03 · What You Need to Know
Understanding Why Participants Discontinue Research Participation
Participant Dropout Is Not a Single Event
The term dropout is frequently used to describe several forms of discontinuation, but these should be distinguished when planning and reporting research.
A participant may stop receiving an intervention while continuing to provide outcome data. Another may miss one scheduled assessment but return later. Someone else may withdraw consent for further participation.
These events have different implications for the study.
Study withdrawal
A participant discontinues participation according to the applicable consent and withdrawal procedures.
Loss to follow-up
Scheduled follow-up information cannot be obtained, potentially because the participant cannot be contacted or assessed.
Intervention discontinuation
A participant stops receiving or adhering to an assigned intervention but may continue outcome assessments.
Intermittent missingness
A participant misses one or more measurements but provides data at other time points.
Researchers should define these events in the protocol and avoid treating them as interchangeable.
The distinction is particularly important for longitudinal studies and randomized trials, where continuing outcome assessment after intervention discontinuation may remain scientifically valuable and ethically permissible.
Study Burden Can Increase the Difficulty of Remaining Enrolled
Participant burden refers to the demands that research participation places on individuals. These demands may involve time, travel, physical effort, emotional discomfort, financial costs, or cognitive workload.
A study requiring several lengthy questionnaires may be difficult for participants managing employment, caregiving, or academic responsibilities. Repeated clinic visits may impose substantial travel costs. Sensitive interviews may require emotional effort that participants did not fully anticipate.
Burden is not determined solely by the number of research activities. A short but invasive procedure may be more demanding than a longer, noninvasive assessment.
Likewise, a procedure that is manageable for one population may be difficult for another.
Researchers should evaluate burden from the participant's perspective rather than assuming that procedures are acceptable because they appear routine to the research team.
Reducing unnecessary measurements, simplifying instructions, and offering appropriate scheduling flexibility may help address avoidable demands.
Longer Follow-Up Creates More Opportunities for Discontinuation
As follow-up extends over months or years, participants have more opportunities to experience changes that interfere with participation.
They may relocate, change employment, graduate, develop health problems, or lose access to the communication channels used by researchers.
Longer studies may also require repeated decisions to continue participating.
However, follow-up duration should not be interpreted as an independent causal explanation in every study. Longer studies often differ from shorter studies in participant populations, assessment schedules, intervention demands, and retention procedures.
Researchers should examine these related factors rather than assuming that elapsed time alone determines dropout.
When estimating expected attrition, the relevant evidence should come from studies with reasonably comparable follow-up conditions. This is central to planning a defensible attrition allowance.
Competing Responsibilities Can Make Follow-Up Difficult
Participants may value the research but find it difficult to prioritize appointments or questionnaires alongside other responsibilities.
Work schedules, caregiving obligations, examinations, household responsibilities, and unexpected emergencies can interfere with participation.
These pressures may change during the study.
For example, university students recruited during an ordinary teaching period may encounter substantially different time constraints during examinations or internships.
Similarly, participants working irregular shifts may struggle to attend assessments scheduled exclusively during standard office hours.
Researchers should distinguish unwillingness to participate from an inability to meet the study's scheduling requirements.
Where scientifically and operationally appropriate, flexible assessment windows or alternative appointment arrangements may improve accessibility.
Financial and Transportation Barriers Can Affect Retention
Participation may involve direct and indirect costs.
Direct costs include transportation, mobile data, parking, or meals associated with research visits. Indirect costs may include lost wages, childcare arrangements, or time away from other responsibilities.
These costs can accumulate during repeated follow-up.
A participant who can afford one baseline visit may not be able to attend six subsequent appointments.
Reimbursement of reasonable participation expenses may reduce some barriers, subject to the protocol and applicable ethics requirements.
However, reimbursement and incentives serve different purposes. Compensation arrangements should be described accurately and should not create inappropriate pressure to continue participation.
Researchers should also avoid assuming that financial barriers affect every participant in the same way.
Accessibility and Digital Exclusion May Contribute to Dropout
Remote research can reduce travel requirements, but it may introduce other barriers.
Participants may experience unstable internet connections, limited device access, inaccessible questionnaires, unfamiliar software, or difficulties receiving authentication messages.
Digital participation may also become less feasible when individuals lose access to a shared device or change their mobile service.
