Q: Is it acceptable to printout emails for the trial master file? How should emails be archived?
Q: What format should be chosen for the electronic record?
A: Where possible, the archive format chosen should be one – such as Extensible Markup Language (XML) – that permits access to the record content over the entire period of its retention time without requiring data migration as software is updated. The use of proprietary formats such as Outlook will require the ongoing monitoring of compatibility of future software versions with the original archived records to ensure ongoing retrievability and integrity over the retention period.
Q: Our emails are all on backup tape. Isn’t that adequate for archival purposes?
A: It is recommended best practice that emails for long-term retention are transferred from the email system and archived in an appropriate electronic repository or document management system. Any email copies remaining in the email system should then be deleted. However, copies of those emails will inevitably reside on back-up media. Since this is not a controlled archive this solution presents the additional risk that records may be retained longer than they should. They must then be located and retrieved if requested in legal discovery.
It is important to ensure that the system used for rotation and purging of back-up media is consistent with the email retention policy (as opposed to any generic security back-up practices) and that it is implemented by the IT department. Any audits of records retention procedures should include an audit of back-up procedures to ensure that deleted emails are not retained beyond their required lifetime.
Q: ICH GCP says to retain “relevany correspondence”. How do I decide what is relevant?
(a) records which fall within the scope of the ICH definition of a “relevant communication” and must therefore be included in the essential trial master file, and
(b) records which records fall outside this definition and may be managed as transient communications that are not necessarily part of the essential trial master file.
Although the definition of “relevant correspondence” can be ambiguous and open to interpretation, it is possible for clarity to provide clear guidance to staff in order to minimise the volume of records that may be retained unnecessarily. The GCP-RMA has provided its own guidance on this topic; their “best practice” paper can be accessed online at http://www.gcp-rma.org/.
Q: Some email trails can get very long and complicated. Do we have to keep everything?
A: In order to reduce the volume of records retained, it is preferable to minimise the retention of duplicated email trails where possible. However, in doing so, there is a risk that the meaning of the email content could be misinterpreted if earlier correspondence is removed.
The GCP-RMA has reviewed various different ways of minimising duplicated email trails but acknowledges that in reality, any solution is very difficult to implement. It would be almost impossible to change the now common practice of using the “reply” option which includes all previous emails in that “string” of emails.
Ideally, staff should use the same writing style that is common practice for hard-copy communications, including the use of quotation marks and citations, where appropriate. When replying to an original email, he/she should cite the initial correspondence in the same manner and style as in a hard-copy letter, citing the date of the initial communication and using references and quotes that clearly relate to the content of the reply. In this way, the original communication and any redundant responses by others are excluded in the replies.
Realistically, it comes down to educating staff so that they apply to email those same writing skills expected in any other formal correspondence.
Q: Are email ‘read receipts’ acceptable proof that a document has been received?
A: In some circumstances, it is necessary to retain confirmation that a recipient has received and/or read an email; for example, to demonstrate that somebody has received an electronic ’Dear Doctor’ notification. However, we do not advise using automatic read receipts. Some email systems disable such receipts whilst others may dispatch a read receipt if an alias opens the email rather than the intended recipient. Such a receipt cannot be taken as confirmation that the intended recipient actually received and read the email.
Taking best practice from the hard-copy environment, it is not acceptable to simply assume that the recipient received the communication; there must be independent action taken by the recipient to confirm that he/she has received and read the communication.
Therefore, if you need to obtain and retain proof that an individual has actually received and read a specific email, you should request that the recipient send back a separate email to acknowledge unambiguously that the email was received and has been read.
