#CaseoftheWeek Episode 66: Addressing Email Threading in ESI Protocols

Episode 66 of #CaseoftheWeek is an analyzes the importance of addressing email threading in your ESI protocols.

The case we discuss is In re Actos End Payor Antitrust Litig., 2022 WL 949798 (S.D.N.Y. 2022) from March 30, 2022 with the decision from United States Magistrate Judge Stewart D. Aaron.

Good morning, and welcome to episode 66 of our Case of the Week series, published in partnership with ACEDS. My name is Kelly Twigger. I am the founder and CEO of eDiscovery Assistant, as well as the Principal at ESI Attorneys, and I am thrilled to be with you today. Thanks so much for joining me.

As always, we welcome your comments and feedback in the comments section no matter what platform you’re viewing us on. You can see us live on LinkedIn, YouTube, Facebook, and Twitter.

If you want to sign up to receive that blog, you’ll get these weekly updates in the form of a newsletter on Thursday, the week that they come out. This week, there are a number of links associated with the decision that we’re going to cover in the In re Actos End Payor Antitrust Litigation matter.First, you’ll see a link to the actual decision itself that we’re discussing today from the Southern District of New York just from the end of March, just a few weeks ago here in 2022. You’ll also see a link to an article from eDiscovery Today in which Doug Austin did write up about this particular case.

You’ll see two links to the eDiscovery Assistant blog; one is a blog post that I wrote on email threading back in 2020. I was talking about the importance of it and highlighting some tips on using email threading from one of my dear colleagues and friends, Tessa Jacob at Husch Blackwell. Second, you’ll see a link to our series on ESI protocols and how we discuss email threading and the importance of including it in protocols, which is particularly apropos for today’s case law discussion.

This decision comes to us, as I mentioned, from the In re Actos End Payor Antitrust Litigation. This is a case that is pending in the Southern District of New York in front of United States Magistrate Judge Stewart Aaron. Judge Aaron has 23 cases in our eDiscovery Assistant database, so he is well versed in eDiscovery issues and is very thoughtful. This decision is also very thoughtful that we’re going to talk about some of the things that I see that are differences between sitting on the bench and actually engaging in eDiscovery and practice.

As always, we issue tag all of the decisions in our eDiscovery Assistant database, and this particular decision is tagged with failure to produce, manner of production, format, privilege log, attorney-client privilege, metadata, ESI protocol, and attorney work product.

All right, so what are the facts before us? We are before the Court on a letter motion on a motion to compel. On the motion to compel the plaintiffs asked for two things from the defendants. First, they ask that the defendants produce all non privileged responsive earlier in time emails that are part of the most inclusive email threads that the defendant has already produced or will produce to the plaintiffs. Second, the plaintiffs are asking for privileged log entries for those same earlier in time emails that are part of the email threads that were redacted or withheld for privilege.

In essence, really what we’re talking about with this decision is the use of email threading as a technology and whether or not you are required to produce all the lesser included emails when you use email threading technology. If you look at the blog posts that I cited to you earlier that are in the posts on the various platforms, you’ll see that we’ve talked about email threading a bunch on the blog.

Email threading is a fantastic tool because it really eliminates the need for you to continuously review the same messages over and over again. Essentially, email threading takes all of the emails in a particular thread and pulls them all together to be reviewed by one person at one time. That eliminates you having multiple reviewers seeing messages from the same thread, perhaps coding them differently. It also allows you to apply a redaction to an email at one time that can then be populated across that same email.

The issue becomes when you produce emails that are part of an email thread, and generally the receiving party wants to receive each individual email as a separate document, because they want to receive the associated metadata with that document to allow them to search and filter.

Imagine, for example, if you have an email thread that has ten people on it and you’re preparing for a deposition of John Doe, and John Doe is on two of those emails, but he’s not on the other eight. Essentially, when you’re doing that preparation, you want to have that metadata to have those emails be able to be segregated as just the two that went to John Doe be a part of his deposition production. Whereas if you don’t have the metadata from those two lesser included emails and you only have the highest thread, if that thread doesn’t list John Doe on it, or there’s other metadata that would prevent it from being contained in a search for John Doe’s materials, then you really lose out on the ability to use that information.

The Federal Rules of Civil Procedure provide that essentially both parties should have the same information. That’s been a topic of discussion among the Sedona conference. I’ve had that discussion multiple times here on the case of the week and also on panels that I’ve spoken on. Essentially, both parties should have access to the same amount of information.

