MHRA produced FAQs for Trial Master Files (TMF) and Archiving

[MHRA Super Moderator]

5. How should the TMF be organised?

The organisation of a sponsor file can become quite complex such as when the trial is multi-country or multi-centre. Typically documentation is organised in the sponsor file at three levels: Global, Country and Investigator Site. This is a recommended approach to take as it allows particular sections of the TMF to be made available upon request, although amalgamation of the files may be a more suitable approach to take if, for example, a single country trial is being performed (with combination of global and country level files) or when running a single site study, where all three levels can be combined.

The sponsor organisation should identify where all of the potential documentation that should be filed in the TMF is located. It does not all necessarily need to be in the same location, but it should be clear where it is from TMF procedures/indices etc as the TMF must be readily available both during the trial and during the archiving retention period following the trial. For example, it may be appropriate for an organisation to determine that SAE cases will all be retained in the PV department, including the SUSAR receipt confirmation rather than printing it off the PV database and filing in the TMF. This detail, may, dependent upon its complexity require formal documentation in a procedure (e.g. SOP). In large commercial organisations, the TMF could include documents from across a variety of different departments other than clinical operations, for example, Data Management, Statistics, Pharmacovigilance, Clinical Trial Supplies, Legal, Regulatory Affairs etc., as well as those provided or held by CROs. This contrasts with a small single centre non-commercial trial, where the documentation is likely to be much less and could be limited to just the investigator and pharmacy files. Some documentation may be non study specific, for example the validation of a computer system that is used for numerous trials, but is still needed to demonstrate the quality of the trials.

Version 1: 17th December 2012

[MHRA Super Moderator]

6. Is it necessary to have SOPs and standard indices for the TMF?

The extent of the potential distribution of TMF documentation and the number of trials a sponsor conducts should inform the decision as to whether a formal procedure is necessary to document the locations of all records associated with the TMF as well as the content of the TMF (e.g. standard indices) or whether the protocol providing an overview of the TMF arrangements together with a trial specific index in the TMF is sufficient. The use of a formal procedure, such as a SOP, and a standard indexing system (rather than creating “study specific” indices repeatedly) in organisations sponsoring several trials is recommended as it tends to facilitate compliance.

It is essential to have a suitable indexing system in place for the TMF. This ensures that the documentation is appropriately sorted and filed, which facilitates audit, inspection and trial management. It is recommended that the documentation is filed in date sequential order to enable a clear audit trail of, for example, communications with sites, REC and Regulatory Authorities.

Investigator site files are typically indexed and organised by the trial sponsor and maintained with the assistance of the trial monitor. As a result an investigator who is working with several sponsors may have several different files organised in slightly different ways. Some non-commercial trials may not provide investigator site files and in such circumstances the investigator will need to organise the filing themselves. In such cases, if the host organisation (NHS Hospital) has a standard indexing system available for use, this would prove useful to the investigator.

Version 1: 17th December 2012

[MHRA Super Moderator]

7. How should the TMF be stored and who should have access?

The sponsor’s TMF is the repository of all the information that is necessary to reconstruct the trial and therefore its security and maintenance is vitally important. The arrangements for who should access the TMF in order to add or remove documentation are important points to consider and must be controlled on a risk based approach. The two key factors that influence risk are:

• Number of trials the organisation is conducting (creation of a specific department to manage TMFs may be appropriate compared with an investigator maintaining their own TMF in their office).

• Use of the trial data (if the trial is to be used to support a marketing authorisation or will be important in determining treatment via its publication).

The investigator’s TMF should be stored securely such that only study staff (and monitors, auditors and inspectors) can gain access to the documentation.

Some organisations have controls in place for placing or removing documentation from the TMF whilst the trial is in progress. It can take the form of formal submission of TMF documents with some form of TMF tracking. For example generation of a submission record or a simple signature and date on the index of the file. Some organisations may archive the documentation on an ongoing basis to prevent loss, particularly where eTMFs are in use. The risk of a lack of control would obviously be missing documentation at the end of the trial. It is recommended that this risk and the implications of the loss should be evaluated before the implementation of additional paperwork or systems to track or manage the placing of documents in the TMF.

