The eTMF, Operationalized: Completeness, Timeliness, and Quality That Survive a GCP Inspection
The vendor pages will tell you an eTMF is "the digital version of the TMF, with version control and audit trails, and it helps with GCP compliance." That is where they stop, and it is exactly where the real work starts. This article is for the clinical-ops lead, the TMF specialist, the QA reviewer, and the CRA or CRC who has just inherited an eTMF or is staring down an inspection and needs to know one thing: is mine actually inspection-ready, and where will it bite me.
Aileen
Aileen writes practical guidance for clinical trial teams at GCP Blog.
On this page · 9 sections
- 01 At a glance
- 02 The eTMF, defined: the regulated repository, not the software
- 03 What “GCP-compliant” actually requires of an electronic system
- 04 The three yardsticks inspectors apply
- 05 What goes in it: essential documents and the TMF Reference Model
- 06 eTMF reconciliation and ongoing QC: the workflow that keeps it inspection-ready
- 07 The findings that actually surface in inspections
- 08 Ownership, oversight, and retention
- 09 Sources
At a glance
- An eTMF is the regulated repository of essential records for a trial, not the software you bought to hold it. The system is a tool; the obligation to keep a complete, reconstructable record sits with the sponsor and the investigator/institution.
- Inspectors do not grade your vendor. They grade three things: completeness (is everything that should be filed actually filed), timeliness (was it filed contemporaneously, not back-filled before the audit), and quality (is each document the right version, legible, correctly indexed, with an intact audit trail).
- An electronic system earns GCP-equivalence with paper only when it carries the controls that ICH E6(R3) and the EMA computerised-systems guideline require: validation, access control, a secure audit trail, certified copies, and controlled electronic archiving.
- Reconciliation and ongoing QC are the workflow that keeps the triad true between today and the inspection. They are a planned, risk-based activity, not a pre-inspection scramble.
- “eTMF” is the repository of essential documents; “TMF Reference Model” (a DIA industry standard, not a regulation) is the indexing convention you map it to; “CTMS” is operational tracking and is not the regulated record.
- Most inspection findings trace back to one broken leg of the triad. Fixing the software rarely fixes the finding.
The vendor pages will tell you an eTMF is “the digital version of the TMF, with version control and audit trails, and it helps with GCP compliance.” That is where they stop, and it is exactly where the real work starts. This article is for the clinical-ops lead, the TMF specialist, the QA reviewer, and the CRA or CRC who has just inherited an eTMF or is staring down an inspection and needs to know one thing: is mine actually inspection-ready, and where will it bite me.
The eTMF, defined: the regulated repository, not the software
Start with what the regulation actually regulates. Under ICH E6(R3), essential records are the documents and data, in any format, that collectively allow evaluation of how a trial was conducted. ICH E6(R3) §C.2.3 states that these essential records should be maintained in or referred to from repositories held by the sponsor and by the investigator/institution, and that these repositories may be referred to as a trial master file (TMF), with the investigator-held repository also called the investigator site file (ISF). The “e” in eTMF only describes the media. It does not change what the repository is for or who owns the obligation.
That obligation has a purpose, and the purpose is the yardstick. ICH E6(R3) §C.1.3 explains that essential records permit and contribute to the evaluation of a trial in relation to the compliance of the investigator and sponsor with GCP and the reliability of the results produced, and that these records are used during inspections by regulatory authorities to assess trial conduct. The EMA TMF guideline puts the same idea operationally: a TMF should be sufficient to adequately reconstruct the activities undertaken in conducting the trial, along with the decisions and justifications made. If a reviewer cannot reconstruct the trial from your eTMF, the system has failed its only job, regardless of how slick the interface is.
Three things get conflated constantly. Hold them apart:
| eTMF | Paper TMF | CTMS | |
|---|---|---|---|
| What it is | The regulated repository of essential records, on electronic media | The same regulated repository, on paper | An operational system for tracking trial activities (sites, enrolment, monitoring visits, payments) |
| What it is NOT | Not just a shared drive; not the CTMS; not “the software” | Not a substitute for contemporaneous filing | Not the regulated record of essential documents; not a place inspectors reconstruct the trial from |
| Who owns it | Sponsor and investigator/institution each own their respective records (ICH E6(R3) §C.2.7) | Same | Operational team / sponsor; not a record-keeping obligation in itself |
| Regulatory status | Essential records repository under ICH E6(R3) and the EMA TMF guideline | Same | Tooling; tracked metadata may inform the TMF but the CTMS is not the TMF |
The practical consequence: a document that lives only in your CTMS, or only on a shared drive, is not “in the TMF” in any sense an inspector will credit. The TMF is defined by what it is for, not by which screen it appears on.