With all of that in mind, getting back to the decision that we’re on, the underlying case here is really a very complex commercial antitrust action in which the plaintiffs allege that the Defendants prevented competitors from timely marketing a generic version of the defendant’s diabetes drug Actos by falsely describing two patents to the Food and Drug Administration.

This is a case that was filed back in December of 2013. It’s an old case. We’re in 2022 at this point. On March 18th of 2015—two years and three months later—the Court approved an order governing the protocol for the discovery of ESI and hard copy documents, which the Court refers to as the Discovery Protocol.

In that protocol, which the Court approved, the parties negotiated for the production of ESI in native formats, together with metadata and coding fields that are set forth as an exhibit to the protocol. That’s pretty much how we generally handle protocols as well. The protocol that the parties agreed upon requires the parties to deduplicate ESI that is produced so as to avoid the production of exact duplicate documents, but the protocol did not provide for the production of only the most inclusive email threads. Meaning that the parties did not include any specific language about using email threading or about what would be produced following a review using email threading.

Beginning in February of 2022, seven years after that protocol was entered, the Defendant began making multiple rolling productions of non privileged documents from 25 agreed upon custodians, including six in house lawyers using email threading. The email threading, as the Defendants describe it in their motion, is by which a party reviews and produces the most inclusive email in a thread.

Essentially what Defendants did was they conducted their review using an email threading tool, and then they produced only the most inclusive email thread. That would mean that all of the metadata from the lesser included emails would not be included as part of a production, because those would not be separate documents that would be produced. That’s really the issue here.

The plaintiffs then objected to the Defendant’s use of email threading and are really asking the Court to require the Defendants to produce the earlier in time emails, as well as the metadata that’s associated with those emails. Secondly, the Plaintiffs also seek to compel the Defendants to provide privilege log entries for all of those emails, including the earlier in time emails. The Defendant argues that compelling that approach on the privilege log would impose an enormous burden on them to itemize each one of those individual emails. The Court also knows that the parties exchange drafts of a privileged log protocol, but have never been able to agree to the terms of such a long call.

All right, the two issues before the Court are 1) whether the defendant’s production of lesser included emails is appropriate or whether they should be required to produce those lesser included emails, and 2) how email threads should be included on a privilege log.

Now, the second issue, the emails included on a privilege log is one that has been discussed a lot. It’s something in particular that Judge Jones out of the Florida Federal District Court has raised and that we’ve discussed multiple times on the UF eDiscovery Conference. I think it’s really challenging; I think the Courts want to try and find solutions on how to deal with privilege logs, and one of them that we’re going to discuss is the use of categorical privilege logs. I still have a lot of problems with that from a practical perspective, but I do agree that there is a definite need to try and come up with a better solution on how to deal with privilege logs.

All right, let’s start with the email threading issue here. The Court first looks at Rule 34 of the Federal Rules of Civil Procedure, which requires a party to produce designated ESI. The Court notes that the Federal Rules do not contain any provisions regarding the use of email threading, nor do the local civil rules of the Court. However, the Court does note that Rule 29 of the Federal Rules Civil Procedure permits the parties to stipulate to procedures governing discovery by, for example, the use of ESI protocols. The Court notes that the Sedona principles encourage the parties to have early discussions regarding potential issues related to the form of production and to enter into, “an agreed upon protocol governing the production of ESI and avoid downstream misunderstandings or disputes.”

The Court notes, the defendant argues that the email threading is often used in ESI protocols, and the defendants point to a protocol in which they were a party that included a specific provision allowing for email threading. I’m going to quote that language to you that was in that previous protocol because it’s useful, and I encourage you to take a look at this decision. It’s very short; it’ll take you less than ten minutes to get through it. It’s really important from a perspective of understanding how ports are viewing email threading.

Here’s the language that was in the previous protocol during which the defendants were a party.

Email threading: The producing party may identify email threads where all previous emails which make up the thread are present in the body of the final email in the thread. Any party electing to use this procedure must notify all receiving parties that email thread suppression has been proposed to be performed on a specified production, and the parties agreed to meet and confer regarding format of this production and reserved the right to seek court guidance on the issue should agreement not be reached.

That’s the sum total of the language that’s in the previous ESI protocol.

