#CaseoftheWeekCase Law

#CaseoftheWeek Episode 87: Whether a Party Can Utilize TAR after Applying Search Terms

Kelly Twigger discusses on episode 87 whether the Special Master in a class action case agreed to allow defendants to utilize TAR on a review set culled by search terms where the plaintiffs argued for linear review. The matter we analyze this week is In re Allergan Biocell Textured Breast Implant Prods. Liab. Litig., 2022 WL 16630821 (D.N.J. 2022). The decision is from October 25, 2022 and was presided over by Special Master Joseph A. Dickson.

Keep reading or watch the video to understand the eDiscovery issues.


Good morning and welcome to Episode 87 of our Case of the Week series, published in partnership with ACEDS. My name is Kelly Twigger. I am the CEO and founder of eDiscovery Assistant, as well as the principal at ESI Attorneys. Thank you so much for joining me this morning.

We have a couple of events that are coming up that I wanted to let you know about. I’m incredibly honored to be able to host judges William Matthewman and Xavier Rodriguez on the Judicial Panel, which is part three of a webinar series on Emojis. That series is being brought to you by multiple chapters of ACEDS. That webinar was originally scheduled for this Thursday, November 10. This morning, we needed to reschedule it to November 30th at noon Eastern due to Hurricane Nicole. Fingers crossed that the storm weakens and that there’s very little danger to everyone in Florida.

The second event I wanted to let you know about is I will also be moderating a panel on Topics in eDiscovery at the Georgetown Advanced eDiscovery Institute next Wednesday, November 17th. If you’re going to be there, please plan to stop by and say hello.


Let’s dive into this week’s decision, which comes to us from In re Allergan Biocell Textured Breast Implant Products Liability Litigation. This decision today is from Joseph Dixon, who is the Special Master in this case. He is a former federal magistrate judge for the United States District Court for the District of New Jersey.

Now, this decision is likely to cause quite a splash in the world of ediscovery. Our friends over at eDiscovery today, Doug Austin, already included it on his blog from late last week. In this decision, Special Master Dixon denied defendants’ motion to utilize TAR on a set of 560,000 documents that were already culled using search terms.

In denying that motion, the Special Master noted that the parties had agreed to cooperate on discussing TAR and did not, and that the case law does not definitively state whether applying TAR after search terms is universally accepted. The Court here goes into great detail in exploring pre-existing case law on this particular issue, and we’re going to talk about some of that.

As always, we tag each of the cases in our eDiscovery Assistant database with issue tags. And the issue tags for this case include search terms: sampling, ESI protocol, cooperation of counsel, technology assisted review and special master.

All right, let’s talk about the facts of the case. The plaintiffs allege, among other things, that Allergan’s BIOCELL® textured, breast implants and tissue expanders caused the plaintiffs to contract breast implant large cell lymphoma, which is the cancer of the immune system that develops in the area around the implant.

For discovery in the case, the defendants collected and indexed more than 9.3 terabytes of data from 60 custodians. That’s a lot of information, but it’s pretty consistent in these large scale class action litigations. The Court does note here that the parties understood the scope of discovery when they got started.

After collection, the defendants applied search terms to the document set to cull it to a subset for review. There is no specific information in the decision on whether the search terms were agreed to between the parties, but in reading between the lines and the language that’s included in the ESI protocol, it seems like the parties did agree on the search terms that were used to cull the initial collected set to the set for review.

Defendants had already reviewed some of the documents, but asserted on this motion that there are still approximately 560,000 documents left to be reviewed by its eDiscovery vendor. The review of those documents is estimated to take another 20 weeks to be able to complete it.

The parties did agree to an ESI protocol in this case that stated that the parties would confer on both the application of TAR and that the parties would cooperate to identify methods like TAR or search terms and bring to the court any disagreements that they were not able to resolve amongst themselves. Now, the defendants are moving, as we mentioned, to have the court allow them to apply TAR to this remaining set of 560,000 documents. This is after they’ve begun a linear review of a set of documents that was culled by search terms.

The defendants made three arguments in support of their position that TAR should be allowed to be applied to the balance of the documents:

  1. That applying TAR after applying search terms is standard practice and commonly used for efficiency and to reduce costs;
  2. That as the responding party, they are in the best position to determine the best review methodology, and that the best review methodology, based on the large number of documents that have already been reviewed, is to apply TAR to the remaining unreviewed custodial documents after the application of search terms; and
  3. That the case law, including In re Biomet, Livingston vs. City of Chicago, and other cases, allow a party to apply search terms prior to applying TAR, and that the process of doing so complies with the Federal Rules of Civil Procedure.

