info@hytanpeptides.com
Hytan Peptides — Peptide CDMORequest Project Quote

Technical resources

Reviewing chromatography audit trails in a peptide CDMO project

A final chromatogram may omit the events that produced its reported result. A focused audit-trail review connects acquisition, processing changes and approvals, allowing reviewers to investigate unexplained events while preserving the complete analytical record.

A reviewer receives a passing analytical report, but the associated record shows several processing changes. Those changes may have a legitimate explanation. The review should establish what happened, why it happened and whether the approved procedure was followed. A clean final printout cannot answer those questions on its own.

The FDA data-integrity guidance explains the role of metadata, audit trails and static versus dynamic records in drug CGMP. It also addresses review of changes and incomplete or aborted analyses. These principles inform the proposed review approach here; they do not certify any instrument, laboratory or supplier.

Start from the reported result

Identify the sample, acquisition sequence, method revisions and processing version supporting the reported value. Follow those links through the system with an authorized reviewer. Record gaps that prevent reconstruction, including a missing method or an unclear relationship between sample labels and electronic identifiers.

Ask whether the reviewer can see the relevant events in their native context. A filtered export may be useful, provided its coverage is understood. Document the filters and time interval used, so another reviewer can reproduce the selection and assess whether events were omitted.

Explain the recorded events

Useful questions include why a sequence was stopped, whether an injection was repeated, and what prompted an integration change. Examine the procedural basis and contemporaneous explanation. A large number of events is not proof of misconduct, and a short event list is not proof that the data are complete.

Consider a proposed review example: an analyst changes a peak boundary, recalculates a result and saves a new report. The reviewer should be able to identify both processing versions, the affected result and the documented rationale. The existence of an audit entry establishes that an action was recorded; it does not establish that the action was scientifically justified.

Separate acquisition changes from processing changes

Acquisition determines what was measured. Processing determines how recorded signals were interpreted under the chosen procedure. Keep those histories distinguishable. Reprocessing an existing signal does not create an independent preparation or a new manufacturing observation.

Review permission settings and unusual administrative actions when they affect the records being assessed. A shared account can make individual attribution difficult. Escalate that limitation through the laboratory’s quality system rather than inventing an identity from shift patterns.

Preserve the evidence while resolving exceptions

Use an exception log that names the event, its possible impact and the evidence needed to close it. Request explanations through the approved investigation process. Do not overwrite the original record to make the final report easier to read.

  • Retain the original acquisition and relevant processing histories.
  • Link any amended report to its predecessor and approval record.
  • Document whether the issue affects one result or a wider set of records.
  • Record the review outcome and remaining restrictions on reliance.

Where confidential data cannot leave the laboratory, arrange controlled access sufficient for review. A summary can record the conclusion, but it should not imply access to information the reviewer never examined. Agree how a later question will be answered without depending on a single employee’s memory.

Make review coverage explicit

Define the relationship between routine result review, risk-based system review and quality oversight. A software vendor’s feature list does not show whether audit logging is enabled, protected or reviewed in the actual configuration. Request evidence of the implemented workflow.

The site’s analytical testing scope can be used to discuss review access and deliverables. The outcome should state which records were examined, which exceptions were resolved, and which conclusions remain unsupported. This article concerns record reconstruction, not the scientific assignment of peptide sequences.

Questions and answers

Does an audit-trail entry make manual integration acceptable?

It records the action. Acceptability also depends on the approved method, the scientific rationale, preserved earlier versions and appropriate review. A logged action can still be unjustified.

Is a PDF chromatogram a complete dynamic analytical record?

It may preserve a useful presentation but can omit the ability to inspect processing history and associated metadata. Assess what must remain accessible to reconstruct the analysis and retain that content in a suitable form.

What should a reviewer do with an unexplained aborted injection?

Record the event and request the relevant explanation and supporting evidence through the quality process. Do not omit it because a later injection produced a passing result.

Can a shared login identify the analyst responsible for a change?

It does not reliably distinguish individual users. Treat that as an attribution limitation and assess its impact; do not infer a person’s identity solely from a work schedule.

May confidential audit records be reviewed without sending copies to the sponsor?

Controlled access can support review when its scope is sufficient and documented. Agree access, retained conclusions and future retrieval arrangements, and state any information the reviewer could not inspect.

References