What “GCP-compliant” actually requires of an electronic system
Paper has implicit controls: you can see a wet-ink signature, a date, an obvious erasure. An electronic system has none of that for free. ICH E6(R3) and the EMA computerised-systems guideline spell out what an electronic system must carry to be trustworthy enough to hold essential records. Software that merely stores files is not yet an eTMF in the regulatory sense; it becomes one when these controls are in place and demonstrated.
Validation. ICH E6(R3) §4.3.5 (Validation) and the EMA computerised-systems guideline both require that computerised systems used in a trial be validated against their specified requirements. The EMA computerised-systems guideline (§Validation) states the validation approach should be based on a risk assessment that takes into account the intended use of the system and its potential to affect human-subject protection and the reliability of trial results. Validation is not a one-time vendor certificate; if you rely on a vendor’s validation documentation, the EMA computerised-systems guideline (§Validation of systems) expects you to confirm it covers your intended use and your defined requirements.
Audit trail. The EMA computerised-systems guideline (§Audit trail) requires that an audit trail be enabled for the original creation and subsequent modification of all electronic data, as a secure, computer-generated, time-stamped record allowing reconstruction of creation, modification, and deletion events. Critically, it also requires (§Audit trail review) that procedures for risk-based, trial-specific audit-trail reviews be in place. An audit trail you never review is evidence you are not using, and inspectors know it.
Access control. The EMA computerised-systems guideline (§Security and access control) requires security processes and features that prevent unauthorised access and unwarranted data changes, including unique user accounts. The EMA TMF guideline reinforces this with the expectation of role-based permissions and restricted access to files and documents.
Certified copies. When an electronic copy permanently replaces an original, it has to be a certified copy. ICH E6(R3) §C.2.9 requires that when a copy is used to permanently replace the original essential record, the copy must fulfil the requirements for certified copies, and ICH E6(R3) defines a certified copy as a copy verified (by dated signature or by a validated process) to contain the same information as the original including relevant metadata. The EMA TMF guideline says the same for documents scanned into the eTMF: copies that irreversibly replace originals should be certified copies. This is where casual scanning practices fail, because “we scanned it and shredded the paper” without a certified-copy process means you may have destroyed the original record.
Electronic archiving. ICH E6(R3) §4.2.7 (Retention and Access) requires that trial data and relevant metadata be archived so they remain retrievable and readable and are protected from unauthorised access and alteration throughout the retention period. The EMA TMF guideline adds the practical maintenance burden: access and readability must be preserved for the entire archiving period, which may mean maintaining old systems, emulating them, or migrating data into new formats.
The three yardsticks inspectors apply
Everything above feeds three questions an inspector actually asks. Frame your own eTMF health against them.
Completeness. Is everything that should be filed actually filed? ICH E6(R3) §9.5 requires that essential records be retained and be available to regulatory authorities, monitors, auditors, and ethics committees upon request to enable evaluation of trial conduct. The EMA TMF guideline expects sponsors and investigators to have procedures to assure the TMF is complete, legible, and accurate, and to have reviewed that filed documentation is complete before archiving.
Timeliness (contemporaneous filing). Was it filed as the trial happened, not reconstructed afterward? ICH E6(R3) §C.2.5 requires the sponsor and investigator/institution to ensure that essential records are collected and filed in a timely manner. This is the leg vendor pages skip entirely, and it is the one that most often catches teams out, because a back-filled TMF can look complete on the day of inspection while failing the contemporaneousness test the audit trail then exposes.
Quality. Is each document the correct version, legible, correctly indexed, and traceable? The EMA TMF guideline (§Security and control) requires documents to be correctly indexed, accessible only per assigned roles, and, for an eTMF, subject to audit-trail review. ICH E6(R3) §C.2.1 adds that records should be identifiable and version-controlled, with authors, reviewers, and approvers as appropriate, and ICH E6(R3) §C.2.6 requires that any alteration to essential records be traceable.
