TMF Completeness and Inspection Readiness

Part of Good Clinical Practice (GCP) — ICH E6(R3): Clinical Trial Closure and Archiving.

About 8 minutes. Free, in your language, no account needed.

Why this matters

An inspector may arrive years after the last participant left. By then the coordinator has moved on, the paper is in an off-site store, and the electronic system has been through two upgrades. Nothing you say on the day will fix a record that cannot be retrieved. Completeness is not a state you reach at close-out. It is a habit you keep for the whole retention period.

Always treat the site file as though an inspector could arrive tomorrow.

What the guideline requires of an archive

ICH E6(R3) is the current revision of the Good Clinical Practice (GCP) guideline. It sets the outcome the archive must deliver.

Section 4.2.7, Retention and Access, states that the trial data and relevant metadata "should be archived in a way that allows for their retrieval and readability and should be protected from unauthorized access and alterations throughout the retention period."

Four requirements sit in that one sentence: retrievable, readable, protected from unauthorized access, and protected from alteration. All four apply for the whole period, not only on day one.

Section 4.2.8 completes the life cycle. Data and relevant metadata may be permanently destroyed when no longer required, as determined by applicable regulatory requirements.

Appendix C.2.4 supports retrieval in practice. The sponsor and the investigator or institution should maintain a record of where essential records are located, including source records. The storage systems used during the trial and for archiving, whatever the media, should provide for appropriate identification, version history, search and retrieval.

Appendix C.2.6 sets the availability standard: records remain complete, readable and readily available, and directly accessible on request by regulatory authorities, monitors and auditors, with any alteration traceable.

Direct access is not improvised on the day. Section 3.16.4(a) requires the sponsor to ensure the protocol or another documented agreement specifies that the investigator and institution provide direct access to source records for trial-related monitoring, audits, regulatory inspection and, where applicable regulatory requirements allow, IRB/IEC review. Section 3.16.4(b) requires the sponsor to ensure that trial participants have consented to that direct access.

Section 2.12.14 states the site's own half of the same duty, and states it without conditions: upon request of the monitor, auditor, IRB/IEC or regulatory authority, the investigator or institution should make available for direct access all requested trial-related records. "All requested" includes the electronic ones and the ones held off site.

Consent and the protocol are what make an inspector's request lawful at your site. Check both before you need them.

The nine sections of a site repository

Appendix C.2.3 says the repository held by the investigator or institution may be referred to as an investigator site file (ISF), and the sponsor's and site's repositories may be referred to as a trial master file (TMF). Whatever you call yours, a stable structure makes records retrievable. Nine sections work well.

• A. Investigator and site staff — curricula vitae, professional licenses, GCP training records, and the delegation of authority log.
• B. IRB/IEC — approvals and favorable opinions for the protocol, amendments, consent materials and recruitment materials; correspondence; continuing review documentation.
• C. Sponsor-provided — the protocol and amendments, the Investigator's Brochure, study manuals, pharmacy instructions and safety letters.
• D. Site-specific operational — the clinical trial agreement, financial disclosure records, insurance documentation and site procedures.
• E. Pharmacy and investigational product — accountability records, temperature records, shipment and return records, and pharmacy training.
• F. Laboratory — laboratory certifications or accreditations, current normal reference ranges, and laboratory staff delegation.
• G. Safety — serious adverse event reports, suspected unexpected serious adverse reaction (SUSAR) notifications, and periodic safety reports.
• H. Screening and enrollment — pre-screening logs, screening and enrollment logs, and the participant identification record.
• I. Other — correspondence, monitoring visit records, and equipment maintenance or calibration records.

Two cautions. First, this is a working structure, not a checklist from the guideline. E6(R3) uses the term essential records and its Appendix C carries the Essential Records Table. Second, under Appendix C.3.3 some listed records may not be produced at all for a given trial, because presence and nature depend on trial design, conduct and risk-proportionate management. Appendix C.2.12 also allows non-trial-specific records, such as the Investigator's Brochure, master services agreements, standard operating procedures and validation records, to sit outside the trial-specific repository.

