It is frequently mentioned that the majority of costs in litigation are the result of the document review process. These costs are often exacerbated by the efforts undertaken to avoid production of privileged work product or communications.

Privilege review typically involves a document-by-document review to fully identify the privileged communications and work product that will be withheld from production to the other parties involved in a matter.

This document-by-document review is critical to mitigating the risk of privileged information being produced.

Inadvertent privilege production can potentially be used against the producing party in the matter in which the inadvertent production occurred or in other cases and may go beyond the particular document at issue to other communications involving the same subject matter.

With the ever-increasing volume of electronic data, the challenges and associated costs of privilege review to avoid inadvertent privilege production become greater and more complex.

The recently-signed Federal Rule of Evidence 502, attempts to address the costs of review while preventing against the waiver of privilege in federal court matters.

The Rule offers definition of intentional waiver versus inadvertent disclosure. Rule 502(b) specifically seeks to protect against privilege waiver depending upon:

  •  whether the waiver was inadvertent,
  •  whether reasonable steps were taken by the producing party to avoid inadvertent privilege production, and
  •  whether steps were taken to promptly correct the inadvertent production, including enforcement of any Clawback or Quick Peek Agreement as provided by FRCP 26(b) (5).

The Explanatory Note on Rule 502(b), prepared by the Judicial Conference Advisory Committee on Evidence Rules, specifically calls out the use of technology as a potential “reasonable step” in avoiding inadvertent production: “Depending on the circumstances, a party that uses advanced analytical software applications and linguistic tools in screening for privilege and work product may be found to have taken ‘reasonable steps‘ to prevent inadvertent disclosure.” 

Indeed, there is a growing body of case law which supports the need to utilize technology in the document review process and to understand the application of that technology in helping to identify responsive data and protect against inadvertent privilege waiver. 

The Explanatory Note also references the value of an “efficient records management system before litigation” as a potential “reasonable step” in the avoidance of privilege waiver.

So, what might be considered “reasonable” steps in preventing inadvertent privilege production? As the Explanatory Note suggests, “reasonable steps” may begin long before actual litigation commences. They may, in fact, begin with an organization‘s records management system.

As an example, an organization utilizing an automated records management system for e-mail may have the ability to tag potentially privileged information within that system as part of the records retention protocol.

As a result, those privileged communications can possibly be excluded at the time of collection for litigation. Or, the privileged items in the records management system can be collected for attorney review with the privilege “flag” already assigned, saving attorney review costs in litigation and increasing the likelihood that privileged communications will be identified as such.

In addition, many organizations are beginning to use early case assessment tools to reduce the overall volume of documents ultimately reviewed by outside counsel. The extent to which these tools can also be used to identify privileged information prior to collection of the potentially relevant universe of documents is valuable to avoiding privilege waiver and may constitute a reasonable step in avoiding errors in production.

Further, as we continue to see greater convergence of enterprise search technology with litigation support/electronic discovery collection—expect to also see enhancements to collection processes and the technology that allow privileged documents to be identified at the time of collection and adverted from the responsive/non-responsive review process.

Regardless of the procedures in place during a document review to protect against any waiver of privileged information, it is a best practice to negotiate a Clawback Agreement, as provided in the amended FRCP Rule 26(B)(5). Alternatively, the Rules provide for a “Quick Peek” option, allowing a party to make documents available without a non-responsive or privilege review – and without waiving privilege - prior to fully reviewing and producing. In my experience, Clawbacks are the best practice and “Quick Peeks” are not widely used.

Certainly, the use of technology and a managed privilege review process constitute reasonable steps in preventing privilege waiver. In fact, no privilege review of ESI should be undertaken without consideration of the review protocol, the technology implemented and the quality control measures in place to avoid inadvertent production. Indeed, technology can help to identify potentially privileged documents, and it can be used to quality control the effectiveness of the privilege review.

However, documented review procedures are also critical to avoiding inadvertent production of privileged material. In selecting an electronic document processing and review approach, the legal team (attorneys, paralegals, litigation technology specialists and clients) should assess the following with regard to the technology: Does the technology not only allow you to eliminate exact duplicates (de-duplication), but does it provide near duplicate and/or e-mail thread suppression capabilities?

Near duplication gives the ability to identify, for example, drafts of a contract – some of which may contain attorney edits of contract language and could potentially be privileged.

E-mail thread suppression allows you to suppress all e-mails in a chain except for the final e-mail. E-mail chains are often the most challenging to review – and can be a source of inadvertent production - as privileged communications may exist within the threads of an extended e-mail communication among various e-mail participants.

The ability to review the final string only – and apply privilege calls related to exact copies of that string or to other copies containing portions of the same string – ensure consistency in the privilege calls and avoidance of privilege production.

Does the technology allow native review? While a few years ago native review was less common, most review platforms allow for native review. But, the ability to be able to review the native version of a document means you are more readily able to see “track changes”, comments, and other metadata which may, in fact, contain privileged information. This is, of course, critical to consider if an agreement to produce native files is in place.