A concrete timeliness cadence, because “timely” is not a platitude: file each essential document as it is finalized, with a target that a document is in the eTMF within days of signature or finalization, not at the next monitoring visit and never at study close. The test you should apply to yourself: if an inspector walked in unannounced this week, would the audit trail show your last filings clustered suspiciously around the inspection notice date? If so, you have a timeliness problem the software cannot cover for you.
What goes in it: essential documents and the TMF Reference Model
“What goes in it” is answered by what is essential, and ICH E6(R3) Appendix C answers that with criteria plus an Essential Records Table. ICH E6(R3) §C.3.1 sets out the criteria for whether a record is essential, and §C.3.2 lists the records considered essential when produced, while noting the list is not exhaustive. The EMA TMF guideline adds that documents not named in the ICH list but necessary to reconstruct and evaluate trial conduct, such as completed forms, checklists, and reports, should still be filed.
To index that content consistently, the industry uses the DIA TMF Reference Model. It is an industry standard, not a regulation, and it does not carry regulatory force. Treat it as the indexing convention that maps your essential records into zones, sections, and artifacts so that completeness gaps become visible. The regulatory requirement is to have a complete, reconstructable, well-indexed repository; the Reference Model is simply the most common way teams operationalize the indexing the EMA TMF guideline already expects.
eTMF reconciliation and ongoing QC: the workflow that keeps it inspection-ready
Reconciliation is the discipline that connects “we have a system” to “we are inspection-ready.” Two distinct activities live under this word, and both are grounded.
Ongoing QC review. The EMA TMF guideline (§Security and control) requires the sponsor and/or investigator/institution to implement risk-based quality checks or review processes to ensure the TMF is being maintained up to date and that all essential documents are appropriately filed. This is not annual housekeeping; it is the recurring pass that keeps completeness and timeliness honest between today and the inspection.
Audit-trail review. The EMA computerised-systems guideline (§Audit trail review) requires procedures for risk-based, trial-specific audit-trail reviews. In an eTMF this is where you catch the back-dated upload, the document filed in the wrong version, the deletion without explanation. The EMA TMF guideline explicitly lists review of the audit trail as part of eTMF QC.
Document-vs-document reconciliation. Beyond the TMF, ICH E6(R3) recognizes reconciliation as a control on data integrity: §4.2.5 (Data Transfer, Exchange and Migration) names reconciliation as one of the processes used to ensure electronic data retains its integrity when moved between systems. The same mindset applies inside the TMF: reconcile expected-vs-filed against your index, reconcile site-level filings between the ISF and the sponsor TMF, and reconcile at milestones (site activation, database lock, study close).
A practical running cadence: per-milestone completeness reconciliation against the Reference Model index, a rolling timeliness check using the audit trail, and a quality pass on version control and certified-copy status. Run it continuously, document it, and you walk an inspector through a system that proves its own health rather than asserting it.
The findings that actually surface in inspections
Most eTMF findings are not exotic. They are the predictable failure of one leg of the triad. Map each to the leg it breaks and the control that fixes it.
| Common finding | Which leg it breaks | The fix (and its basis) |
|---|---|---|
| Essential documents missing or only in the CTMS / a shared drive | Completeness | Reconcile expected-vs-filed against the index; file everything essential per ICH E6(R3) §C.3.2 and the EMA TMF completeness expectation |
| Documents back-filled; audit trail shows a cluster of uploads before inspection | Timeliness | Enforce contemporaneous filing per ICH E6(R3) §C.2.5; rolling timeliness QC via audit-trail review |
| Wrong or superseded version filed; no version history | Quality | Version control and traceable alteration per ICH E6(R3) §C.2.1 and §C.2.6 |
| Scanned copies replaced originals with no certified-copy process | Quality | Certified-copy process per ICH E6(R3) §C.2.9 and the EMA TMF guideline before destroying originals |
| Audit trail not reviewed (or reviewable) | Quality | Implement risk-based audit-trail review per the EMA computerised-systems guideline (§Audit trail review) |
| Broad or shared access; no role-based permissions | Quality | Unique accounts and role-based access per the EMA computerised-systems guideline (§Security and access control) |
| System never validated for its intended use | Quality | Risk-based validation per ICH E6(R3) §4.3.5 and the EMA computerised-systems guideline (§Validation) |
Where teams get it wrong, in one sentence: they treat the eTMF as a filing cabinet to be tidied before an inspection, when the regulation treats it as a continuously maintained, contemporaneously filed record whose audit trail will testify against any last-minute tidying.