Now, interestingly, that language really says that the parties will meet and confer if one party says they’re going to use email threading, and that’s not what happened here. What happened here is that defendants used email threading, produced documents with only the top email with all of the individual lesser included emails in it, and didn’t confer with the other side. Interestingly, they’re proposing to the Court that it should follow language from a previous protocol that is not the same situation that’s being presented to the Court here.

The Court does note that there is no similar agreement on ESI threading in this particular ESI protocol that was entered by the Court back in 2015. The Court also notes, and this is really important because it demonstrates the fact that the Court is understanding the technical issues here that the defendant’s exclusion of lesser included emails from production has resulted in the exclusion of metadata associated with the earlier emails in the chain. The Court notes that that could even be weeks or even months prior to the last email in the chain, presumably from looking at some examples, but that’s not noted in the decision.

The Court notes that the exclusion of metadata has multiple effects on the plaintiff’s ability to search and filter produce data, including three things that it notes, specifically. One, that it materially reduced the plane of the ability to search for all correspondence within a date range, because, again, we don’t have the individual dates of those lesser included emails. Second, the Court notes that in certain email chains, only the sender of particular emails earlier in a chain are reflected, not the recipients of such emails. Third, the lack of metadata removes plain stability to see if anyone was blind copied unless they’re included emails, even though this information was among the metadata the parties agreed to in the discovery protocol to produce.

The Court’s noting a few things here. One, it’s understanding the impact of not having the metadata on those lesser included emails to the receiving party, which is here at the plaintiff. Second, it’s also noting that the production by the defendants here not including the metadata for the lesser included emails is contrary to the provision in the discovery protocol, which says that certain metadata fields will be produced for each document. Those are two important points that the Court’s making.

How does the Court come down on the email threading? The Court notes that if the issue of email threading had been raised at the time the parties were negotiating the discovery protocol, they may have been able to resolve the issue by including, for example, a protocol provision that’s requiring the exchange of certain metadata to be included for those lesser included emails. That’s hard to do from a production perspective if you’re not producing the actual documents. Again, that’s where I mentioned we see courts trying to come up with alternative solutions to these problems, but not necessarily understanding exactly how they would play out in terms of production in resolving the situation.

The Court then goes on to note that because the plaintiffs were not provided the opportunity to negotiate how email threading might be accomplished in an acceptable manner, the Court declined to impose email threading on the plaintiffs. The Court goes on to say that although it recognizes that the production of earlier in time emails will cause some burden on the defendants. The Court finds that any additional burden is not undue since they agreed to the discovery protocol and likely has already reviewed many of the emails at issue. The Court required that the defendants produce all of the responsive ESI to plaintiffs, including the earlier in time emails.

That’s where we come out on the email threading issue. Essentially the court is saying if you don’t have this in your discovery protocol, you cannot leave out lesser included emails to the opposing party.

Now, turning to the privilege log issue, this is really impacted by the Court’s decision on the email threading issue, and the Court really directs the parties to meet and confer on coming up with a revised privilege log protocol based on the Court’s ruling on email threading. The Court does go on to provide some guidance to the parties regarding the terms of such protocol.

The Court cites the two particular rules, Rule 26(b)(5), that lays out the requirements of a privilege log, as well as the local rule on the Southern District of New York. The Court cites to a decision in Rekor Systems v. Loughlin that allows the use of categorical privilege logs in the Southern District of New York. That’s very important and the Court goes on to cite a specific language regarding that, and that is this, “A categorical privilege log is adequate if it provides information about the nature of the withheld documents sufficient to enable the receiving party to make an intelligent determination about the validity of the assertion of the privilege.”

In that case, the categorical privilege log was found acceptable by Judge Liman and Rekor, it contained groupings of multiple documents over discrete date ranges to and from individuals at identified firms listing the general subject matter, and that was acceptable to the Court in that particular case. The Court is really just providing that as guidance to the parties here on how to deal with the privilege law.

The Court rejected the plaintiff’s proposal of permitting categorical logging of emails only where all emails involve the same participants as subject matter as being inconsistent with the Rekor case. Since there’s no requirement that all the participants be identical for categorical logging to be appropriate. The Court also rejected the defendant’s proposal of only logging the threaded emails as also being inconsistent with the Rekor case and the rules, since it’s unlikely the log would contain sufficient information to plaintiffs to assess the claim of privilege for each email in the thread.

