by Julia Romero Peter, Esq.
In re Biomet M2a Magnum Hip Implant Prods. Liab. Litig., No. 3:12-MD-2391 (N.D. Ind. Apr. 18, 2013), the court held that defendant Biomet’s use of keyword culling and de-duplication in conjunction with predictive coding (technology assisted review) satisfied its discovery obligations.
In 2012, defendant Biomet began producing documents in a multidistrict products liability action that ultimately became consolidated in the United States District Court of the Northern District of Indiana. Biomet started identifying and producing documents despite certain plaintiffs’ counsel warning Biomet not to start production until the Judicial Panel on Multidistrict Litigation decided where to centralize the cases.
To identify relevant documents, first Biomet used keyword searching and de-duplication. The keyword culling decreased the document population from 19.5 million to 3.9 million documents and attachments. De-duplication reduced this number further to 2.5 million documents and attachments. Biomet performed statistical sampling of a random sample. It projected at a 99% confidence level that between .55 and 1.33 percent of the unchosen documents would be relevant and 1.37 and 2.47 percent of the initial 19.5 million documents were relevant. Biomet’s methods had found a notably higher 16 percent.
Subsequently, from the 2.5 million that resulted from the keyword and de-duplication methods; Biomet used technology assisted review (TAR) to determine the relevant documents to be produced. At the time of the court’s order, Biomet’s ediscovery costs were approximately $1.07 million. The costs are expected to amount to $2 to $3.25 million.
Plaintiffs’ Steering Committee argued that Biomet’s use of keyword culling “tainted the process,” asserting that it was less accurate than using TAR. They sought to have Biomet redo the processing through production process using only TAR, starting from the original 19.5 million documents. In addition, they sought to have plaintiffs and defendant “jointly enter[] the ‘find more like this’ commands” during the TAR process. Biomet objected that “starting over would cost it millions.”
The court denied the Steering Committee’s requests. In finding that Biomet’s methods met its discovery obligations, the court noted, “The issue before me today isn’t whether predictive coding is a better way of doing things than keyword searching prior to predictive coding . . . .” It further noted that even if TAR could identify more relevant documents than keyword searching, as asserted by plaintiffs, it would cost defendant potentially millions of dollars. The court applied Federal Rule of Civil Procedure 26(b)(2)(C) to “find that the likely benefits of discovery . . . equals or outweighs its additional burden on, and additional expense to, Biomet.” The court arrived at this conclusion even after considering a number of ostensibly compelling factors, including “the needs of the hundreds of plaintiffs in this case, the very large amount in controversy, the parties’ resources, the importance of the issues at stake, and the importance of this discovery in resolving the issues . . . .”
The court assumed that Biomet would be amenable to “meeting and conferring on more reasonably-targeted search terms and to producing the non-privileged documents included in the statistical sample.” It also stated that if the Steering Committee wanted documents only retrievable by redoing the processing through production steps the Steering Committee would have to pay the costs for doing so.
Download our FREE white paper: