Leonard Deutchman
Leonard Deutchman ()

Editor’s note: This is the second in a two-part series.

It has been more than 18 months since the American Bar Association amended its comment to Rule 1.1, pertaining to competence, to require that lawyers “keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology.” In my last column, we discussed how to learn about e-discovery, some general rules involving it and how to build a map of the data universe from which you will have to produce responsive documents and withhold privileged ones. In this month’s column, we will discuss the ethics of preservation and collection, searching and review.

Preservation and Collection

The first step a party must take with regard to e-discovery is to preserve its electronically stored information. Counsel’s ethical duty is to advise its client as to how and what to preserve.

Qualcomm v. Broadcom, Case No. 05cv1958-B (BLM) (S.D. Cal. Jan. 7, 2008), makes this painfully clear. The attorneys representing Qualcomm found themselves referred to the state disciplinary board for failure to advise their client and oversee discovery properly. The court and, later, the state bar found that leaving to the client the choice of what files to collect for review was an abdication of counsel’s ethical duty. The determination of what is responsive to a discovery request is counsel’s to make, not the client’s.

The attorneys were eventually exonerated, as it was revealed that the client had withheld information. The takeaway from Qualcomm, nevertheless, is that counsel must control the discovery preservation and collection process, and oversee it closely.

There are two separate but equally important reasons that counsel must take control. One reason, illustrated by Pippins v. KPMG, 11 Civ. 377 (CM) (S.D.N.Y. Oct. 7, 2011) (appeal denied), and Vaughn v. LA Fitness, 285 F.R.D 331 (E.D.PA. 2012), is that smart, informed answers win, and counsel cannot provide a smart answer without knowing everything, which here means the client’s potentially discoverable materials. The ethical duty to control and oversee discovery preservation and collection is inseparable from the best practices counsel needs to follow to prevail.

In Pippins and LA Fitness, discovery was stayed prior to class certification. In Pippins, KPMG complained that it had to preserve 9,000 hard drives at a cost of millions of dollars, and moved to preserve only 100 hard drives that, it claimed, would be representative of the 9,000. In LA Fitness, LA Fitness complained that the cost of preserving all records would be more than $500,000. Relying upon the discovery stay, however, KPMG failed to examine the drives to provide any evidence to support its assertion, and so the court denied KPMG’s motion because it failed to present testable, scientific evidence. KPMG could rely upon the discovery stay or make its motions, the court reasoned, but could not do both.

LA Fitness, by contrast, prevailed when it made a similar motion to KPMG’s and was in the same procedural posture. LA Fitness complained that the cost of preserving all records would be more than $500,000, but, unlike KPMG, it conducted extensive review of its materials and showed that very few responsive records were to be found in the volumes being preserved. The court held that if the plaintiffs wanted to engage in further discovery, they would have to pay for it.

The second reason counsel must control preservation and collection is that failure to do so can lead to the deadliest claim in discovery production: spoliation. Counsel owes this duty not only to its client, but to itself. In the recent Procaps v. Patheon, Case No. 12-24356-CIV-GOODMAN (S.D.FL. Feb. 28, 2014), the court, in awarding the requesting party attorney fees for having to litigate the spoliation motion, found the producing party’s counsel so derelict in failing to oversee the implementation of a litigation hold and the collection of data that it ordered that half of those fees be paid by the producing party, the other half by its counsel.

What must you know to preserve? You must know when the duty to preserve arises, where the universe of data to be preserved resides­—which means creating the data map discussed in last month’s column—and the best way to preserve.

You can preserve in place by notifying all data custodians, in writing and using simple, clear language, that there is a litigation hold and nothing subject to it can be deleted. The best way to promulgate the hold is to have it as a log-in screen on each user’s workstation that the user will have to read and acknowledge. IT can track each acknowledgement and produce a log should the issue arise.

If you do not trust preservation in place, the best way to preserve is to collect the data. The best way to do that is to have IT or a vendor make forensic images, i.e., exact copies of the media imaged, which capture everything on the media, so if two years after collection there is a question of whether everything was produced, the client can always return to the forensic image.

Processing, Searching 
and Review

The next step in e-discovery production is to process the collected files into a searchable review database. The review platform allows reviewers to tag files as responsive, nonresponsive or privileged, as well as with other tags customized for the matter.

Processing and review also generally involve searching, so as to try to locate the potentially responsive material and eliminate review of nonresponsive files. Presently, there are two categories of searching. The simpler searches include keyword, Boolean (X but not Y), proximity (X within five of Y), and fuzzy (typical misspellings or spelling variation) searches, as well as date restrictions, while the more complex, technology-assisted review searches include email threading, predictive coding and concept searching.

With regard to searching, you discharge your ethical duties in two fundamental ways. The first, as with all e-discovery matters, is to know both the matter and the technology so that you will know how the technology applies to the matter. To determine whether to use predictive coding, for example, you must know the cost compared to attorney review, whether anything in the matter does not lend itself to predictive coding (a complex patent matter, for example, may not), and whether litigation problems—production deadlines, for example—are solved or made worse by using predictive coding.

The second way is to position yourself best to avoid challenges from the requesting party as to how the searching was conducted, or prevail if challenges are unavoidable. Challenges are costly to litigate and, if you lose, re-searching is even more costly. On both the macro and micro levels, the best way to position yourself strongly is to invite input from opposing counsel and test your data.

At the macro level of searching, you will want to see what sources of data you can eliminate from having to review at all. Imagine you have collected the data from those 15 sources discussed above. At the micro level, you will want to proceed with search terms to reduce the dataset to be reviewed, or with TAR to review the data set.

At the macro level, you will want to rank your custodians in order of importance to the matter. Solicit input from requesting counsel. Then sample those files. If you can show, using representative samples, that some sources contain files of marginal value, you can seek agreement that those sources need to be reviewed. You may obtain agreement only by including a source you think has nothing to add, but what you save your client, in review and potential litigation costs, is worth the compromise. If you cannot reach agreement, you will be in the best position to defend your position before the court.

At the micro level, with search terms, invite input from opposing counsel. If the offered terms are too expansive, test them, or some of them, and provide statistics as to how many false positives they yielded, as well as files that, while not duplicates of others, add nothing new to what is being produced. With TAR searches, when you have refined your coding or concept search results, your results have stabilized, but you have not deployed across all files, remove privileged files and offer to share your results with opposing counsel.

Sharing results can lead to three outcomes. You can get agreement from opposing counsel, thus mooting later challenges. Opposing counsel can propose further refinements that you accept, leading to agreement. Finally, if you obtain no agreement, you have done everything right—tested your results, shared them with counsel, sought input—and that will put you in the best position to pass the Daubert test and prevail before the court.

It is, nevertheless, possible that you will decide not to share results, because you have concluded that so doing may expose a weakness, a potential new issue or something else. Those litigation issues are rare, but they do arise, and if that is the case, you may decide not to share—as we discussed in last month’s column, one size does not fit all.

Broadening Duty

As e-discovery has faded from being the next big thing, the volume of ESI and the percentage of written materials and communications generally that are ESI in their native format have exploded, and each graduating class of newly barred attorneys has grown up with computers, the requirement that litigators understand e-discovery has broadened. The steps discussed can help you discharge that ethical duty, and win your cases. 

Leonard Deutchman is vice president and general counsel of LDiscovery LLC, a firm with offices in New York City, Fort Washington, Pa., McLean, Va., Chicago, Atlanta, San Francisco and London that specializes in electronic digital discovery and digital forensics.