Essential Documents: How Much is Enough?

Sponsors with whom we work note a trend among Contract Research Organizations (CROs) who conduct site monitoring: They are collecting fewer site-level essential documents for the Trial Master File (TMF), and collecting them less frequently.  For example,

1. Start-up documents for sub-investigators added mid-study (CVs, financial disclosures, records of GCP training) are collected infrequently throughout the study, or at the end.
2. Site submissions to the IRB/IEC for protocol deviations and safety reports are not collected.
3. Normal ranges for local labs are not collected.
4. IP dispensation and accountability records are collected infrequently or at the end of the study.
5. Delegation logs are collected only at the end of the study, or one is collected up front and another at the end.
6. Site training records are collected at the end of the study, or not collected after SIV training.
7. Submissions to the IRB/IEC for protocols, ICFs, IBs, and patient-facing information are not collected (approvals are collected).

CROs give various reasons for this trend:

• CRAs are conducting on-site visits less frequently, so there are fewer opportunities to collect documents.
• CRAs are checking delegation logs, training records, and IP records during monitoring visits and noting the checks in their monitoring reports, so there is no need to collect those records on an ongoing basis.
• Logs are “living documents,” and per SOP the TMF is only intended to include final documents.
• Sites in large institutions delegate responsibilities to a long list of staff members; it does not make sense to collect documentation for every person who might draw blood or process samples.
• Site training is the Principal Investigator’s responsibility.
• IRB/IEC submissions do not any value, assuming IRB/IEC approvals/acceptances are filed.
• Normal ranges are printed on lab reports or made available on websites, so there is no need to download them.

We are all for improving efficiency in document collection, so we decided to take a look at this trend from a regulatory perspective and with an eye toward the value of these activities.

ICH GCP E6 R2 Section 8.3 provides a list of documents that “should be added to the files during the trial as evidence that all new relevant information is documented as it becomes available” (emphasis mine).  For sponsor files, this list includes the following:

1. Curriculum vitae for new investigators and/or sub-investigators (with a reference to section 8.2.10, which includes “other relevant documents evidencing qualifications,” typically records of GCP training
2. Notification by sponsor and/or investigator to IRBs/IECs of unexpected serious adverse drug reactions and other safety information
3. Updates to normal values/ranges for procedures included in the protocol
4. Documentation of investigational products and trial-related materials shipment
5. Site signature sheet “to document signatures and initials of all persons authorized to make entries and/or corrections on CRFs”

These five requirements correspond fairly closely to the first five items in our list above, with one exception.  E6 R2 does not require that sites maintain a “delegation log,” only a “site signature sheet,” which documents “signatures and initials of all persons authorized to make entries and/or corrections on CRFs,” a quaint concept. Section 4.1.5 states, “The investigator should maintain a list of appropriately qualified persons to whom the investigator has delegated significant trial-related duties”; however, that list is not enumerated as required documentation in Section 8.0 either for the site or the sponsor.

Thus, we are satisfied that there are explicit GCP requirements for collecting the first four documents on our list throughout the study, but no explicit requirements to collect delegation logs, site training logs, monitoring visit logs, or IRB/IEC submissions.  Is there an implicit requirement to collect those documents?

Section 8.1 of E6 R2 tells us that essential documents “individually and collectively permit evaluation of the conduct of a trial and the quality of the data produced,” and that the minimum list provided in the following sections “should be supplemented or may be reduced…based on the importance and relevance of the specific documents to the trial.”  Thus, we are to be guided by our old friend, the “risk-based approach.”

From a risk perspective, the delegation of authority log is a useful document to collect every time it updates, because it provides the sponsor with a record of who is working on their trial and what those people are authorized to do.  In the event of a serious breach of GCP, urgent safety measure, or product recall, it is the definitive list upon which all other lists (contact lists, CTMS entries) are based. The information on the delegation log is also critical for checking the completeness of Clinical Trial Agreements, 1572s, CVs, financial disclosure forms, GCP training requirements, privacy agreements, and site training.  Approximately one-third of the completeness checks for the site-level Trial Master File cannot be completed without an updated delegation log.  If we discover only at the end of the study that we have no documentation that sub-investigators were submitted to the 1572, received critical training, or completed financial disclosure statements, it may be too late to address those issues.

Similar arguments could be made for the site training log.  Site monitors are supposed to verify that required training is taking place, but the only ALCOA-compliant training documentation is the actual log.  With the log filed in the TMF on an interim basis, the site monitor has documentation to back up the monitoring report.  Without it, gaps can hide until the end of the study–or until a site inspection, if the log is never collected–when it is too late to address them.

IRB/IEC submissions are a different case.  In theory, approval/acceptance communications should be sufficient to document that protocol, ICF, and IB updates “have been subject to IRB/IEC review and were given approval/favorable opinion” and “to identify the version number and date of the documents.” In practice, IRB/IEC communications do not always identify all documents that were submitted, or do not include document identifiers and version dates, or (worse), include their own version numbers. Submission letters prove that documents were submitted; frankly we’re not sure why any sponsor would want to toss away proof that they fulfilled their regulatory obligations with regard to the IRB/IEC.

Another consideration for collecting site documents on an ongoing basis vs. the end of the study is the risk of site closure. PIs are people, too.  They go on medical leave.  They die unexpectedly.  They clash with owner of their site network and quit in a huff.  They get divorced, buy a boat,  and sail to the Seychelles. Bottom line – sponsors and CROs need to take a risk-based approach to ongoing collection of site documents that aren’t explicitly required by GCP.