03 · What You Need to Know
Understanding the Relationship Between Participant Adherence and Intervention Fidelity
What Is Participant Adherence?
Participant adherence refers to the extent to which participants follow the intervention activities, participation requirements, or treatment recommendations specified for them.
In educational research, adherence may involve attending required sessions, completing assigned learning activities, or following prescribed practice procedures.
In clinical research, adherence may concern medication use, attendance at therapy sessions, or compliance with treatment recommendations.
In behavioral interventions, adherence may involve performing recommended activities or applying strategies between sessions.
The appropriate definition depends on the intervention.
Researchers should specify what participants are expected to do rather than use adherence as a general label for every form of engagement.
What Is Intervention Fidelity?
Intervention fidelity concerns whether an intervention is implemented according to its intended design.
In narrower usage, fidelity often emphasizes delivery by intervention providers, including adherence to prescribed content, dose, and quality requirements.
Broader treatment fidelity frameworks may include study design, provider training, delivery, participant receipt, and enactment.
The NIH Behavior Change Consortium framework, described by Bellg and colleagues, explicitly considers treatment receipt and enactment alongside delivery.
Consequently, researchers should state which framework they are using before deciding whether participant adherence is part of fidelity.
The terminology should not obscure the practical need to measure provider and participant processes separately.
Why Are Adherence and Fidelity Sometimes Treated as the Same Thing?
Both concepts concern whether an intervention was implemented or followed as intended.
However, they often refer to different actors and behaviors.
Providers may be expected to deliver specified components, while participants may be expected to attend, practice, or apply those components.
Some implementation frameworks use adherence to describe provider adherence to intervention content. In other contexts, adherence refers to participant behavior.
This variation in terminology can create confusion.
Researchers should identify whose adherence is being assessed and which intervention requirements provide the reference standard.
How Do Provider Fidelity and Participant Adherence Differ?
| Dimension |
Central question |
Illustrative indicator |
| Provider delivery fidelity |
Was the intervention delivered as intended? |
Required components delivered |
| Provider competence |
How well were procedures performed? |
Structured quality ratings |
| Participant exposure |
What intervention content was encountered? |
Sessions attended or materials accessed |
| Participant adherence |
Were prescribed activities followed? |
Required tasks completed |
| Participant engagement |
How did participants interact with the intervention? |
Behavioral participation and relevant engagement indicators |
| Treatment enactment |
Were intervention skills applied in intended settings? |
Observed application of learned strategies |
These dimensions may overlap, but they should not be assumed interchangeable. Each answers a different implementation question.
Can an Intervention Have High Fidelity but Low Participant Adherence?
Yes. Consider a program requiring instructors to deliver eight workshops and provide structured feedback.
All workshops may be delivered correctly, and feedback may meet the intended quality criteria.
Nevertheless, some participants may attend only a few sessions or fail to complete the assigned activities.
The providers' delivery fidelity may be high while participant adherence is low.
This distinction matters because the observed treatment effect may reflect both the intervention offered and the extent to which participants engaged with it.
Researchers should not automatically interpret low adherence as evidence that the intervention was poorly delivered.
Can Participant Adherence Be High When Provider Fidelity Is Low?
Yes. Participants may complete every activity they are given even when providers omit or modify essential intervention components.
For example, students may attend all sessions and complete every worksheet, but the instructor may fail to provide the diagnostic feedback required by the intervention protocol.
Participant adherence to the activities delivered may be high, while delivery fidelity is incomplete.
The distinction demonstrates why attendance or task completion alone cannot establish that the intended intervention was implemented.
Researchers should consider how intervention delivery fidelity is assessed independently of participant participation.
Is Participant Adherence the Same as Intervention Engagement?
Not necessarily. Adherence concerns following specified requirements, while engagement may describe behavioral, cognitive, or affective involvement with the intervention.
For example, a student may complete every required digital activity but engage only superficially with its content.
Another student may participate deeply in fewer activities.
Completion records can describe adherence but may not capture the quality of engagement.
Researchers should define engagement according to the theoretical construct and measurement approach used in the study.
They should avoid interpreting digital activity counts as direct evidence of cognitive engagement or learning.
What Is the Difference Between Treatment Receipt and Enactment?
Treatment receipt concerns whether participants receive, understand, or acquire the intervention content or skills, depending on the framework.
Enactment concerns whether participants apply those skills or behaviors in the intended setting.
For example, a student may understand a strategy for evaluating AI-generated research suggestions during a workshop.
That understanding does not establish that the student subsequently uses the strategy when preparing an independent research proposal.
Researchers may therefore need separate measures of receipt and enactment.
The NIH Behavior Change Consortium framework emphasizes these distinctions as part of comprehensive treatment fidelity assessment.
How Does Participant Adherence Relate to Intervention Dose?
Adherence and dose received often overlap but represent different concepts.
Dose received describes the amount of intervention exposure or participation.
Adherence evaluates whether participants followed the intervention requirements applicable to them.
For example, an adaptive program may assign four activities to one student and eight to another according to diagnostic performance.
Both students may be fully adherent if they complete their assigned activities, even though their received doses differ.
Researchers should therefore interpret intervention dose relative to the intended treatment strategy.
How Should Participant Adherence Be Measured?
Adherence measurement should correspond to the intervention activities participants are expected to perform.
Possible measures include session attendance, prescribed task completion, medication records, practice logs, and documented application of required procedures.
