Researchers should not change it selectively for individual participants after observing their scores.
A justified threshold can still create borderline cases because measurement is rarely perfectly precise.
How Does Measurement Error Affect Borderline Eligibility?
Observed measurements contain varying degrees of error.
For example, a digital literacy assessment score may vary because of temporary distraction, testing conditions, item sampling, or other measurement factors.
Consequently, a person whose true competence lies near a cutoff may receive scores on either side across repeated assessments.
Researchers should examine the reliability and precision of the instrument when interpreting near-threshold results.
However, measurement uncertainty does not automatically justify admitting everyone close to the cutoff.
The protocol should specify how uncertainty will be handled consistently.
Should Borderline Participants Be Retested?
Retesting may be appropriate when the original measurement is uncertain and a repeat assessment can provide useful additional evidence.
For example, a laboratory result affected by a suspected procedural error may warrant confirmation according to the relevant clinical protocol.
Similarly, an educational assessment interrupted by a documented technical failure may reasonably be repeated.
However, repeated testing can introduce new problems if participants are tested until they obtain an eligible score.
Researchers should specify the circumstances permitting reassessment, the number and timing of tests, and how multiple results will be interpreted.
Retesting should resolve uncertainty rather than function as an informal mechanism for bypassing eligibility criteria.
Can Repeated Screening Create Selection Bias?
Yes, depending on how repeated assessments are used.
For example, allowing only preferred participants to repeat an eligibility test may create inconsistent opportunities for enrollment.
Similarly, repeatedly testing individuals until they cross a cutoff can select favorable measurement fluctuations rather than the intended underlying characteristic.
Such procedures may distort the distribution of baseline characteristics among enrolled participants.
Researchers should establish uniform reassessment rules and document all relevant attempts.
Where repeated measurement is necessary, the classification procedure should be justified by the instrument's properties and the research question.
How Should Conflicting Eligibility Evidence Be Resolved?
Conflicting evidence should be examined for differences in timing, definitions, accuracy, and completeness.
For example, a student may report being enrolled while an institutional register has not yet reflected a recent registration change.
Both statements may be understandable if they refer to different administrative stages.
Researchers should identify the reference date and source required by the criterion.
Where appropriate, additional verification may resolve the disagreement.
The decision process should be established before recruiters encounter difficult cases, particularly when conflicting sources are likely.
Should Self-Report Override Official Records?
Neither source should automatically override the other in every study.
The appropriate source depends on the characteristic being assessed and the operational definition.
For example, official enrollment status may be best established through an authorized institutional register.
However, a participant's firsthand experience of using generative AI may not be adequately captured by administrative records.
Researchers should evaluate which eligibility assessment method provides appropriate evidence rather than relying on a universal hierarchy of sources.
What Is Eligibility Adjudication?
Eligibility adjudication is a structured process for resolving uncertain or disputed eligibility decisions.
It may involve a designated investigator, qualified clinician, independent reviewer, or adjudication committee, depending on the study.
For example, conflicting clinical assessments may be reviewed against prespecified diagnostic criteria.
In qualitative research, ambiguous experience requirements may be evaluated using a documented screening framework.
The adjudication process should identify who makes the decision, what evidence is considered, and how disagreement is resolved.
Where feasible and relevant, adjudicators may be kept unaware of information that could improperly influence eligibility decisions, such as anticipated outcomes or treatment assignment.
Should the Principal Investigator Decide Every Borderline Case?
Not necessarily. Principal investigators are responsible for appropriate study conduct, but unrestricted individual discretion may produce inconsistent decisions.
Routine cases should be resolved through established screening procedures.
Complex cases may require specialist review or a designated adjudication process.
For clinical eligibility, the responsible decision-maker should have appropriate professional qualifications.
Researchers should avoid relying on informal personal judgments when objective criteria or structured review can provide a more consistent basis.
The decision-making authority should be defined in the protocol or relevant study procedures.
What If a Participant Meets Most but Not All Criteria?
Eligibility generally requires satisfying all mandatory inclusion criteria and none of the applicable exclusion criteria.
For example, a participant who satisfies four of five required inclusion conditions is not automatically eligible because most conditions are met.
However, researchers should examine whether every criterion is genuinely necessary.
If one requirement is unjustified, it may be appropriate to revise the eligibility framework through a documented protocol process.
The researcher should not selectively waive a criterion for an individual participant merely because that person appears suitable.
Any legitimate exceptions should be explicitly permitted by the approved protocol and applied consistently.
How Should Changing Eligibility Status Be Handled?
Some eligibility characteristics change over time.
For example, a student may complete enrollment after initial screening, or a patient's disease severity may change before intervention initiation.
Researchers should specify the relevant eligibility reference point and whether reassessment is required.
A participant who becomes eligible before enrollment may be treated differently from someone who becomes ineligible after randomization.
These situations should not be conflated.
Baseline eligibility, continued intervention suitability, withdrawal, and analytical inclusion are distinct methodological decisions.
Can Borderline Eligibility Affect Randomized Trials?
Yes. Eligibility should generally be established before randomization according to the approved protocol.
Admitting participants through inconsistent screening can alter the randomized population and complicate interpretation.
More seriously, changing eligibility decisions after treatment assignment is known may undermine allocation-related comparability.
Researchers should distinguish genuine pre-randomization screening errors from post-randomization events.
Participants should not automatically be removed from an assignment-based analysis merely because eligibility concerns are discovered later.
The appropriate analytical handling depends on the estimand, protocol, and relevant reporting requirements.
