Duplicate Documents: How to Avoid Seeing Double

March 8, 2017

  • eTMF Resources

If you have attended a conference presentation where an MHRA representatives spoke about TMF and TMF inspections, you will no doubt remember the speaker saying:

Duplication of documents should be avoided.

You can certainly appreciate the frustration they will feel if they encounter the same document several times during the course of an inspection and have to determine themselves if it is a duplicate or if there are subtle differences.

What are duplicates? 

This part may sound like a dumb question.  But consider:

What if the same document applies to several, but not all, sites?  Should it exist one, or several times, in an eTMF?

Ideally, the document should only exist once in the eTMF, and be linked to other locations.  This will minimize the time spent to upload and review the document, and allow an agency inspector to look at it only once.  It should be obvious which sites it applies to.

If your eTMF does not support this way of handling a document, you will be forced to duplicate the document.  In this case, if the document must be updated, take care to update all instances of it.

What about program/product level documents such as IBs or safety documents? 

Again, ideally, this document would exist once and be linked into relevant studies (these might not be all studies – e.g. you would not want to link a new version of an IB into a study that is going to be closed tomorrow.)

What if a document is of the same type and has all the same metadata?

That means it may be a duplicate – not that it is a duplicate, even if it has the same date.  For example, it is unlikely but possible you receive two separate IRB approvals on the same date, but they are actually different documents.  Only a person can make this judgement.

What if multiple documents have the same hash code (this is a computed value based on the unique contents of the document)?

This is an almost certain indication that the document is a duplicate.  However, the converse is not true – documents having different hashes could be duplicates.  This will happen if the same document is scanned twice.

Who is responsible for detecting duplicates, and when?

The best strategy is to detect as early as possible – upon upload of documents.  At this point, the eTMF should alert the submitter to any potential duplicates, provide the opportunity to view the potential duplicate(s), and to decide whether to continue with the upload or cancel.  If the eTMF does not have this capability, the submitter may need to query for potential matching documents manually, which can be time consuming.

Alerts during QC are also helpful as a second check or to detect duplicates which entered the system through some automatic process.  QC specialists should be trained in whether they need to do specific checks for duplicates and how to handle them.

The worst approach is to scan the TMF for potential duplicates as this is time consuming, error prone, and could never really be over until every document is received.

As with most TMF business processes, avoiding duplicate documents is a combination of technology and business process.  It’s important to understand your eTMF’s capabilities, take advantage of them, and build manual steps to work around any deficiencies.