A Creative Screening of Electronically Stored Information May Determine the VictorAugust 01, 2008 The Corporate Counselor
This article first appeared in the August edition of Corporate Counselor and is posted here with permission.
For all corporate counsel, the prospect of litigation that requires the preservation, collection, review, and potential production of voluminous electronically stored information (“ESI”) can be overwhelming. The time and cost considerations are significant, and many in-house counsel are unsure of the appropriate first steps. One thing is certain: This is not an issue that can be put on the back burner and dealt with in a piecemeal fashion after litigation ensues. The painful results of such an approach were the subject of Magistrate Judge Grimm’s recent decision in Victor Stanley, Inc. v. Creative Pipe, Inc., et al., 2008 WL 2221841 (D. Md. 2008). In that case, the United States District Court for the District of Maryland was forced to decide whether the defendants’ inadvertent production of certain categories of ESI waived the protections of the attorney-client privilege and work-product doctrine. After carefully analyzing the events that led to the inadvertent production, the court in Victor held that the defendants waived any claim to attorney client privilege or work-product doctrine protection with respect to the inadvertently produced documents.
Defendants, Creative Pipe, Inc. (“CPI”) and Mark and Stephanie Pappas (“M. Pappas and S. Pappas”) (collectively, the “defendants”) initially responded to plaintiff’s discovery requests with a traditional “paper production.” The plaintiff took issue with the sufficiency of the defendants’ paper production and the parties ultimately identified a joint protocol for searching and retrieving relevant and responsive ESI. The joint protocol included nearly five pages of keyword/phrase search terms that were intended to help locate responsive ESI, but not designed to identify privileged or work-product protected documents. After completing their search, defendants expressed concern to the court that an individualized privilege review of the responsive documents would unnecessarily delay production and cause undue expense. As a result, defendants provided their computer expert with a list of keywords that could be used to search for and retrieve privileged documents. Defendants’ counsel, recognizing the possibility of inadvertent disclosure, requested that the court approve a “clawback agreement” in the event of such a disclosure. The court’s subsequent decision to extend the discovery deadline by four months convinced the defendants that they would be able to undertake a document- by-document privilege review and, they believed, made a clawback agreement unnecessary — a decision that would ultimately subject them to judicial criticism.
Soon after receiving and reviewing defendants’ ESI production, plaintiff’s counsel discovered documents that were potentially privileged or work-product protected and notified defense counsel accordingly. Defendants asserted that any production of protected documents was inadvertent, but plaintiff filed a motion seeking a ruling regarding the discoverability of inadvertently produced documents. The court undertook a close examination of the defendants’ retrieval and production of ESI, which included the execution of the joint protocol and an electronic search for privileged material on those files that were in a “textsearchable” format. The electronic privilege search consisted of seventy keywords identified by defendant M. Pappas, one of his former attorneys, and another attorney. Additional, non-text searchable files were turned over to defendants’ current attorney for a manual, “page-by-page” privilege review; however, due to time constraints and the need to review thousands of documents, the defendant and defense counsel chose only to review the page titles of the documents. Only if they determined from the title that a privilege might be applicable, did the defendants review a document in its entirety.
The court in Victor immediately identified numerous problems with the defendants’ explanation of their ESI search protocol, as well as the search itself. First, the court found that the defendants were “regrettably vague” in their description of the 70 keywords that they used for the textsearchable ESI privilege review; specifically, the defendants failed to inform the court how the search terms were developed, how they conducted the search itself, and what quality controls, if any, were used to assess the reliability and accuracy of the search. Second, the court questioned whether the defendants and the two attorneys who created the keyword search were qualified to create a search and information retrieval strategy designed to yield a reliable privilege review. Finally, the court in criticized the defendants for failing to assert that they sampled the text-searchable ESI files to determine whether the electronic keyword search was reliably identifying privileged documents.
THREE POSSIBLE APPROACHES
In evaluating the plaintiff’s motion, Magistrate Judge Grimm identified three possible approaches when deciding whether a party waives the protection of the attorney client privilege or work-product doctrine as a result of an inadvertent production. Judge Grimm described a spectrum of tests that ranges from the most lenient to the strictest.
Under the most lenient approach, a court will find no waiver because there is no knowing and intentional relinquishment of the privilege or protection. Under the strictest approach, courts find waiver because once the documents have been disclosed, there is no longer any expectation of confidentiality. Between these two divergent views, Judge Grimm identified an intermediate approach that requires a balancing of factors to determine whether the producing party exercised reasonable care under the circumstances to prevent the disclosure of privileged or protected information. If the court believes that the party exercised reasonable care, there is no waiver. Judge Grimm in Victor evaluated the disclosure and determined that the defendants waived the protections afforded by the attorney-client privilege and work-product doctrine because they failed to exercise reasonable care to prevent the disclosure of these items.