Version 1: 17th December 2012

[MHRA Super Moderator]

8. Should the TMF be kept up to date?

The TMF must be kept up to date, with documents placed in the TMF in a timely manner as the UK legislation states that, “The master file shall at all times contain the essential documents relating to that clinical trial” (SI 2004/1031 [as amended] 31A, (3)). This can greatly assist in the successful management of a trial by the investigator, sponsor and monitor. The contemporaneous nature of the TMF may vary in trials and could be more anachronistic in trials that have more complex TMF arrangements with multiple parties involved. In this scenario it may be useful to define the timescales for submission and filing of documents to the TMF in procedural documents or TMF plans. Documentation in the TMF that is relied upon for subsequent activities should be in the TMF prior to the activity taking place, for example, monitoring visits rely on the information in the previous report, so the previous report should be completed and filed in the TMF prior to the next visit. MHRA GCP inspectors would raise concerns if the TMF appeared out of date such that the ability to manage and oversee the trial conduct was questionable. It should be remembered that MHRA GCP inspections can be unannounced or performed at short notice and even routine inspections may not contain a notification of the trials to be inspected until a short period prior to the inspection, thus the TMF must be maintained in an “inspection ready state”. This particularly applies to TMFs after the trial has completed, as these TMFs must be “complete and legible” (SI 2004/1031 [as amended] 31A, (7)).

There should be a process to ensure that any documents that have been, or are currently contained, in the TMF have traceability of any changes made to them to comply with UK legislation (SI 2004/1031 [as amended] 31A, (6)).

Version 1: 17th December 2012

[MHRA Super Moderator]

9. What is regarded as an essential document?

Any documentation that facilitates evaluation that the trial has been conducted in accordance with EU/UK legislation and guidance documents must be retained in the TMF. Often encountered on inspection is the view that if a document is not listed in ICH E6 guidance as being an ‘essential document’ then it need not be contained in the TMF. This is not the expectation of GCP inspectors. Firstly, the UK legislation (SI 2004/1031 [as amended] 31A) defines essential documents as follows:

The essential documents relating to a clinical trial are those which (a) enable both the conduct of the clinical trial and the quality of the data produced to be evaluated; and
(b) show whether the trial is, or has been, conducted in accordance with the applicable requirements of Directive 2001/83/EC, the Directive, the GCP Directive and Commission Directive 2003/94/EC.

Two of the principles of GCP, which are a UK legal requirement (SI 2004/1031 [as amended] Schedule 1, Part 2, (4) and (9)) are:

The necessary procedures to secure the quality of every aspect of the trial shall be complied with.

All clinical information shall be recorded, handled and stored in such a way that it can be accurately reported, interpreted and verified, while the confidentiality of records of the trial subjects remains protected

These principles and definitions together can be interpreted to mean that all records created from following trial procedures as well as those listed in guidance relating to the conduct the trial should be retained to demonstrate compliance. The documentation listed in section 8 of ICH GCP and section 3 of the Volume 10 TMF guidance is a useful guide for the minimum documents that are considered essential, but it is not comprehensive, examples of documents that are essential to reconstruct the trial but that are not contained include, QP certification, the regulatory green light document to release and ship IMP and the database lock documentation. The ‘essential documents’ list can therefore be regarded as a minimum subset of the potential documentation that could be regarded as essential for reconstruction of the trial conduct. This list should not be used as a definitive checklist for TMF content. It is important to note some documentation listed in ICH GCP or Volume 10 may be inappropriate, such as a randomisation schedule, or have been replaced for a particular trial, such as an investigator’s brochure for trials of marketed product where the SPC is being used. Where a risk adaptive approach is being followed, some documents listed may not be in the TMF, for example IMP temperature storage records, IMP accountability, laboratory reference ranges etc. and the rationale for this would be in the trial risk assessment.

Additionally, during inspections, verbal information provided during interview, particularly at investigator sites, reveals activities that occurred that were not transparent from the documentation in the file, such as the training that the principal investigator provides to his/her team members. It is therefore recommended that an assessment of all activities is undertaken to determine whether they need to be documented to enable reconstruction of the trial conduct from the paperwork alone. This is an important point to consider, because if a trial needs to be reviewed by the regulatory authority some years after its completion, personnel involved may no longer be available, therefore the TMF must be a robust record of all aspects of the clinical trial.

Version 1: 17th December 2012

[MHRA Super Moderator]

10. How should we deal with correspondence, in particular, emails?

The conduct of a clinical trial can generate large of amounts of correspondence such as emails, letters, meeting minutes and telephone call reports, which may be internal within an organisation (e.g. sponsor, vendor or investigator site) or between organisations (e.g. sponsor and CRO, sponsor and investigator). Correspondence is an important component in reconstructing the trial conduct with some CRO organisations relying solely on email correspondence to confirm sponsor approval of processes, documents, and decisions. Only relevant correspondence that is necessary for reconstruction of key activities and decisions (for example the medical monitor allows an ineligible subject to remain in the trial) or that contains other significant information must be retained. If some correspondence is not retained (particularly certain emails which may not add any value by being retained, such as, email correspondence between investigator site staff and the trial monitor discussing holidays or suitable hotels to stay in near the site), then there should be a formal process to assist individuals in the evaluation of whether it contributes to the reconstruction of the trial or not. Electronic correspondence may be retained electronically (see question concerning eTMF).