Accessibility considerations extend beyond technology. Research materials may need appropriate language, readable formatting, assistive-technology compatibility, or alternative communication arrangements.
These barriers can be particularly consequential when follow-up depends on a single platform.
Researchers should consider whether participants can continue using the selected procedures throughout the study rather than evaluating accessibility only at enrollment.
Changes in Contact Information Can Interrupt Follow-Up
Participants may change telephone numbers, email addresses, residences, or institutions during the research period.
Some changes are predictable. Students may graduate or lose institutional email access. Employees may change workplaces. Patients may transfer care to another facility.
Other changes occur unexpectedly.
When follow-up depends on outdated contact information, researchers may classify participants as lost even when they remain willing to participate.
Appropriate contact verification, participant-authorized alternative communication methods, and clearly explained update procedures may reduce this risk.
These arrangements should be planned in accordance with confidentiality requirements and the approved protocol.
The challenge becomes especially important when participants' contact details or locations change during follow-up.
Intervention Experiences May Influence Continued Participation
In intervention studies, participants' experiences with the assigned condition may affect whether they continue.
Some may discontinue because of adverse effects, discomfort, inconvenience, disappointment, or perceived lack of benefit. Others may stop because their circumstances improve and they no longer consider participation necessary.
In educational research, students may disengage from an intervention that conflicts with their learning preferences or requires substantial additional effort.
In clinical research, treatment tolerability and disease progression may affect intervention adherence and follow-up availability.
These mechanisms can produce differential attrition across study groups.
Importantly, intervention discontinuation does not necessarily require withdrawal from outcome follow-up. Where consent and the protocol permit, continued measurement may remain essential to the study's intended comparison.
Researchers should avoid treating nonadherence, dissatisfaction, and complete study withdrawal as the same event.
Perceived Relevance and Participant Expectations May Change
Individuals may enroll because they consider the research meaningful, interesting, or potentially beneficial.
As the study progresses, their perceptions may change.
Some may discover that the procedures differ from what they expected. Others may feel that repeated assessments provide little personal value or that the research no longer relates to their circumstances.
These experiences can influence willingness to continue, although their effects vary across participants.
Recruitment materials should therefore communicate participation requirements accurately from the beginning.
Overly promotional descriptions can create expectations that the study cannot fulfill. Such concerns connect to expectations created by research advertisements.
Researchers should not promise individual benefits merely to encourage enrollment or retention.
Trust, Privacy, and Researcher Relationships Can Affect Retention
Participants may reconsider participation when they become concerned about confidentiality, the use of their information, or their relationship with the research team.
Unclear communication, unexpected procedural changes, or perceived disrespect may undermine confidence in the study.
Conversely, transparent communication and respectful interactions may help participants understand what participation involves.
Trust should not be interpreted as an obligation to remain enrolled. Participants retain the rights established by the applicable consent framework, including the right to discontinue participation when relevant.
Researchers should also avoid assuming that participants who withdraw because of privacy concerns are simply insufficiently committed.
Such concerns may indicate that study procedures or explanations require review.
Participant Characteristics May Predict Dropout Without Causing It
Studies sometimes identify associations between attrition and characteristics such as age, socioeconomic circumstances, health status, education, or baseline outcome measures.
These associations may be useful for planning, but they require careful interpretation.
For example, an association between lower income and missed follow-up visits might reflect transportation costs, inflexible work schedules, or other contextual barriers.
The demographic characteristic itself should not automatically be treated as the causal mechanism.
Furthermore, predictors identified in one study may not generalize to another population or research setting.
Researchers should avoid using demographic risk profiles to justify excluding individuals who may be difficult to retain.
Doing so could intensify selection bias and reduce equitable access to research participation.
Dropout May Be Related to the Outcome Being Studied
One of the most consequential concerns arises when the probability of missing follow-up data is associated with the outcome or factors related to it.
Consider a study measuring psychological well-being. Participants experiencing worsening difficulties may become less able or willing to complete follow-up questionnaires.
If their outcomes are disproportionately missing, the observed sample may present an overly favorable picture.
Alternatively, individuals who improve substantially may discontinue because they no longer perceive a need to participate.
The direction of bias therefore cannot be assumed from the presence of attrition alone.
Missing-data methods distinguish among mechanisms such as missing completely at random, missing at random conditional on observed information, and missing not at random.
