Attrition And Dropout Monitoring¶
Track who leaves a study, when they leave, why they leave, and from which condition so dropout cannot silently distort causal or comparative conclusions.
The Diagnostic Story¶
Symptom: One arm or condition is losing participants faster than another, and the people leaving differ from completers on baseline characteristics. The final analysis denominator is smaller and less representative than the enrollment denominator, but the plan for handling missing outcomes was never specified before data collection ended.
Pivot: Instrument the study so every loss event is recorded, classified, and compared across conditions from the start, with pre-specified responses that may include retention outreach, burden reduction, or missing-data sensitivity analysis tied to observed attrition evidence.
Resolution: Differential dropout is detected early enough to respond, the analysis denominator remains transparent and traceable from eligibility through final reporting, and missingness assumptions are stated as assumptions rather than disguised as observed facts.
Reach for this when you hear…¶
[clinical trial management] “We're at week six and the dropout rate in the intervention arm is already twice what we planned for — if we don't investigate why now, the primary analysis will be uninterpretable.”
[education research] “The students who stuck with the tutoring program showed big gains, but we have no idea what happened to the half who dropped out, which means we can't claim the program works for the population we started with.”
[survey methodology] “Non-response in this income bracket is systematic, not random, so any analysis that ignores who didn't return the survey is going to overestimate the effect.”
When This Archetype Applies¶
No catalog groundingNone of the structural conditions is currently represented by an accepted prime or domain-specific abstraction.
Diagnostic problem
A comparison is designed around an eligible or assigned population, but some units stop contributing data before the endpoint. When dropout depends on treatment condition, baseline characteristics, burden, adverse experience, or emerging outcome status, the analyzed sample is no longer exchangeable with the intended comparison frame.
Show the applicability expression
Applicability expression3 distinct conditions
groundedpartly groundedopen
3 conditions, all required.
3Required in every casenumbered 1–3
These hold no matter which pattern applies.
Completer-only outcomes · open
Final outcomes are observed only for completers or for units still reachable at endpoint.
A comparison is designed around an eligible or assigned population, but some units stop contributing data before the endpoint. The narrower requirement in this condition set is: Final outcomes are observed only for completers or for units still reachable at endpoint.
Excess differential dropout · 2 cases · 0 matched
Preliminary dropout rates exceed planned assumptions or begin to diverge between comparison groups.
Experimental design tries to hold comparison groups stable, but participation is itself a dynamic response to the intervention, context, incentives, and measurement process. The narrower requirement in this condition set is: Preliminary dropout rates exceed planned assumptions or begin to diverge between comparison groups.
Informative follow-up loss · open
Loss to follow-up is outcome- or exposure-dependent enough to destroy exchangeability between completers and the intended comparison frame.
This condition preserves a load-bearing part of the diagnostic problem that was not captured by a source-condition atom. It remains explicit because omitting it would weaken the sufficient condition set.
Other requirements and context (4)
Why these sit outside the expression
Supporting context — it may accompany or help interpret the situation, but it is not a load-bearing condition in a sufficient diagnostic set.
Goal — a goal states an intended outcome or evaluation criterion, not a pre-existing situation that independently summons the archetype.
Supporting contextThe study requires repeated contact, follow-up measurement, attendance, adherence, or continued platform use.
Supporting contextTreatment conditions differ in burden, risk, attractiveness, stigma, cost, inconvenience, or perceived benefit.
When dropout depends on treatment condition, baseline characteristics, burden, adverse experience, or emerging outcome status, the analyzed sample is no longer exchangeable with the intended comparison frame. In this archetype, the relevant contextual consideration is: Treatment conditions differ in burden, risk, attractiveness, stigma, cost, inconvenience, or perceived benefit. It helps interpret the situation or strengthens the practical case for examining the archetype.
Supporting contextSubgroups face unequal practical barriers to continued participation, such as transportation, device access, schedule flexibility, language, or trust.
GoalRegulators, reviewers, or decision makers will need to know whether loss to follow-up threatens the study claim.
Coverage
0 of 3 conditions grounded · 3 open.
