Documentation Retention in the Lab: A Practical Guide

By Multimod Labs.

A scientist finishes a late experiment, wipes down the bench, and leaves a half-filled notebook beside the instrument. Two years later, an FDA inspector asks for the original dilution record. The page is missing. A scanned PDF exists, but nobody can confirm whether it includes the complete entry, who made the correction, or where the supporting instrument file went.

That situation captures the core purpose of documentation retention. Retention isn't just deciding how long a file stays in a folder. It connects the moment a scientist records an observation with the later duties of review, retrieval, audit, legal hold, and approved disposal. The record has to remain readable, attributable, complete, and recoverable for the period that applies to its purpose.

A practical rule helps: capture now, classify early, retain by purpose, dispose deliberately, and hold when in doubt.

Table of Contents

• What Documentation Retention Means at the Bench
  • The record begins before the ELN
• Why Retention Periods Are Long and Variable
  • Named timelines show the difference
• Building a Retention Schedule That Fits Your Lab
  • Build the schedule in five passes
• Paper, Digital, and ELN Records Side by Side
• Writing and Enforcing a Real Retention Policy
  • Start with triggers, then define controls
  • Make disposal deliberate
• Data Integrity, Audit Trails, and Contemporaneous Capture
• Ready-to-Use Retention Checklist and Schedule Template
  • Checklist before publishing the schedule
• Putting Retention Into a Daily Lab Workflow
  • A repeatable bench rhythm

What Documentation Retention Means at the Bench

Documentation retention is the combination of a policy, a retention clock, and controlled storage that keeps a laboratory record usable from creation until lawful disposal. The record might be raw analytical data, a protocol, an SOP version, an instrument log, a calibration file, an ELN entry, an image, or an exported report.

The clock matters because different records begin their retention period at different events. A raw observation may need to remain tied to the moment it was captured. A retired SOP may follow the date it was superseded. A clinical-trial record may remain active until a study or product-related trigger occurs. Without a defined trigger, “keep it for years” isn't an operational rule.

Retention also differs from three related ideas:

• Backup creates a recoverable copy after loss or failure. It doesn't automatically define how long the record must remain readable or whether the copy preserves metadata and audit history.
• Archiving moves inactive material into managed long-term storage. An archive can be part of retention, but retention also covers active records and the rules for final disposal.
• Intellectual property protection controls access to inventions and confidential information. It doesn't, by itself, establish a scientific record's retention period or prove that the record is complete.

The record begins before the ELN

A scientist often thinks the official record starts after review or sign-off. In practice, important evidence may appear earlier, in a typed note, voice capture, timer, image, instrument output, or deviation observed during the experiment.

The protocol says what should happen. The bench record shows what happened. That distinction matters when the scientist changes an incubation time, notices an unexpected color, substitutes a material, repeats a measurement, or pauses a procedure because the sample behaves differently.

Practical rule: The first trustworthy capture is part of the record's history, even if it later gets organized into an official ELN entry.

A lab can use a bound notebook, a validated ELN, or an ELN companion workflow for that first capture. Verbex, made by Multimod Labs, is a private, on-device lab documentation app for iPhone that supports voice notes, typed notes, timers, and images, then organizes those source captures into a structured record for human review and export. It isn't an ELN, LIMS, QMS, inventory system, or validated regulatory system.

The working definition is simple: documentation retention keeps the right record, with its context and history, in a usable form for the required period, then permits deliberate and documented disposal when no hold or override applies.

Why Retention Periods Are Long and Variable

A scientist may finish the bench work, close the project, and still need the record years later. A published result may require verification. A product may remain on the market. An adverse event may trigger review. An inspector may need to reconstruct a decision from raw data, calculations, instrument files, and supporting records.

That is why retention begins at the capture moment but often ends much later. The record's value can change over time, even when no one is actively using it.

There is no single global clock. The applicable period depends on the record type, study purpose, product lifecycle, patient-safety exposure, sponsor agreement, and jurisdiction. A lab that supports academic research, clinical studies, and regulated production may face overlapping rules. In practice, the longest applicable requirement usually controls unless a documented exception applies.