These are assumptions about the missing-data process, not labels that can ordinarily be established from the observed data alone.
Understanding plausible dropout mechanisms is therefore important for both retention planning and the interpretation of missing outcomes.
Different Study Groups May Experience Different Dropout Risks
Comparative studies may experience differential attrition when participants in one group face greater burden, different expectations, or less tolerable procedures.
For example, an intervention group may require additional weekly activities while the comparison group completes only scheduled assessments.
The resulting difference in follow-up availability could reflect intervention burden rather than a general lack of participant motivation.
However, identical retention procedures do not necessarily produce identical retention rates. Participants may experience the same procedures differently.
Researchers should assess whether retention strategies should be consistent across study groups and how any justified differences affect interpretation.
For randomized trials, missing outcome data can threaten the credibility of treatment-effect estimates when missingness depends on outcomes or other relevant factors.
Excessive Follow-Up Contact Can Become a Retention Problem
Reminders and follow-up communication may help participants remember appointments and resolve practical difficulties.
Yet frequent or poorly timed contact can become intrusive.
Participants may disengage when they receive repeated messages, experience pressure to respond, or find that communication disregards their preferences.
The appropriate contact frequency depends on the study, participant population, follow-up requirements, and consent arrangements.
Researchers should establish reasonable communication procedures, provide ways to update contact preferences, and honor refusals or requests to stop contact.
Retention should not be pursued at the expense of participant autonomy.
How Should Researchers Identify Dropout Risk Factors?
A defensible assessment combines previous evidence with knowledge of the proposed study.
Researchers can review comparable longitudinal studies, examine pilot data, consult relevant participant representatives, and assess the demands of the planned protocol.
They should consider both factors that are difficult to modify and procedural barriers that may be reduced.
| Potential dropout factor |
What researchers should investigate |
Possible design response |
| Repeated assessments |
Whether frequency or duration creates unnecessary burden |
Review assessment schedules and remove nonessential procedures |
| Transportation difficulties |
Whether visits impose substantial travel costs or time |
Consider appropriate reimbursement or alternative assessment arrangements |
| Changing schedules |
Whether appointments conflict with work, study, or caregiving |
Offer feasible scheduling flexibility |
| Unstable contact details |
Whether participants can reliably receive follow-up messages |
Establish approved contact-update procedures |
| Digital access barriers |
Whether devices, connectivity, or accessibility affect participation |
Provide accessible and appropriate participation options |
| Intervention burden |
Whether procedures differ in tolerability or required effort |
Monitor discontinuation and preserve follow-up where appropriate |
| Privacy concerns |
Whether participants understand and trust data-handling procedures |
Improve explanations and review confidentiality safeguards |
These responses should be selected according to the study's scientific requirements and ethical approvals. Not every modification will be feasible or appropriate.
Monitor Dropout Without Assuming That Correlation Establishes Cause
During follow-up, researchers may examine retention by study group, site, assessment period, and relevant baseline characteristics.
Such monitoring can identify emerging patterns.
For example, lower retention among participants who work evening shifts may suggest scheduling difficulties, but the observed association does not establish the reason.
Where appropriate, researchers may ask participants about barriers to continued participation. Responses can inform procedural improvements, although individuals who cannot be contacted may remain absent from this information.
Researchers should document changes to retention procedures and obtain protocol amendments when required.
Any statistical investigation of dropout predictors should account for the study design, sample size, missing information, and the possibility of confounding.
Post hoc associations should not be presented as established causal explanations.
Retention Strategies Should Be Planned Before Dropout Occurs
Waiting until participants begin leaving can limit the available options.
Researchers should anticipate plausible barriers during protocol development and establish appropriate procedures for reminders, scheduling, participant communication, accessibility, and follow-up after intervention discontinuation.
They should also define how withdrawals, missed assessments, and unsuccessful contact attempts will be recorded.
These arrangements are part of developing a participant retention plan.
Planning does not guarantee that dropout will be prevented. It does, however, make the research team better prepared to respond consistently and ethically when difficulties arise.
Watch Out
Do not exclude participants merely because their circumstances suggest that follow-up may be difficult, unless an eligibility restriction is scientifically and ethically justified. Such exclusions may systematically remove populations whose experiences are especially relevant to the research.