In response, the plaintiffs made their own set of three arguments:

  1. That applying TAR to the unreviewed documents is prejudicial and unreasonable because it will exclude documents from review and reduce the “efficacy and accuracy of the process”;
  2. That defendants have not provided any evidence in support of its contention that applying TAR will save time or money; and
  3. That the court should follow In re Valsartan and other case law that precludes the application of TAR following the culling by search terms.

Those are the arguments in front of the Court.


Let’s talk about the analysis that the Court undertakes. The Special Master begins immediately by addressing the notion of whether defendant’s proposal is consistent with previous case law, several of which we’ve covered on the Case of the Week. The Special Master states specifically:

There is no such general principle espoused by the courts or the commentators. That is not to say that cases do not exist where the parties are permitted to apply TAR after culling by the application of search terms. The courts find solutions to the problems confronting them, but do not settle the question of which method is better.

What the Court essentially says is there is no consistent position from the courts on whether TAR can be applied after the application of search terms. Courts have addressed the problems that have faced them, but there is no definitive statement from the courts on which approach is better applying search terms first and then TAR, or applying TAR to the entire set of documents prior to applying search terms.

The Court then undertakes an analysis of the different cases that are cited by the plaintiffs and distinguishes all of them. The Court starts with the In re Biomet ruling, where Biomet provided a cost benefit analysis to the Court on why TAR should be applied. That kind of analysis is not present here. The defendants in this case do not present any kind of cost analysis that was present in the In re Biomet case.

Here defendants have only provided an affidavit or a declaration of cost that had been incurred to date. The Court also notes that the Biomet court specifically stated that, “it was not deciding whether predictive coding or keyword searching is better. Rather, the issue was whether Biomet had satisfied its ediscovery obligations.”

So again, the Court is pointing out to defendants that there is no case law that simply stands for the proposition that applying TAR after search terms is a valid and enforceable way to go.

The Court also questioned the defendant’s position that it would have to recollect the entire 9.3 terabytes of data to start over the process with TAR. The defendants made this argument that they would basically have to recollect all of their data, and it is a mystery to the Court, and to me, frankly, why that would be the case.

Regular practice, best practice, is always to keep the full set of collected data separate and apart as a backup so that you can reload it at any time that’s necessary. It would seem to me that if you were going to start over—which I’m not sure makes sense and I’m not advocating for it—that it would not be difficult to do so. That seems to be a bit of a ludicrous position, and it likely undermines defendants other arguments in front of the Court.

The Court also looks back at some of the case law and says that the age of some of the cases cited by the defendants are telling. Cases that are from eight to nine years ago, represent, “a lifetime in the world of technological development and the advancement of TAR.” And that’s really true.

When the Bridgestone Arms case came out that the defendants rely on, we were still talking about TAR 1.0. That’s a completely different technology than TAR 2.0 that we’re talking about now, or even the predictive coding that came in between.

The Court then looks at In re Valsartan. The Court takes the overriding principle that, “TAR requires an unprecedented degree of transparency and cooperation among counsel in the review and production of ESI responsive to discovery requests.” Basically, what the Court does is say okay, defendants, you guys are telling me to rely on all these cases. But what these cases actually say is, if you guys are going to engage in TAR, it requires a lot of cooperation, and let’s look at the facts that we have here and see if that’s what we have.

The Court then says the parties didn’t agree to the application of TAR, and it looks to the case management order and the ESI protocol in the case for the language. The Court’s case management order required cooperation between the parties—both to come up with an ESI protocol, but also to continuously cooperate ongoing throughout the ediscovery process. The ESI Protocol, agreed to by the parties, states that the parties will use search terms for filtering and that the parties will confer on using TAR or other methodologies, and where they cannot reach agreement, they’ll seek court intervention.

Based on looking at those two documents, the Court finds that the Defendants had not set forth an adequate basis for ordering the application of TAR after the applications of search terms and therefore denies the defendant’s motion. Finally, the Court states that:

Applying TAR to an already reduced via search terms set of documents will reduce further the identified responsive documents and will certainly not reveal documents that the application of search terms has precluded. Because plaintiffs did not bargain for this at the outset over a year ago, it is inappropriate to force them to accept it now.


What are our takeaways from this case? Well, we talk about this regularly on the case of the week, and the first one is plan, plan, plan. What you agree to in your ESI protocol will be determinative of how you can proceed in your case specifically regarding search terms and TAR. If you agree to search terms or to cooperate in negotiating on TAR, you will be held to that agreement by the Court.

When you read this decision, you’ll see that there is no discussion of Rule 26 or Rule 34. There’s no discussion of the Federal Rules of Civil Procedure at all because the parties made agreements here, and those agreements are what control.