Named timelines show the difference

The examples below come from different research and regulated environments:

• UKRI and MRC research records: basic research records should be kept for at least 10 years after completion. Population-health and clinical studies should be retained for at least 20 years after study completion, according to the UKRI research-data retention framework.
• Harvard Medical School: essential research records should generally be retained for no fewer than 7 years after a project ends.
• Johns Hopkins: original research data should be kept for at least 5 years after publication.
• Northwestern: research records generally require at least 3 years after research completion. Certain HIPAA-linked records are kept 6 years, while FDA-regulated drug, device, or biologic records are kept 2 years after marketing approval or discontinuation.
• UK clinical research: sponsors, investigators, and delegated service providers must retain trial documents in the TMF and medical files for at least 25 years beginning the day after trial conclusion, under UK clinical-trial archiving guidance.
• ICH GCP: essential clinical-trial documents must remain available until at least 2 years after the last approval of a marketing application in an ICH region, and until there are no pending or contemplated applications, or at least 2 years after formal discontinuation of clinical development, under ICH Good Clinical Practice E6.

These timelines are not interchangeable. A publication-based clock differs from a study-completion clock, and both differ from a product or regulatory trigger. The trigger must travel with the record so a later reviewer can understand when retention began.

GMP, GLP, and laboratory quality standards add further floors. FDA cGMP rules require batch-specific production, control, and distribution records to remain for at least 1 year after batch expiration, or 3 years after distribution for certain over-the-counter products without expiration dating, as set out in 21 CFR 211.180. UK GLP guidance connects study-data retention with inspection readiness and describes retaining copies for at least one inspection cycle, effectively 2 to 3 years, when original raw data will not remain available, according to UK GLP retention guidance.

The practical rule is straightforward: identify every applicable clock, record its trigger, and document which requirement controls. Retention, ELN history, audit trails, and eventual disposition then operate as one workflow, rather than separate administrative tasks.

Building a Retention Schedule That Fits Your Lab

A retention schedule should be built from the lab's actual record inventory, not copied from a generic template. Start by listing every place scientific evidence appears, including notebooks, ELN entries, instrument software, shared drives, quality systems, email, training systems, image folders, and sponsor portals.

Then define the trigger. The trigger might be capture, study completion, publication, SOP supersession, the last calibration reading, product approval, market discontinuation, sample exhaustion, or project closeout. A record without a trigger can sit indefinitely because no one knows when the clock began.

Build the schedule in five passes

1. Inventory records. List raw files, source notes, protocols, SOPs, instrument records, training evidence, deviations, investigations, correspondence, and attachments.
2. Classify by risk. Mark records as regulated, clinical, sponsor-controlled, legally sensitive, operational, or research reference material.
3. Define the clock. Record the event that starts retention. NIH's intramural ELN policy illustrates why status matters: active ELNs used for scientific reference are retained indefinitely, and the clock starts only after the investigator no longer needs the record and it moves to inactive archival storage. Most inactive ELNs are then kept for 7 years, while patents can require 30 years and FDA or IRB requirements can add other obligations, according to the NIH intramural ELN policy.
4. Map regulatory minimums. Compare the schedule with applicable FDA, clinical, GLP, sponsor, institutional, and local requirements.
5. Approve and publish. Give the schedule an owner, version number, approval record, distribution method, and review date.

A useful schedule can include broad defaults, but each row should be overridden when a contract, sponsor, legal hold, or local regulation demands more. Raw analytical data might need long-term preservation based on sector and study risk. Protocols should remain connected to the study record. Retired SOPs, calibration logs, and ELN exports need their own named rules rather than being left inside a general “lab documents” category.

Don't overlook training records tied to a protocol, deviation and investigation files, instrument qualification certificates, method validation reports, and chain-of-custody forms for reference standards. These supporting records often explain whether a primary result can be trusted.

Paper, Digital, and ELN Records Side by Side

Paper remains understandable and visible at the bench, but it creates practical work for indexing, access, environmental protection, copying, and retrieval. Fully electronic records improve searchability and metadata handling, but the system must preserve access controls, version history, audit trails, exportability, and readability over time.