THE VICTOR COURT’S APPROACH
The court in Victor employed the intermediate approach, which relies upon a balancing of the following factors: 1) the reasonableness of the precautions taken to prevent inadvertent disclosure; 2) the number of inadvertent disclosures; 3) the extent of the disclosures; 4) any delay in measure taken to rectify the disclosure; and 5) overriding interests of justice. Victor Stanley, Inc., 2008 WL 2221841, *5 (citing McCafferty’s, Inc. v. Bank of Glen Burnie, 179 F.R.D. 163, 167 (D. Md. 1998)). The court then determined that it need not go any further than the first factor, and found that it “militated most strongly in favor of a finding that Defendants waived the privilege in this case.” Id. Specifically, the defendants had not carried their burden of proving that their conduct was reasonable because they failed to provide the court with: 1) information regarding the keywords used for the search; 2) the rationale for the selection of keywords; 3) the qualifications of the individuals who created the search to design an effective and reliable search and information retrieval method; 4) whether the search relied on simple keywords or a more sophisticated methodology such as Boolean proximity operators; and 5) whether the defendants had analyzed the results of the search to assess its reliability, task appropriateness, and quality of implementation. The court adopted the position that “all keyword searches are not created equal,” and emphasized that proper selection and implementation requires technical and scientific knowledge.
SEDONA CONFERENCE ‘PRACTICE POINTS’
To address the known deficiencies associated with basic keyword searching, the court focused on the Sedona Conference “Practice Points” as a guide for a proper search protocol. These practice points included: choosing a search method that is dependent on the specific legal context (Practice Point 3); performing due diligence when selecting an information retrieval product or vendor (Practice Point 4); recognizing that the use of search and information retrieval tools does not guarantee the identification of all responsive documents in a large data collection (Practice Point 5); making a goodfaith effort to collaborate on the use of particular search and information retrieval methods, tools, and protocols (Practice Point 6); and, expecting that your choice of search methodology will need to be explained, either formally or informally, in subsequent legal contexts (Practice Point 7). Victor Stanley, Inc., 2008 WL 2221841, *6; The Sedona Conference Best Practices Commentary on the Use of Search & Information Retrieval Methods in E-Discovery, 8 Sedona Conf. J. 189). The court in Victor reasoned that compliance with the Sedona Conference Best Practices for search and information retrieval would strongly support a parties’ argument that their chosen method was reasonable and reliable, but the defendants failed to demonstrate the reasonableness of their search, or satisfy the remaining criteria of the intermediate approach. Finally, the court criticized the defendants for failing to pursue a clawback agreement that could have provided them with added security in the event of an inadvertent production of protected documents.
Although it may require the expenditure of time and money up front, a corporation will benefit greatly from the creation and institution of a document retention policy that clearly delineates the company’s position of how (and for how long) documents are maintained, where those documents are maintained, and in what format. The policy should be well thought out, reasonable, and strictly adhered to by all employees. Upon receipt of discovery requests seeking ESI from many years prior to the litigation, a company will be able to quickly assess the request in light of its document retention policy and access material in an orderly fashion. Having a well-documented document retention policy will also provide companies with a ready explanation why an otherwise discoverable document may not have been retained.
Another important component of any company’s e-discovery plan is the creation of a standard form litigation hold notice that can be modified, as needed, for specific situations. Once corporate counsel is aware of the potential of litigation, the litigation hold notice should be distributed to all employees as well as all information technology staff. The notice should explain that electronically stored data could become a key source of information and that the deletion of any material, even that which occurs in the normal course of business, should immediately be suspended and all potentially responsive information must be maintained.
Even for a small company that is infrequently faced with litigation, the creation and institution of a document retention policy and standard litigation hold notice will pay for itself in the long run by saving time and money in responding to discovery requests and by preserving the confidential nature of attorney client privilege or work-product doctrine documents. Without such a pre-litigation protocol, a small company will particularly feel the economical pain of developing and implementing such a policy on the fly in the middle of litigation with the added pressure of an opponent and the court.
Finally, Victor teaches us that all ESI searches are not created equal. When conducting a large-scale electronic document search and production, parties must pay particular attention to ensure that attorneyclient privileged and work-product protected documents are not inadvertently produced during the course of discovery. As the defendants in Victor found, prudence dictates that a company develop a protocol to identify and segregate attorney-client privileged and work-product protected documents, including the clear demarcation of these items with a “privilege” or “confidential” stamp, well in advance of litigation.
By having a set protocol, companies can better avoid the pitfalls of the Victor defendants. In-house and outside counsel should work together and employ the “utmost care” in selecting a search and information retrieval methodology to identify and withhold information protected by the attorney-client privilege and work product doctrine. This includes, among other things, careful consideration of the qualifications of individuals who design the search methodology, quality assurance testing once the methodology has been implemented, and an expectation that a party could be called on to explain and defend its chosen methodology in future proceedings. When faced with litigation, the parties should also produce an accurate and complete privilege log that identifies any documents that are being withheld from production and the basis for doing so. Finally, early in the litigation, the litigants should work together towards the entry of an appropriate protective order or “clawback agreement” in the event that documents are inadvertently produced, despite taking the appropriate steps to prevent that from happening.
Although this time-consuming and expensive process does not guarantee against the inadvertent production of protected documents, abiding by document retention, document hold and ESI production protocols will support an argument that an inadvertent production should not result in a wholesale waiver of the attorney-client privilege or work-production protection. Companies who fail to implement such protocols in advance of litigation run the risk of bearing the increased burden of having an inadvertent production be deemed to be a wholesale waiver. Needless to say, a wholesale waiver could have devastating results on the outcome of litigation.