Records, Content & Retrieval
The platforms that hold documents and knowledge: structured content, document services, search, taxonomies, archival, retrieval, and the reuse of what the organisation already knows. Retrieval, not storage, is the obligation — a record that cannot be found within the time an investigation or an inspection allows is functionally missing, and retrieval is the property least often tested.
What an explainer is not
A topic explainer is SPEQ’s synthesis of what a practice involves, cited to the standards that govern it. It does not reproduce their text, and it does not determine which of them apply to your product or process.
[ POSITION IN THE FRAMEWORK ]
7 DIMENSIONS · 24 LINKSRetention is the easy half. The obligation is retrievability — a record that exists on media nobody can read, in a format nothing renders, has been kept and lost simultaneously.
06 · QUALITY MATURITY — RECORDS, CONTENT & RETRIEVAL, REACTIVE TO ADAPTIVE
Records are kept where they were created. Retention is understood as not deleting things.
A retention schedule exists by record type, and archived records are stored without any test of whether they can be retrieved.
Retention derives from the obligation per record type and jurisdiction, and retrievability is periodically demonstrated rather than assumed.
Format and media obsolescence are managed, with migration performed while the original can still be read and the migration verified.
The archive is a working asset: a record from any point in the retention period can be produced, in context, within a defined period.
SPEQ’s shared five-stage progression, labelled synthesis — not the FDA QMM rating scale. Where does your organization sit? Score your quality system →
07 · REGULATORY & EVIDENCE
GOVERNING STANDARDS · 5
Derived from the 5 standards SPEQ maps to this subject, across 5 regulatory bodies: FDA, EMA, ICH, MHRA, WHO.
RECORDS & OBJECTIVE EVIDENCE
- The retention schedule by record type, with the obligation each derives from
- Retrieval testing records, demonstrating archived records can be produced
- Format and media migration records, with verification of completeness
- Metadata and context retained alongside records, including audit trails
- Disposition records where retention periods expired
COMMON INSPECTION FINDINGS
- Archived records on media or in formats no longer readable
- Retention schedules that state periods but not the obligations behind them
- Records migrated with no verification that content and metadata survived
- Audit trails not retained with the records they belong to
- Retrieval never tested, so the archive’s usability is unknown until it is needed
Retrieval is the requirement that is rarely verified
Record-retention programmes are audited for whether records are kept, in what format, for how long. They are rarely audited for whether a specific record can be produced on request within a useful time. Those are different properties, and an archive that satisfies the first and fails the second has met the letter of an obligation whose purpose it has missed.
The test is simple and almost nobody runs it: pick a record from six years ago at random, ask for it, and time how long it takes to arrive complete with its metadata and approval history. Where the answer is days, or where it arrives without its audit trail, the retention programme needs work regardless of how well the storage is documented.
Metadata is what makes retrieval possible at scale
Finding a record depends on the attributes captured when it was filed — product, site, process, date, document type, related batch or study — and those are decided by a taxonomy that is usually designed once and then degrades. New document types are filed under approximate categories, a naming convention drifts, and after several years the collection is searchable only by someone who knows where things went.
Full-text search mitigates and does not solve this, because the question is usually structural: every change control affecting this equipment, every deviation on this product line, every version of this specification in force on a given date. Those are metadata queries, and they only work if the metadata was captured consistently — which requires periodic review of the taxonomy against what is actually being filed.
Archival has to preserve meaning, not just bytes
Long-term retention in regulated industry regularly outlives the application that created the record. WHO guidance on good documentation practice and the electronic-records requirements both point at the same obligation: the record must remain accurate, legible, retrievable and complete for its retention period — including its metadata and audit trail, which is the part that migrates worst.
A record migrated as a flat rendering has lost the approval history and the change record that made it evidence. The practical requirement is that archival preserves the record, its metadata and its audit trail together in a form readable without the original application, and that a restore from the archive is actually tested rather than assumed.
SPEQ interpretation — knowledge reuse is a retrieval problem
Organisations invest in knowledge management as a cultural programme and then fail at it for a mechanical reason: the knowledge exists, in a report or an investigation or a development study, and cannot be found by someone who does not already know it exists. ICH Q10 treats knowledge management as an enabler of the quality system, and the enabling step is retrieval.
The concrete version of this is that a deviation investigation should be able to find every prior investigation of the same phenomenon on the same equipment, and a development team should be able to find what was learned about a formulation before repeating the study. Both are metadata queries against records the organisation already holds, and both fail for the same reason — nothing linked the record to the thing it was about.
FREQUENTLY ASKED
How do you test a retention programme properly?
Ask for a specific record from several years ago and time how long it takes to arrive complete with metadata and approval history. Retention audits check that records are kept; they rarely check that a named record can be produced usefully quickly, and those are different properties.
Why doesn’t full-text search solve findability?
Because the questions are usually structural — every change control affecting this equipment, every deviation on this product line, every version of a specification in force on a given date. Those are metadata queries, and they only work where metadata was captured consistently, which requires reviewing the taxonomy against what is actually being filed.
What does archival have to preserve beyond the document?
Its metadata and audit trail, in a form readable without the original application. A record migrated as a flat rendering has lost the approval history and change record that made it evidence — and a restore from archive should be tested rather than assumed to work.
Why does knowledge management fail for mechanical reasons?
Because the knowledge exists and cannot be found by someone who does not already know it exists. A deviation investigation should be able to find every prior investigation of the same phenomenon on the same equipment; that is a metadata query against records already held, and it fails because nothing linked the record to what it was about.