What Does “The Making of a Surgeon” Have to Do With ESI and “Software Glitches?”

Proportionality in Government e-Discovery – June 2011 IQPC Presentation
June 30, 2011
American Bar Assoc., Publishes Berman, et al., Managing E-Discovery and ESI
July 30, 2011

“The Making of a Surgeon”

In 1968, Dr. William A. Nolen wrote “The Making of a Surgeon” (Mid-List Press 1968, 1990):

How do you make a surgeon? Not by the preliminaries, the four years of college and four years of medical school that have to be gone through to earn an M.D. degree, but by the five, six or seven years that a man [or woman] spends after medical school learning the surgical trade.

Id. He suggested that “[t]he transformation is a slow process. . . . Not by big jumps, just small steps forward.”  Dr. Nolen wrote:

It’s sad that all through a book on surgery, mistakes and error should play such a prominent role.  Patients like to believe doctors are infallible. . . .  Unfortunately, that’s not the case. Not some, not most, but all doctors, at one time or another, make errors.  This is the nature of medicine. It’s isn’t an exact science. .  . . [W]e made errors. We operated when we shouldn’t have, didn’t operate when we should have; sometimes we performed the wrong operation; occasionally, we chose the right operation but did it poorly.  We made all these errors. But, to keep the proper perspective, not often.  We did our damnedest to avoid mistakes – we took every precaution we could to eliminate errors and struggled to keep them down to the absolute minimum.  And for every patient we hurt by one of our mistakes, we did, I hope, help hundreds.

Id. at 57, 65.  The author described “[t]he patient, or better, victim,” of his first major surgery, an appendectomy.  Id. at 18.  When handed the scalpel, all of his confidence disappeared and he could not determine where to make the incision.  Then, his effort to tie off the base of the appendix failed, contaminating the surgical area. Although the recovery was more complicated, the patient survived.  Dr. Nolen concludes: “If he had died, I think I would have given up surgery for good.”

What does that have to do with ESI and software glitches?  “Especially in the highly complex world of e-discovery, even with good faith efforts, it is very easy to fail to preserve or lose relevant information by inadvertence.”  P. Grimm, M. Berman, et al., “Discovery About Discovery: Does the Attorney-Client Privilege Protect All Attorney-Client Communications Relating to the Preservation of Potentially Relevant Information?,” 37 U.Balt.L.Rev. 413,  454 (2008).

Perfection Is Not, and Has Never Been, the Standard of Litigation

In The Pension Committee of the University of Montreal Pension Plan v. Banc of America
Securities, LLC,
685 F.Supp.2d 456, 461 (S.D.N.Y. 2010), the court wrote:

In an era where vast amounts of electronic information is available for review, discovery in certain cases has become increasingly complex and expensive. Courts cannot and do not expect that any party can meet a standard of perfection.

In a forthcoming book, The Hon. Paul W. Grimm wrote that the Advisory Committee Note to Fed.R.Evid. 502(b) suggests that the Rule “does not require perfection. . . merely that what was done in a particular case was reasonable.”   P. Grimm, “The 10 Most Important Things You Need to Remember About FRE 502,” in M. Berman, C. Barton, and P. Grimm, eds., “Managing E-Discovery and ESI: From Pre-Litigation Through Trial” (ABA forthcoming Aug. 2011), 416-17.

Thus, “[i]t would be utopia to expect perfection from everyone: police, lawyers, judges, and society in general. We can take note, however, that while we often strive for perfection it is seldom achieved.”  State v. Scott, 699 S.W.2d 760, 764 (Mo. App. 1985)(failure of police to dictate a complete inventory on property bag)(subsequent history omitted); accord Southern Capitol Enters., Inc. v. Conseco Servs., L.L.C., 2008 U.S. Dist. LEXIS 87618, at *7 (M.D. La. Oct. 24, 2008)(in an “ongoing electronic discovery dispute,” the court wrote that “