Court then goes on to say that now that it’s requiring the production of all responsive ESI, the parties are directed to meet and confer with respect to the privileged log protocol consistent with its decision on the email threading. In essence, the plaintiff’s letter motion is granted in part and denied in part, and the defendants required to produce all the responsive ESI to the plaintiffs, including the earlier in time emails, and the parties are required to meet and confer to come to terms for a privilege log protocol.

All right, so what are our takeaways from this case? Well, I think they’re going to be pretty obvious, and that is that one you’ve got to discuss how you’re going to handle email threading and privilege logs with regard to email threading at the time you’re negotiating your ESI protocol. If you take a look at the blog posts that are linked to this particular video, you’re going to see that we’ve discussed exactly those kinds of discussions, how you need to have them. There’s sample language here in this case that you can use in your ESI protocol to deal with that issue. Oftentimes, as we’ve discussed previously on the Case of the Week, we allow review using email threading if we’re the receiving party, but we require that production be made of all lesser included emails along with the corresponding metadata so that we get the same information but you’re allowed to leverage email threading in order to be able to expedite your review and make sure that it’s more consistent with regards to lesser included emails on the string.

Next takeaway, I want you to really think carefully about the kinds of information that you’ll be producing and receiving and discuss all of these issues upfront in order to be able to preclude the motion practice that we’re seeing on this decision. As always, I still recommend that you go to the Court if you have unresolved issues, but those should be on the protocol so that the language of the protocol can then be followed throughout the case. We tend to refer back to the ESI protocol hundreds of times during a case in terms of every single production that’s made, in terms of the party’s obligations as to what is to be provided. It’s really important to think carefully about what goes into that document.

I mentioned earlier that production of these lesser included emails is going to be kind of a mess for both parties, and that’s really because you’re going to see things that are not going to be produced as they existed in the ordinary course of information, and that’s what’s required by Rule 34. Rule 34 requires that you either produce information as it exists in the ordinary course of business or by request. Here I’m not sure how the defendants have gone about, they probably already have since six weeks later, go about producing those emails as lesser included emails so that they correspond with the existing productions from the defendants. It’s going to be tough.

With regard to the categorical privilege logs. We talked about this a little bit within the context of the Court’s ruling, but I find categorical privilege logs to be very difficult from a perspective of interpreting them upon receipt, and timing of when you get them is also a factor. Some courts are requiring that parties produce privilege logs with each production. Other courts are allowing parties to produce a privilege log at the end of discovery. It’s hard when you’re looking at multiple privilege logs to be able to determine what’s happening, particularly if documents are being produced by custodian and not necessarily sequentially or in a way that makes it easy to understand what category of information is being included on this privilege log, and by that I mean what’s the scope of the actual production of the privileged log is being produced for.

My view is that categorical privilege logs really prejudice the receiving party unless they are finally negotiated. There’s just not a lot of time to review privilege logs and engage in disputes over privilege logs in discovery, especially if you’re getting them at the end of discovery and you’re really rushing to get depositions completed and compile information for summary judgment motion. It’s very difficult to resolve privilege log issues and to feel like you’ve gotten the information that you need. It’s really important that you negotiate what that’s going to look like up front in your ESI protocol.

One of the things that I noticed here is that the parties all make the appropriate arguments here, but they really needed to go to the Court with actual metrics on email, threading and logging in advance of the production or while negotiating the protocol in order to really give the Court what they needed in terms of seeing the burden that would have been on defendants or the prejudice to the plaintiffs. That protocol wasn’t finished until two years after the litigation filed, which may likely be due to motion practice or maybe due to the amount of time that negotiations were strung out. It’s not clear from this decision.

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 are an ACEDS member and interested in using eDiscovery Assistant, there is a discount available to current ACEDS members and a trial for folks who are taking the ACEDS exam. There’s information on how to get to those through your ACEDS portal.

If you’re interested in taking eDiscovery Assistant for a test run and you’re not an ACEDS member, you can sign up for a free trial, or you can also reach out to us at support@eDiscoveryAssistant.com and somebody will be in touch with you to set up a review for you or for your organization. Thanks so much, stay safe and healthy out there and I’ll see you next week. Bye.

Watch Next Episode Live

#CaseoftheWeek streams live on the ACEDS YouTube, Facebook, Twitter, and LinkedIn company pages. Subscribe to the ACEDS social channels 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.