Does the technology vendor processing the electronic data have a standard protocol in place for processing of native files that helps prevent against inadvertent privilege waiver. For example, how does the processing vendor handle “track changes” or comments when converting files to image format for an imaged-based production – are they accepted or revealed in the imaged version of the file?

How does the technology handle production of partial document “families”? To illustrate, if the review protocol calls for only the privilege documents in a family to be withheld while the responsive / non-privileged documents are to be produced, does the technology have the ability to produce only the non-privileged family members?

Does the review platform provide concept grouping technology, increasing the likelihood of documents with similar content (i.e. privileged information) being grouped together for consistent privilege tagging during the review process?

Does the technology provide a “social network” view so communications between lawyers and client might be easily segregated for review?

Does the technology highlight search terms in a document so the review team can easily see why a document is tagged as potentially privileged?

Does the review platform include audit mechanisms that allow you to view who tagged a document as privilege in the review history?

In addition to the above considerations regarding the overall technology platform, technology can be used in other ways to prevent inadvertent production of privileged material. Most commonly, keyword searches for privilege terms, such as “privileged and confidential” or “attorney-client communication”, and attorney and law firm names will be employed to flag documents likely to be privileged.

It is important to consider the following in ensuring privilege waiver does not occur: Not only should searching be conducted for law firm or attorney names, but searches should also include law firm email domains In conducting privilege searches is the technology provider searching both metadata fields and the text of the document for privileged terms?

In the case of e-mail, is header information also searched? Some E-discovery vendors have developed their own cumulative (and proprietary) list of privilege terms and law firm names that may allow for a more comprehensive identification of privilege documents.

Does the search technology allow you to exclude standard disclaimer footers in e-mails?

These disclaimers often contain terms such as “privilege and confidential”, which may lead to many false hits on potentially privilege communications if the privilege search term list includes “privilege” and “confidential”?

How are non-searchable files (i.e. image files without searchable text) segregated for privilege review? Can foreign language documents be searched and segregated for privilege review?

Increasingly, the importance of a managed and documented review protocol is critical to ensuring that not only are the overall costs of review most efficient, but that privilege waiver is avoided.

Again, the following may constitute reasonable steps in avoiding privilege waiver: Development and documentation of privilege search terms for use in flagging potentially privilege documents.

This is critical not only for the legal team‘s future reference but, if necessary, to demonstrate to the Court the methodology used in taking reasonable steps to avoid production of privileged material. Training of review teams as to what constitutes privilege in a particular matter.

Does a fax sheet indicating a document was sent to an attorney require that the entire document or family is privileged or is the fax sheet redacted and the document produced?

Development of a protocol regarding treatment of privilege family members when producing e-mail and attachments. Is the entire family withheld? Are the privileged documents redacted or are the privileged documents withheld and partial “families” are produced?

Does the review protocol include a “Redaction” category so documents containing privileged information are flagged at review time for redaction of the privileged portions.

How will the privilege review be handled? Will first pass reviewers flag privileged documents and confirm potential privileged items identified through the privilege searches? Or, will all potentially privileged items automatically go to a more senior group of attorneys for initial review?

Will documents be flagged for privilege regardless of responsive/non-responsive status? How – or by whom – will foreign language and non-text searchable documents be reviewed for privilege content?

Quality control is also very important in ensuring against inadvertent production. Quality control measures should be implemented throughout the privilege review: Validate the privilege terms used in flagging potentially privileged items by verifying the validity of those terms.

Run a search on the privileged documents identified early in the review process – were they all identified through the privilege keyword searches prior to review?

Does the privilege keyword list need to be revised?

Use “Find Similar” or other near duplicate technology to find documents similar to the confirmed privileged documents and spot-check to ensure the consistency of the privilege/non-privilege calls Run searches on Responsive documents to verify no privilege family members have accidentally been included in the production set.

Be sure to also look for responsive items that were flagged for redaction and ensure they have been directed through the proper work flow.

Be sure to document your production search to prove that it included exclusion of privilege-tagged documents. Randomly check first-pass review team members work product and analyze the number of privilege calls each reviewer is making.

This allows you to determine whether they are applying privilege tags too liberally or insufficiently. Provide feedback to the review team on the appropriateness of privilege calls.

These checks should be conducted throughout the life of the review, not just at the beginning of a review effort.

Conduct a final review of the privileged documents when preparing the final Privilege Log.

While by no means intended to be exhaustive, the above should help when planning a privilege review and ensuring that “reasonable” steps have been taken to avoid inadvertent production of privileged information.

A combination of pre-litigation records management and early case assessment/collection tools employed by a company, the use of technology to enhance accuracy of privilege identification, and the use of managed review protocols are all “reasonable steps” in protecting privilege in this age of electronic discovery. 

This article was originally published in Law360, New York (September 30, 2008).