Mechanisms / Implementations¶
- Attrition Dashboard: Tracks dropout as it happens — sliced by arm, site, subgroup, time, and reason — so selective loss surfaces while the study is still running, not after it ends.
- Completer Balance Table: Lines up the people who stayed against the people who left, covariate by covariate, to show whether the two groups were ever the same population.
- Data Monitoring Review: An independent body that periodically reads the attrition evidence against pre-set triggers and decides whether to continue, adapt, or stop when loss threatens the inference or the participants.
- Missing-Data Sensitivity Analysis: Re-runs the conclusion under a range of assumptions about the missing outcomes — including deliberately adverse ones — to see whether the finding survives the people who are gone.
- Participant Flow Diagram: Draws the study as a cascade of boxes — assigned, retained, measured, analyzed — so every unit lost between stages is visible on one page.
- Retention Outreach Protocol: A pre-specified, evenly-applied routine for reducing avoidable burden and recovering endpoints — without turning a participant's right to leave into a defect to be eliminated.
- Withdrawal Reason Survey or Interview: Asks the people who left why they left — in their own words, coded but uncertainty-preserving — so a withdrawal is recorded as a diagnosis rather than a blank.
Related Abstractions¶
Abstractions this archetype builds on — directly (a source ingredient) or as a related pattern. Links follow the typed catalog namespace.
Built directly on (3)
- Experimental Design: Structuring an investigation through deliberate intervention, controlled assignment, and measurement so that causation can be distinguished from mere correlation and confounding.
- Missing Data Mechanisms (MCAR, MAR, MNAR): MCAR, MAR, MNAR.
- Selection Bias: Skewed sampling.
Also references 10 related abstractions
- Confounding: Hidden variable interference.
- Data Integrity: Accuracy and consistency preserved.
- Effect Size: Magnitude of effect.
- Measurement Uncertainty and Observational Noise: Measurement noise arises from instrument and observation limits.
- Monitoring: Continuously observing a system's state to detect deviation from expected behavior and trigger a response, separating genuine signal from routine noise.
- Randomization: Assign by chance.
- Statistical Inference: Reasoning from a finite, noisy sample back to the underlying population or process while explicitly quantifying the uncertainty that sampling introduces.
- Statistical Power: Probability of detecting effect.
- Type I & Type II Errors: False positive/negative.
- Validation: Confirming that an artifact actually solves the intended problem in its real operational context, as distinct from confirming it was merely built to specification.
Variants¶
Narrower or domain-specific specializations that share this archetype's core structure. Recognized variants are established; candidate variants are provisional.
Differential Arm Attrition Monitoring · comparison group variant · recognized
Focuses on dropout imbalance between treatment, control, cohort, or policy arms.
Loss-to-Follow-Up Endpoint Recovery · endpoint recovery variant · recognized
Separates discontinuation from endpoint missingness by attempting ethical final outcome collection even when normal participation stops.
Burden-Induced Dropout Monitoring · participant burden variant · recognized
Tracks whether measurement burden, intervention intensity, logistical friction, or psychological cost is causing selective exit.
Nonresponse Attrition Monitoring · survey and panel variant · recognized
Monitors wave-to-wave survey or panel nonresponse and its relationship to observed covariates or previous responses.
Administrative Censoring Audit · censoring variant · recognized
Checks whether administrative rules, data-system cutoffs, transfers, eligibility changes, or record linkage failures remove units unevenly.
Editorial Notes¶
Problem Classification¶
Classification: Uncertainty, Evidence & Inference Failure → Sampling, Selection, Missingness & Generalization
Problem kernel: outcome-dependent dropout corrupts the comparison sample
Rationale: Units cease contributing in ways related to condition, burden, baseline, or emerging outcome, making the analyzed survivors nonexchangeable.
Independent corroboration: The earliest necessary condition in the frozen evidence is: A comparison is designed around an eligible or assigned population, but some units stop contributing data before the endpoint. That is a sampling selection missingness and generalization problem because Observed cases differ systematically from the target because entry, dropout, missingness, case choice, or reuse beyond the sampled domain is ungoverned.
Review outcome: Independent reviewer agreement; high confidence.