Hybrid systems combine both burdens. A scientist may write on paper, scan the page, upload the file, attach instrument data separately, and then wonder which copy is authoritative.

NIH states that no new paper lab notebooks may be created for intramural research after June 30, 2024, and ongoing research had to transition to an approved ELN by that date, as described in its ELN transition and recordkeeping guidance. That policy doesn't eliminate paper everywhere, but it shows why hybrid practices are increasingly treated as transition arrangements rather than permanent design.

The scan-and-shred habit needs a controlled decision, not an assumption. A PDF can preserve the visible page, but it may not preserve the original binding, physical signature context, attached labels, ink changes, or evidence of page order. If the paper is destroyed, the lab needs a documented basis for treating the scan as the controlled copy.

For practical guidance on the trade-offs of paper records, a lab manager can review data storage on paper.

The decision rule is straightforward: retain the native record format when duration, searchability, metadata, and audit risk make conversion unsafe. If conversion is necessary, preserve the source context, verify the copy, retain required audit history, and test retrieval before disposing of the original.

Writing and Enforcing a Real Retention Policy

A retention policy becomes useful when named people can answer four questions quickly: who owns the record, what starts the clock, where the record lives, and who can approve disposal.

The records steward maintains the schedule and repository map. A designated approver authorizes disposal. A legal or compliance authority issues and releases legal holds. Scientists remain responsible for capturing and routing records correctly, but they shouldn't have to guess which repository or period applies.

Start with triggers, then define controls

Triggers should be specific. Examples include project closeout, publication, product approval, market discontinuation, sample exhaustion, SOP retirement, or the end of a defined study. A legal hold overrides ordinary disposal as soon as litigation, investigation, or another preservation duty is known or reasonably anticipated.

Storage controls should cover:

• Access control: Limit viewing, editing, export, and deletion rights by role.
• Encryption: Protect records at rest and in transit where the system supports it.
• Redundancy: Maintain a recovery path that doesn't depend on one device or location.
• Restore testing: Prove that retained records can be recovered and opened.
• Format continuity: Preserve the software, metadata, and context needed to interpret the record.
• Auditability: Log transfers, changes, approvals, holds, and disposal actions.

A policy should also state how the laboratory handles mixed records. For example, an image captured at the bench may be source evidence, while a reviewed ELN entry is the controlled study record. Both may need retention, but they don't necessarily share the same owner, format, or clock.

Make disposal deliberate

Disposal should require documented approval, preferably by two authorized people when the record has regulated, clinical, sponsor, or legal significance. The laboratory should retain a disposal log and, where appropriate, a certificate of destruction. That log needs its own retention rule because it proves what was removed, when, under which schedule row, and with whose approval.

Teams handling personal or health-related records may also benefit from a carefully scoped GDPR and HIPAA retention schedule, while labs can use a controlled lab SOP template to document responsibilities, approvals, and review points.

Annual attestations and new-hire training help keep the policy active. The policy becomes shelfware when scientists don't know the repository, the retention trigger, or the hold process, or when the schedule exists but no one checks whether the system can restore an old record.

Data Integrity, Audit Trails, and Contemporaneous Capture

ALCOA+ becomes practical when each principle maps to a decision at the bench:

• Attributable: The entry identifies the person who made it.
• Legible: The record remains readable throughout its retention period.
• Contemporaneous: The observation is recorded when it is generated or observed.
• Original: The source data or a verified true copy is preserved.
• Accurate: The entry reflects the event and is checked for errors.
• Complete: Supporting metadata, attachments, repeats, and exceptions remain connected.
• Consistent: Dates, sequence, units, and record flow make sense.
• Enduring: The record is protected from loss or uncontrolled alteration.
• Available: Authorized reviewers can retrieve it when needed.

WHO good-data principles define contemporaneous data as information recorded at the time it is generated or observed, as stated in WHO GMP data-integrity guidance. That principle affects ordinary actions: a scientist records the unexpected precipitate immediately, starts a timer when the procedure starts, attaches the instrument file rather than typing only a summary, and avoids reconstructing the sequence from memory at the end of the shift.