Researchers should distinguish objective records from self-reported behavior and consider the limitations of each.
For example, a digital platform may record that a student submitted an activity, but the submission does not necessarily establish that the activity was completed independently or with meaningful engagement.
Self-reports may capture activities occurring outside formal systems but can be affected by recall or reporting error.
Measurement should be proportionate and consistent with participant privacy.
Can Participant Adherence Be Expressed as a Percentage?
Yes, when adherence is defined through countable prescribed activities.
For example, researchers may calculate the proportion of required activities completed during a specified period.
The percentage describes completion of prescribed activities, not necessarily comprehension, engagement, or effectiveness.
Researchers should also avoid assuming that 80% represents an adequate adherence threshold without intervention-specific justification.
Does Every Intervention Have a Minimum Acceptable Adherence Rate?
No universal threshold applies across interventions.
The importance of adherence depends on the intervention's content, intended dose, timing, and causal mechanism.
For example, missing one essential safety-related procedure may be more consequential than missing several supplementary activities.
Researchers may establish operational thresholds for monitoring or defining protocol adherence, but those thresholds should have a defensible rationale.
They should not be presented as evidence that a particular level of adherence guarantees treatment effectiveness.
Why Might Participants Have Low Adherence?
Low adherence can arise from practical barriers, participant needs, intervention burden, perceived usefulness, accessibility limitations, competing responsibilities, or adverse experiences.
For example, students may struggle to complete an online program because of unreliable internet access or demanding academic schedules.
Others may find the activities insufficiently relevant or difficult to understand.
Researchers should avoid interpreting every instance of nonadherence as a lack of motivation or personal commitment.
Adherence may reflect the interaction between intervention design, delivery conditions, and participant circumstances.
Process evaluation can help investigate these mechanisms without assuming that one explanation applies to every participant.
Can Intervention Design Influence Participant Adherence?
Yes. The burden, accessibility, clarity, timing, and perceived relevance of intervention activities may influence participation.
For example, a program requiring lengthy activities during examination periods may experience lower completion than one integrated into ordinary course schedules.
However, observed differences in adherence across delivery formats do not automatically establish that the format caused those differences.
Researchers should consider intervention usability and feasibility during development and pilot testing.
Adherence should be treated as an implementation outcome worth understanding rather than merely a participant characteristic to be controlled statistically.
Should Researchers Exclude Nonadherent Participants?
Not automatically. In randomized trials, adherence occurs after treatment assignment and may be associated with participant characteristics or emerging outcomes.
Excluding nonadherent participants can undermine the comparability created by randomization.
For example, students who discontinue a learning program may differ from those who complete it in prior achievement, available time, or other factors related to outcomes.
A comparison restricted to adherent participants may therefore be biased.
The primary analysis should follow the prespecified estimand and appropriate analytical strategy.
Additional analyses estimating effects under different adherence conditions require suitable methods and assumptions.
Can Statistical Adjustment Correct for Nonadherence?
Ordinary adjustment for adherence does not automatically recover the causal effect of receiving the intervention.
Adherence may be influenced by treatment assignment, participant characteristics, previous outcomes, and changing circumstances.
Conditioning on adherence can therefore introduce selection bias or alter the estimand.
Depending on the research question, investigators may consider instrumental-variable approaches, causal methods for treatment switching, or other specialized analyses.
These approaches require assumptions that should be justified and assessed where possible.
Researchers should distinguish the effect of assignment from effects under hypothetical or specified adherence conditions.
How Does Adherence Affect Interpretation of a Null Result?
Low adherence may help explain why an intervention produced little observed improvement, particularly when substantial exposure is needed for the intended mechanism.
However, low adherence does not establish that the intervention would have worked under better participation.
The intervention may be ineffective even when fully implemented, or the study may have insufficient precision to detect an effect.
Researchers should consider provider fidelity, dose, adherence, outcome measurement, and contextual conditions together.
The distinction between intervention failure and implementation failure cannot be resolved from adherence data alone.
Can High Adherence Guarantee Intervention Effectiveness?
No. Participants may follow every prescribed activity without achieving the intended outcome.
For example, students may complete all required research writing exercises while showing little improvement in independent proposal quality.
High adherence establishes that prescribed activities were followed according to the measurement definition.
It does not establish that the activities were delivered competently, that participants understood them, or that the intervention's causal theory was correct.
Effectiveness requires appropriate outcome evidence and a credible research design.
Should Adherence Be Monitored During Implementation?
Monitoring may be useful when adherence problems can be addressed through procedures consistent with the intervention strategy.
For example, automated reminders or facilitator follow-up may help participants complete activities.
However, adherence support can itself become an intervention component.
If one group receives intensive encouragement or additional contact while another does not, those differences may affect the treatment contrast.
Researchers should specify adherence-support procedures and document important changes during implementation.
Monitoring should avoid unnecessary surveillance or pressure that compromises voluntary participation.
How Should Participant Adherence Be Reported?
Reports should define adherence operationally and identify the activities, periods, and measurement sources involved.
Researchers should describe the distribution of adherence rather than report only a single average where substantial variation exists.
For example, the proportion completing all prescribed activities may be informative alongside median completion or component-specific adherence.
Missing adherence records should be distinguished from confirmed noncompletion.
Researchers should also explain whether adherence was included within a broader fidelity framework or assessed as a separate implementation construct.
Transparent reporting helps readers understand both the intervention offered and the extent to which participants followed it.