How Does Borderline Eligibility Affect Observational Studies?
In observational studies, inconsistent eligibility assessment can alter the source population and create differences between exposure groups.
For example, patients receiving one treatment may undergo more extensive diagnostic testing than patients receiving another.
If eligibility depends on those tests, the groups may be selected through different mechanisms.
Researchers should evaluate whether eligibility information is measured comparably across exposure groups.
Differences in screening intensity, documentation, or clinical assessment may affect the validity of the intended comparison.
Standardized definitions and sensitivity analyses may help assess these concerns.
How Should Borderline Eligibility Be Handled in Qualitative Research?
Qualitative research may involve eligibility criteria based on experiences, roles, or contextual relevance rather than numerical thresholds.
For example, a study of researchers experienced in AI-assisted systematic reviews may encounter individuals who have used AI for literature searching but not for formal review synthesis.
Whether these individuals qualify depends on how the relevant experience was defined.
Researchers should establish clear conceptual boundaries while recognizing that experience may not fit simple categories.
Where the study uses purposive sampling, borderline cases may be considered for their relevance to the research question, provided the eligibility framework permits such decisions.
Flexibility should be methodologically justified and documented rather than used selectively to recruit preferred participants.
Can Borderline Cases Be Included as a Separate Group?
Sometimes, if the research question and design support such an approach.
For example, a study investigating classification uncertainty may deliberately include individuals whose measurements fall near a diagnostic threshold.
However, creating a borderline group after examining outcomes can introduce concerns about selective analysis.
Researchers should define the group, its purpose, and analytical treatment in advance where feasible.
Borderline participants should not be added informally to an otherwise restricted population merely to increase sample size.
Should a Gray Zone Be Used Around Eligibility Cutoffs?
A gray zone may be useful when measurement uncertainty makes a sharp classification insufficiently reliable.
For example, a protocol might identify a narrow range around a test cutoff that requires additional assessment.
However, the width of the gray zone should be justified using measurement precision, clinical evidence, or other relevant information.
It should not be selected simply to admit more participants.
The protocol should specify whether gray-zone results lead to retesting, adjudication, temporary deferral, or another decision.
A gray zone is a structured uncertainty-management procedure, not permission to disregard eligibility criteria.
How Can Borderline Eligibility Affect Generalizability?
Strict thresholds can create a study population that excludes individuals with characteristics close to those of enrolled participants.
For example, an intervention evaluated only among students scoring below a digital literacy cutoff may not directly establish effectiveness among students just above it.
However, this does not necessarily make the restriction inappropriate.
The issue is whether the threshold defines a scientifically meaningful population and whether the intended application includes people outside it.
Researchers should avoid treating small measurement differences as proof of substantial differences in intervention response.
Broader applicability requires evidence or assumptions about how outcomes and effects vary across the eligibility boundary.
Can Borderline Eligibility Create Inequitable Selection?
Yes, particularly when uncertain eligibility is resolved differently across participant groups.
For example, individuals with incomplete records, limited language proficiency, or less access to diagnostic testing may be more likely to remain unclassified.
Automatically excluding these individuals can disproportionately affect groups already facing participation barriers.
Researchers should consider accessible verification procedures and consistent opportunities to provide missing information.
However, equitable selection does not justify ignoring genuine safety requirements or mandatory eligibility conditions.
The objective is fair application of scientifically defensible criteria.
What Should Be Done When Recruitment Is Slower Than Expected?
Recruitment pressure should not determine individual eligibility decisions.
For example, admitting students whose screening scores exceed the cutoff simply because enrollment targets have not been met changes the study population.
Researchers should first assess recruitment procedures, access barriers, and whether the original criteria remain scientifically justified.
If substantive changes are necessary, they should be considered through a documented protocol amendment and applicable ethics or registration procedures.
Any revision should be applied prospectively and consistently according to the approved design.
Borderline eligibility should not become an informal mechanism for solving recruitment difficulties.
What If a Participant Was Incorrectly Enrolled?
Discovering an eligibility error after enrollment requires a different response from resolving uncertainty before enrollment.
Researchers should document the error, assess participant safety, and follow the protocol's procedures for deviations or violations.
Where relevant, the research ethics committee or other oversight body may need to be informed.
However, an eligibility error does not automatically justify deleting the participant's data from every analysis.
The appropriate analytical treatment depends on the estimand, study design, timing of discovery, and prespecified analysis plan.
Researchers should distinguish participant protection from decisions about retaining observations for valid inference.
How Should Borderline Eligibility Decisions Be Documented?
Screening records should identify the relevant criterion, available evidence, assessment date, and final decision.
For uncertain cases, document the reason for uncertainty, additional assessments performed, and the applicable decision rule.
Where adjudication occurs, identify the authorized reviewer and record the basis of the determination.
Researchers should preserve an appropriate audit trail while protecting confidential information.
Documentation enables consistency checks and helps distinguish legitimate verification from arbitrary exceptions.
How Can Researchers Prevent Borderline Eligibility Problems?
Prevention begins with clear operational definitions and appropriate assessment methods.
Researchers should pilot screening procedures to identify ambiguous wording, likely near-threshold cases, and inconsistent interpretations.
Where numerical cutoffs are used, consider measurement precision and establish reassessment rules.
Where judgment is necessary, provide structured guidance and an adjudication process.
Recruiters should be trained to distinguish unknown eligibility from confirmed ineligibility.
Most importantly, decisions should be made without reference to anticipated participant outcomes or pressure to achieve a desired sample size.