It is recommended that correspondence is effectively organised, for example, it could be segregated into topic area (e.g. protocol development). It is essential that some correspondence is filed in an appropriate section in the file (for example, correspondence with MHRA, REC etc). It is also recommended that correspondence is always filed in chronological order and that duplication of documentation (e.g. chain emails) is avoided. Electronically attached documents that are not present elsewhere within the TMF must be included where necessary for reconstruction (for example reports/data reviewed in dose escalation meetings).

Sections including correspondence must be complete. During GCP inspections it is often seen that only copies of letters received rather than those sent out and received are filed (such as those sent from the sponsor to REC or investigator to REC), such that the TMF only contains half of the audit trail.

For emails, these could be printed on to paper or retained electronically as a pst file (preferable) or a pdf. See also FAQ numbers 11 and 17.

Version 1: 17th December 2012

[MHRA Super Moderator]

11. Is it acceptable for the TMF to be electronic?

Yes, either completely or a mixture of paper and electronic, provided this structure is sufficiently defined. There is an increasing use of eTMFs instead of paper based TMFs in the management of clinical trials and the same controls must be available for managing the eTMF as for a paper TMF in terms of security, control over unauthorised edits and access, ease of retrieval of documents etc.

Some eTMF are really the storage of documents within folders in a computer systems’ operating environment. Such systems have been typically shared areas where everyone has access in the organisation and documents could easily be moved, deleted, replaced etc and with a lack of version control so there is ambiguity over which is the current version and which is a superseded version. These systems would require appropriate controls and security to be in place, which would include role based permissions for activities being undertaken through the use of secure passwords. Accounts should be created and deleted within a formal approval process and in a timely manner. This is analogous to the paper system where the TMF is kept in a secure location, with restricted access. There should also be a back up of the system. A more acceptable solution is where the eTMF is a document management system containing further controls. For example, where the system is being used for approval of documents via a workflow system, password enabled electronic signatures should be used, the system must have an audit trail in place to identify date/time/user details for creation, uploading, approval and changes to a document. It would also be recommended that a system for locking documents or the entire eTMF is considered to prevent changes to documents, such as when the eTMF is archived. Finally, all members of staff involved in the conduct of the trial must receive training in using the system and this should be documented.

The eTMF could contain digital documents in their original format, potentially with digital signatures, or records that have been converted from another format (such as paper documents that have been converted to digital images, which may contain wet-ink signatures). Records that only exist in a digital format are often only printed onto paper at the time of an inspection as they normally are accessed in their digital form. This can result in loss of any version control applied in the computer system (e.g. dated filename). An example of this could be the “edit checks” specification used by data management. Such documents, when uploaded into the eTMF may also have the same potential issue which re-enforces the requirement that all documents should have clear version control which is maintained on transition from one media, or from one system, to another.

Unlike a paper TMF, documents loaded into the eTMF will require the addition of metadata to enable the document to be identified within the system. Metadata is that data that is associated with the document (but not the document itself) so typically would be identifying codes for the trial, site, protocol, product etc. The type of metadata is recommended to be formally defined to ensure consistency across all documents.

The eTMF will require validation to demonstrate that the functionality is fit for purpose, with formal procedures in place to manage this process and for change control. The validation of the system should follow previously published standards. The documentation for this process must be retained.

There is also the potential for the investigator site file, held by the principal investigator, to become electronic, with the system either provided by the sponsor, a vendor or by the host site organisation. The documentation in the investigator site file will contain source documents, for example, subject screening and identity logs, consent forms, drug accountability records etc, and the control of these must remain with the investigator unless specific consent had been obtained from trial subjects for their personal information to be released from the trial site. A situation where all the site records are sent to the external sponsor for uploading onto an eTMF system, which the investigator then accesses via a portal, would breach this requirement. The sponsor should consider the EMA GCP Inspectors Working Group Reflection paper on expectations for electronic source data and data transcribed to electronic data collection tools in clinical trials (EMA/INS/GCP/454280/2010), as the considerations and recommendations will have applicability to source documents contained in eTMFs.

Version 1: 17th December 2012