A bounded measure for peptide sample storage-location agreement

Sample inventories are useful only when their location fields can be checked against defined evidence. Agreement is measurable, but its meaning is limited. A matching freezer position does not prove sample identity or condition.

Method and evidence scope

I reviewed FDA guidance on data integrity, the current electronic text of 21 CFR Part 11, and ICH E6(R3). The review focused on attributable records, controlled electronic evidence, and management of trial-related information. From those principles, I derived a point-in-time comparison method for administrative sample-location records.

This work is a qualitative synthesis. It is not an audit, inspection, observational study, or representative survey. No sample records were collected. The proposed measure is an inference for workflow design and requires local validation by qualified scientific and quality owners.

Establish the population before looking

Define the study, storage area, sample status, and time window. State whether archived, shipped, destroyed, quarantined, or temporarily removed samples are included. Capture sample identifier, expected location, observation reference, timestamp, reviewer, and both source systems.

Use four primary outcomes: agreement, documented mismatch, excluded under the written rule, and unable to verify. A missing inventory entry and a sample not observed at the expected position should remain separate subtypes. They support different questions.

Control the observation and handoff

Only trained, authorized staff should observe or handle controlled storage. An administrative analyst may reconcile approved reports but should not move, relabel, or dispose of a sample. Preserve the source extracts and their timestamps so the result can be reconstructed.

Route mismatches to the designated custody, scientific, or quality owner. Record who accepted the item, the decision date, any approved correction, and closure evidence. Do not overwrite the original observation when a record is corrected.

Calculate and report without overreach

The agreement proportion is the number of reviewed samples with matching defined locations divided by the eligible, verifiable sample count. Report the selected population, exclusions, and unable-to-verify count alongside it. A second measure using the full selected population may help show access gaps, but it answers a different question.

The result applies only to the records and observation time reviewed. It cannot demonstrate sample identity, temperature control, stability, chain of custody, or suitability for analysis. A mismatch also does not identify when or why the records diverged.

Trend comparisons need stable definitions, observation methods, and system states. A bulk inventory cleanup could improve agreement while making comparisons with earlier periods difficult.

Limitations

The cited sources do not prescribe this exact metric, a peptide-specific target, or a universal sampling frequency. ICH E6(R3) concerns clinical trials and may not apply to every peptide research setting. Part 11 applicability depends on the electronic records and regulated context.

The method was not tested for observer agreement, sampling bias, transient movements, system latency, or false mismatches. It does not account for the safety burden of physical observation. Organizations should pilot the method on a small authorized population and have qualified staff resolve ambiguous cases.

References

• FDA: Data Integrity and Compliance With Drug CGMP
• eCFR: 21 CFR Part 11, Electronic Records and Signatures
• ICH E6(R3): Guideline for Good Clinical Practice