An audit trail adds another layer. It should preserve who changed what, when the action occurred, and the prior value, without obscuring the original information. FDA-oriented electronic-record guidance says audit trails should be retained for at least as long as the underlying electronic records and remain available for agency review, as summarized in electronic records and audit-trail guidance.

A static PDF may be useful for reading, but it can't automatically replace source data, metadata, or audit history. In a multimodal workflow, source captures can include typed notes, voice notes, timers, and images. Supported material-label images in Verbex's Materials section can be processed on-device into structured Materials entries when the text is sufficiently legible, but the scientist still reviews the result and remains responsible for its meaning.

For a broader treatment of the connection between trustworthy records and review, see data integrity assurance.

Ready-to-Use Retention Checklist and Schedule Template

A schedule should be usable by a scientist who has just finished an experiment, not only by a records administrator. The following template gives each record category a default direction, but sponsor terms, contracts, institutional rules, legal holds, and local regulations override any default.

Checklist before publishing the schedule

• Inventory: Every source location, including instrument software and personal working folders, is listed.
• Classification: Each category has a risk, regulatory, sponsor, and privacy classification.
• Clock: The trigger is an event, not a vague calendar date.
• Override: Sponsor, contract, legal, and local requirements have a visible override field.
• Evidence: Cross-referenced ELN attachments, instrument qualification certificates, method validation reports, deviations, and investigations are included.
• Access: The repository has named owners, role-based permissions, and a recovery process.
• Hold: A legal-hold overlay can suspend normal deletion immediately.
• Disposition: The schedule names the approver, method, date, and retained disposal record.

A records steward should review the schedule at least annually and after every audit, major system change, or relevant regulation change. A new instrument, ELN migration, sponsor agreement, or paper-to-digital transition can create a retention gap even when the written policy hasn't changed.

Putting Retention Into a Daily Lab Workflow

Retention works best when it starts with the scientist's actual capture moment. During the experiment, the scientist records the observation in the bound notebook, approved ELN, or controlled capture workflow, then keeps the supporting evidence attached instead of relying on a later reconstruction.

At sign-off, the record gets classified. The scientist or reviewer identifies whether it is raw data, a deviation, a protocol execution record, a training-linked record, an instrument file, or another category. That classification routes the record to the named repository and connects it to the appropriate retention trigger.

A repeatable bench rhythm

1. Capture contemporaneously. Record what happened while the work is underway.
2. Attach evidence. Link raw instrument files, images, timers, labels, and relevant deviations.
3. Review and complete. A human checks the organized draft against the source captures and corrects errors without hiding the history.
4. File and index. Store the completed record with metadata that supports retrieval.
5. Retain or dispose deliberately. Follow the schedule, pause disposal for a legal hold, and document approved destruction.

Verbex supports this capture-to-record pattern without becoming the official system of record. Users select Objective, Materials, Procedure, Observations, Conclusion, or a custom section before capturing information. Voice and typed notes preserve source context and timestamps, timers document time-sensitive procedures, and images can remain attached as source evidence. Review & Complete creates a source-backed organized draft, and supported devices may offer an additional ELN-style draft when local Apple Intelligence processing succeeds. The scientist reviews and edits before completion, then can export the record as PDF, DOCX, or Markdown for an existing documentation workflow.

The app processes information on the iPhone, requires no account, and uses no cloud AI, cloud storage, advertising, analytics, or tracking. It doesn't replace an official ELN or validated system, interpret scientific results, perform regulatory sign-off, manage inventory or samples, or guarantee compliance.

A retention policy only works when the bench team lives it. The sentence worth repeating to a colleague is: capture now, classify early, retain by purpose, dispose deliberately, and hold when in doubt.

---

Verbex helps bench scientists capture voice notes, typed notes, timers, and images on an iPhone, then review and export source-backed records into an existing ELN workflow. Visit Verbal Experiment or Verbex to see how contemporaneous capture can fit into a practical documentation-retention process.