Inspection-readiness criteria

Apply these to every section, on a routine cycle, not at close-out:

• Records are filed in the correct section.
• Records are clean, complete and legible.
• Records are organized chronologically within each section.
• There are no loose pages and no missing signatures or dates.
• Records are accessible on demand, including from the electronic systems.
• The location record required by Appendix C.2.4 is current, so anyone can say where a record is.
• Version control is visible, and superseded versions are marked as superseded rather than discarded.

A useful test: give a new colleague a record name and ask them to produce it. If it takes more than a few minutes, the record is not readily available, whatever the file index says.

Mock inspections

A mock inspection is an internal rehearsal run by someone who did not do the work. It converts hidden problems into findings you can still fix.

Run one at three points: before the first monitoring visit, at the trial's midpoint, and at close-out. A simple method works.

1. Pick five participants at random and follow each from screening to last visit through source, the data acquisition tools and the consent records.
2. Ask for ten named records from different sections and time how long each takes to produce.
3. Ask a coordinator to explain who was delegated a specific activity on a specific date, and to show the evidence.
4. Test the electronic systems: can you retrieve and read an audit trail from a year ago?
5. Write findings the way an inspector would, then run a corrective and preventive action for anything systemic.

Do not fix a mock finding by writing a record after the fact. That is the same error as creating documentation during a real inspection, and it is worse than the gap it hides.

In practice

Three years after Dr. Sarah's hypertension trial closes, a regulatory authority notifies the institution of an inspection. Part of it will be conducted remotely.

Amina, the coordinator, has left. The institution's research records manager, named to the sponsor under the retention requirements, holds the archive.

Four problems appear.

The paper file is in an off-site store with a 48-hour retrieval time. That fails "readily available" under Appendix C.2.6 for a remote review starting in three days. The records manager requests immediate retrieval.

The electronic data capture system was decommissioned and its contents exported. Nobody has confirmed the exported audit trails are still readable. Section 4.2.7 requires retrieval and readability throughout the retention period, so this is a genuine risk, not a formality. The sponsor tests a sample export and confirms it opens.

The location record required by Appendix C.2.4 was never updated after the off-site move. It still points to a cupboard that no longer exists.

Nobody can immediately confirm that participants consented to direct access. Section 3.16.4(b) makes that the sponsor's assurance, and section 3.16.4(a) requires the protocol or an agreement to specify direct access. The records manager finds both, in the protocol and in the approved consent form, within an hour.

Three of these four failures were failures of retrieval, not of retention. The records existed. That is the pattern inspections keep finding.

An archive nobody has opened in three years is an untested archive.

Key takeaways

• ICH E6(R3) section 4.2.7 requires trial data and relevant metadata to be archived so they can be retrieved and read, and to be protected from unauthorized access and alteration throughout the retention period.
• Section 3.16.4 requires direct access to be specified in the protocol or another documented agreement, and requires the sponsor to ensure participants have consented to that direct access; section 2.12.14 requires the investigator or institution to make all requested trial-related records available for direct access on request by the monitor, auditor, IRB/IEC or regulatory authority.
• Appendix C.2.4 requires the sponsor and the investigator or institution to maintain a record of where essential records are located, and requires storage systems to support identification, version history, search and retrieval.
• A site repository is inspection-ready when records are correctly filed, legible, chronologically organized, free of missing signatures and dates, and accessible on demand.
• Under Appendix C.3.3, a record listed in the Essential Records Table may not have been produced for a given trial, so an absence must be explainable rather than assumed to be a gap.
• Mock inspections run by someone who did not do the work should test retrieval time and the readability of archived electronic records, not only whether a record exists.
• A mock inspection finding must never be closed by creating a record after the fact.