Now, we’ve had some pretty spirited debates at various events, including the University of Florida eDiscovery Conference that’s coming up in February, about whether or not an ESI protocol is in the party’s best interest. Issues like TAR and search terms can certainly tip the balance of the argument against, but you can also decline to agree to those issues and simply rely on the producing party’s obligations under Rules 26 and 34. There is no rule that requires parties to agree to use search terms or TAR, or the process for either one, only that a producing party meets its obligations.

So think about that when you’re negotiating your ESI protocols. Think about what your position is relative to the other side. In these large class action litigations, it needs to be a very thoughtful process as to how discovery is going to be handled and how you’re going to meet these cost obligations that your clients are going to have.

The biggest issue here for the defendants was the complete lack of evidence that was submitted to the court to show the cost of ongoing linear review and the potential savings of TAR. In addition to that, defendant’s reliance on case law where those facts that were in the record showing a cost benefit analysis highlighted the lack of any facts being on record here. So that In re Biomet case, on which the defendants tried to really hedge their bets, the defendants put in information about the efficiencies that would be created by TAR. No such evidence was presented to the court here, and as such, there really was no basis for the court to be able to order TAR at all.

I’m not really sure why the defendants didn’t try to find some middle ground here to advance their plan. It really seemed like they took a hard line in the sand that TAR should be applied to this last set of documents. Instead of working with plaintiffs on a method to validate the search terms or a sampling of some sort that might have alleviated their concern, they just kind of put their foot down and said, this is the plan that we’ll go with.

In these types of situations, to be able to move the ball forward, you’ve got to try and find some middle ground, some sort of cooperative approach, because that’s what you agreed on in your ESI protocol. That failure to cooperate with that was also in direct contradiction to what the Court had ordered on cooperation and it frustrated the Court.

Now, I will say that there’s one area of this decision that I did not agree with the Special Master on and would love to hear any of your comments on it. The Special Master here raised the issue that applying TAR to this set of 560,000 documents would not allow the parties to uncover any search terms that wouldn’t have been captured by the original search term culling. So in essence, the Court is saying if you use linear review, you’re likely to identify additional search terms that weren’t included in the initial set that should have been included and might therefore expand the scope of what you want to review. If you use TAR, you won’t be able to do that.

Well, that really seems to fly in the face of the fact the parties agreed on search terms to cull the set. There’s nothing that the Court cites to as to whether or not there will be additional sampling done on search terms or whether that continued iterative analysis on search terms is to happen. And so without that language, it’s really odd to hear the Special Master focus on this particular point.

We have discussed multiple times on Case of the Week that search terms are fraught, and once you agree to them, absent a plan to continue to sample and test those search terms, that cold set is the set for review. Now, there may be more going on here than is in this actual decision, but I really think that that’s a point to be considered from a perspective of using this decision on an ongoing basis.

If the parties had put together any kind of process for sampling or revisiting search terms, then I could understand the Special Master’s position. But a producing party has to have some consistency for how it will move forward, even though production is always a moving target. And unless there is a plan for an iterative sampling of search terms, once you get to that review set, there has to be some boundaries around what a party has to review and produce. A constantly moving target is just way too difficult and way too expensive. Now had the defendants set forth evidence on cost savings and efficiency and tried to address the scope of the search term issue, this is a motion that could have succeeded.

And the takeaway here is that you really need to know and understand what you need to do to make an effective, successful argument on a motion trying to have the Court apply TAR in a situation like this. Otherwise, you are essentially just stalling the case even further and costing your client more on motion practice.


All right, that’s our Case of the Week for this week. Thank you so much for joining me. We’ll be back again next week with another decision from our eDiscovery Assistant database.

If you’re interested in doing a free trial of our case law and resource database, get started or reach out to us at support@ediscoveryassistant.com.

Thanks so much. Stay safe, healthy. Best of luck to those in Florida. Have a great rest of the week, and I’ll see you on the next episode.

Watch Next Episode Live

#CaseoftheWeek streams live on the ACEDS YouTubeFacebookTwitter, and LinkedIn company pages. Subscribe to the ACEDS social channels and turn on notifications to be notified when we go live.

Catch the Replay

You’ll be able to read episode summaries and watch the recorded show right here on the eDiscovery Assistant blog or access the recorded versions on our Vimeo channel.

Privacy Settings
We use cookies to enhance your experience while using our website. If you are using our Services via a browser you can restrict, block or remove cookies through your web browser settings. We also use content and scripts from third parties that may use tracking technologies. You can selectively provide your consent below to allow such third party embeds. For complete information about the cookies we use, data we collect and how we process them, please check our Privacy Policy
Consent to display content from - Youtube
Consent to display content from - Vimeo
Google Maps
Consent to display content from - Google
Consent to display content from - Spotify
Sound Cloud
Consent to display content from - Sound