Ownership, oversight, and retention
Ownership is shared and must be documented. ICH E6(R3) §C.2.7 states that each party retains the essential records required to fulfil their responsibility and that original records should generally be retained by the party who generated them. ICH E6(R3) §3.16.3 places record-keeping and retention obligations squarely on the sponsor for sponsor-specific essential records, while §2.12.11 makes the investigator/institution responsible for maintaining and controlling the essential records they generate. When a CRO holds the TMF, the EMA TMF guideline expects the contractual arrangement to ensure the sponsor retains the full set of documents and can make them readily available for inspection, including inspections of the CRO’s own duties. None of this is delegated away by buying software, and the sponsor remains responsible for oversight.
Retention is long and the system must outlive its own technology. Under the EU Clinical Trials Regulation, as the EMA TMF guideline notes, the sponsor and investigator shall archive the content of the clinical TMF for at least 25 years after the end of the trial unless other Union law requires longer. ICH E6(R3) §2.12.12 requires the investigator/institution to retain essential records for the required period and to take measures ensuring availability, accessibility, and readability while preventing unauthorised access and premature destruction. For an electronic archive, the EMA TMF guideline’s practical point bears repeating: you must preserve access and readability for the entire period, which can mean maintaining, emulating, or migrating systems. At study close, the transfer of the TMF (or parts of it) between sponsor and CRO must itself be reconciled so nothing is lost in the handover.
A note on alignment between the in-scope regulations: on the substance covered here they reinforce rather than contradict each other. ICH E6(R3) sets the essential-records and certified-copy obligations; the EMA TMF guideline operationalizes completeness, timeliness, QC, and archiving; the EMA computerised-systems guideline supplies the electronic-control detail (audit trail, access, validation, certified copies) that gives an electronic system equivalence with paper. Where they touch the same point, certified copies for instance, they say the same thing. No conflict required surfacing here. If your own situation surfaces a genuine tension between two requirements, the discipline is to document both and resolve it with QA, not to smooth it over in prose.
For deeper treatment of the indexing convention, the underlying essential-documents obligation, the system-selection questions, and inspection-readiness more broadly, see the sibling articles on the TMF Reference Model, essential documents in clinical trials, eTMF systems, CTMS vs eTMF, the clinical-trial audit trail, and GCP inspection readiness.
Sources
- ICH E6(R3) Good Clinical Practice (ICH), version r3 — https://www.ich.org/page/efficacy-guidelines
- EMA Guideline on content management and archiving of the clinical trial master file (EMA), version 2018
- EMA Guideline on Computerised Systems and Electronic Data in Clinical Trials (EMA), version 2023 — https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/guideline-computerised-systems-and-electronic-data-clinical-trials_en.pdf
- TMF Reference Model (DIA) — industry indexing standard, not a regulation; referenced as the indexing convention only
Written by
Aileen
Aileen writes practical guidance for clinical trial teams at GCP Blog.
Continue reading
The Clinical Research Coordinator as a GCP-Accountability Map: Which CRC Duties Carry Regulatory Weight Under ICH E6(R3) and 21 CFR 312
If you are an early-career coordinator, you have read a dozen pages that list what a CRC does. This one does something different. It treats each duty as a line item in a GCP-accountability map: the task you perform, the regulation that makes it mandatory, and the part of that obligation the principa...
ReadEDC, eCRF, eCOA and IWRS as One Regulated Data Flow: Who Owns the Source Record and Where ALCOA Breaks at the Seams
Five acronyms recur on every modern study, and practitioners are routinely asked to "define the difference" as if it were trivia. The disambiguation is worth doing, but only as a setup for the real question: where does the source record sit, and who put it there.
ReadClinical Trial Design and Management as One Quality-by-Design Chain: From Critical-to-Quality Factors to Risk-Based Oversight
This is the pillar for clinical-operations leads, study managers, and CRAs who are scoping a new trial or onboarding to one. It is deliberately deep on the connective logic and the tables, and it names sibling playbooks (quality by design, CtQ factors, quality tolerance limits, monitoring plans, dec...
Read