Documents Identified by Agreed Upon Search Terms Do Not Necessarily Require Production of Those Documents
March 30, 2016
In Gardner v. Continental Cas. Co., (2016 WL 155002 [D. Conn. Jan. 13, 2016]), the District Court was called upon to decide two different issues raised by Plaintiffs in a motion to compel. The case itself concerned the long term care insurance coverage for five Connecticut residents for stays at Connecticut Managed Residential Care (“MRC”) facilities. As is relevant here, after some negotiation, counsel agreed to a list of search terms to use to search the emails of twenty-three custodians. The result was the return of approximately 38,000 documents. Defendants reviewed the documents for relevance and privilege and produced 2,214 pages of documents – many of which were copies of the complaint and other filings in the lawsuit. Plaintiffs sought to compel the production of the balance of the 38,000 documents, all of which were found using the agreed-upon search terms, and argued the smaller production was the result of defendant “cherry-pick[ing]” documents. Plaintiff also agreed it should not be forced to accept the “just trust us” approach defendant endorsed. Further, the plaintiffs argued that the purpose of the agreed-upon search terms was “to avoid prolonged and detailed debate over what ESI documents [were] ‘responsive’ . . .” The plaintiffs supported their position regarding the scant production by pointing out that the defendant’s third-party claims adjustor submitted a “far more comprehensive and informative” production, while the defendant argued that it had already provided “extensive discovery” and that it had spent “significant resources” reviewing the documents from the agreed-upon search terms.
This discovery issue arises with much consistency in cases with extensive electronically stored information (“ESI”). Overwhelmingly courts conclude that the position taken by plaintiffs is “simply untenable” – the defendant is not obligated to turn over all 38,000 documents, especially where issues of privilege abound. The court did, however, recognize plaintiffs’ “legitimate concern” regarding the limited production, and ordered opposing counsel to confer and discuss approaches for addressing the potential need to turn over results of the search hits including—“sampling and iterative refinement.”
This decision raises an interesting issue. It would seem obvious that only relevant non-privileged documents would be produced irrespective of how many documents (i.e., false responsive or privileged) “hit” upon a search term. However, to avoid motions to compel and protracted discovery disputes, this case reminds us all to spell out precisely what we are agreeing to do when we enter into ESI protocols.