Detention Notice Evidence Completeness: A Cohort Method

Published: September 14, 2026.

This study measures whether eligible detention-notice records contain a predefined administrative evidence set by a fixed cutoff. It does not establish detention entitlement, contract meaning, timestamp accuracy, notice delivery, fault, or payment.

Research question and cohort

The question is: what share of eligible reported detention events has each required evidence element recorded by the carrier-defined cutoff? Select the cohort from an independent load or event source. Do not build it only from notices that reached billing.

Register the business units, facilities, period, qualifying event definition, exclusion rules, cutoff, and expected systems. Keep loads with incomplete or missing event data visible.

Evidence completeness rule

Create a data dictionary before review. Candidate elements might include load identifier, appointment source, arrival evidence, release or departure evidence, notice rule source, notice artifact, destination, delivery evidence, and owner decision. The carrier must choose the actual list based on its process.

Define acceptable source types and usability tests for each field. A populated field is not complete if its required source is absent or unreadable. Conversely, a missing non-required field should not fail the registered measure.

Methodology

Freeze eligible events, then inspect records without changing the rule. Assign each required element one of four states: present and usable, present but unusable, absent, or not applicable under a documented rule. Keep conflicts as an additional flag.

Calculate element-level completeness and whole-record completeness. Publish counts for eligible, fully complete, partly complete, not calculable, excluded, unresolved, and still open records. Report results at the fixed cutoff so later collection does not rewrite the original observation.

Conduct a second analysis after an announced follow-up window if useful. Label it as a later snapshot.

Reviewer agreement

Two reviewers should independently classify a sample containing clean, missing, conflicted, late, and corrected records. Compare their field states and exclusion decisions. Resolve differences through the registered data dictionary and retain the adjudication history.

Interpretation boundaries

Low completeness can reflect source design, access restrictions, late uploads, inconsistent procedures, or capture errors. High completeness means the defined artifacts were located; it does not show that they are accurate, timely, persuasive, or accepted by another party.

Avoid facility or individual rankings when groups are small or process differences were not registered.

Limitations

The metric depends on the chosen evidence list and cutoff. Records outside approved systems may be missed. Later corrections and duplicate files can change classifications. Findings do not generalize beyond the named cohort and procedure.

Source treatment

Regulatory and records-management sources provide context only. They do not supply the local completeness checklist or a benchmark. The study publishes direct links and retrieval dates and identifies every local rule as local.

Conclusion

Evidence completeness is a practical control measure when the study includes missing and unresolved records. It describes administrative record availability, not the merits of a detention request.

References

1. https://www.ecfr.gov/current/title-49/subtitle-B/chapter-III/subchapter-B/part-371, retrieved 2026-09-14
2. https://www.ecfr.gov/current/title-49/subtitle-B/chapter-III/subchapter-B/part-379, retrieved 2026-09-14
3. https://www.archives.gov/records-mgmt/policy, retrieved 2026-09-14