erfection in document production is not required. . . .”); Frazier v. Layne Christensen Co.,  2005 WL 372253, at *4 n. 3 (W.D. Wis. 2005) (in a proceeding for violation of protective order, the court wrote that “[t]his court doesn’t expect perfection from counsel or their staff when thousands of documents are being exchanged. . . .”);  Gilmore v. Oil and Gas Conservation Comm’n., 642 P.2d 773, 781 (1982)(“Appellant seems to expect perfection. Justice was accomplished here, as much as could be under the circumstances. This litigation should end.”)(administrative appeal); State v. Turner, 21 N.C.App. 608, 205 S.E.2d 628 (1974)(“Certainly, the bench and bar cannot expect perfection in the transcription of trial court proceedings, particularly jury instructions.”); Williams v. Russell, 419 F.2d 1092, 1094 (6th Cir. 1969)(in denying habeas petition for alleged failure to request a severance and call a specific witness, the court wrote “[w]e just simply cannot hold lawyers to perfection to guarantee that they will be absolutely free from error. . . .”), cert. denied, 398 U.S. 942 (1970); Dolley v. Ragon, 68 Cal.App. 223, 226, 228 P. 665, 666 (1924)(contempt proceeding against attorney for misrepresenting facts in the record)(“Every one knows that in the course of litigations in which numerous details of extensive transactions have to be referred to, as they were referred to in this case, sometimes counsel, and sometimes the judges of the court, will commit errors, whereby a misstatement of fact may be made, or a fact may be omitted which ought to have been included in the statement. That occurs every day. No one reasonably expects perfection in such matters, and no one reasonably charges corrupt intention in the making of such errors.”); U.S. v. Foster Lumber Co., Inc., 429 U.S. 32, 52 (1976)(“No one expects perfection in income taxation.”)(Blackmun, J., dissenting).

In short, “[t]he goal of our system of justice is fairness, not perfection. We live in an imperfect world and it is unrealistic to expect perfection in the courtroom.”  Gould v. Charlton Co., Inc., 929 S.W.2d 734, 739 (Ky. 1996)(discussing jurors).  Thus, “[s]ometimes, despite everyone’s best efforts, mistakes are made during the course of a trial. We visit an endless array of trial errors with a paradigm that looks for fairness rather than perfection. We do not expect perfection.”  People v. Biggerstaff, 287 Ill.App.3d 813, 818, 679 N.E.2d 118, 121 (1997); accord Bush v. Gore, 531 U.S. 98, 143 (2000)(“we live in an imperfect world. . . .”)(Ginsburg, J., dissenting)(Presidential election).

Sanctions Are Reported To Be At An “All Time High”

It is in this context that a recent article states: “E-discovery sanctions are at an all-time high.” D. Willoughby, R. Jones, and G. Antine, “Sanctions for E-Discovery Violations: By The Numbers,” 60 Duke L.J. 789, 790 (2010).  “[T]here has been a significant increase in
both motions and awards since 2004.” Id. at 790-91.  While “[m]arquee e-discovery
disaster cases. . .  are towering reminders of the most severe sanctions. . . [o]f greater concern to the average practitioner is the increasing frequency of sanction decisions, an issue most recently illustrated by Pension Committee of the University of Montreal Pension Plan v. Banc of America Securities, LLC, in which all thirteen plaintiffs were sanctioned for e-discovery failings not rising to the level of intentional or willful conduct.” Id. at 791.  Mr. Willoughby and his colleagues conclude that “appropriate consideration should be given to the complexity of e-discovery in ruling upon the increasingly frequent e-discovery sanction motion.” Id. at 828.

Sanctions and the Brady Analog

Sanctions motions in civil cases have “developed a civil procedure analog to a Brady [v. Maryland, 373 U.S. 83 (1963)] attack on alleged prosecutorial misconduct.   In
short, under the sanctions rules, civil litigators could obtain a tactical advantage by alleging deficiencies in the performance of opposing counsel.  The opposing attorney, countering such allegations, was often tempted to respond in kind.  Civility suffered.”  M. Berman, “The Sedona Conference Cooperation Proclamation” ABA Technology for the Litigator, reprinted in M. Berman, C. Barton, P. Grimm, eds., “Managing E-Discovery and ESI: From Pre-Litigation Through Trial” (ABA forthcoming Aug. 2011); see P. Grimm, et al., “Discovery About Discovery: Does the Attorney-Client Privilege Protect All Attorney-Client Communications Relating to the Preservation of Potentially Relevant Information?,” 37 U.Balt.L.Rev. 413 (2008).

This is especially problematical because “[a]s a relatively recent concept, e-discovery’s reasonableness standards are still developing.” D. Bassett, “Reasonableness in E-Discovery,” 32 Campbell L.Rev. 435, 436 (2010).  In short, courts and litigators are still “writing the book.”  M. Berman, C. Barton, P. Grimm, eds., “Managing E-Discovery and ESI: From Pre-Litigation Through Trial” (ABA forthcoming Aug. 2011), 2.

The anticlimactic denouement in Qualcomm, Inc. v. Broadcom Corp., 2008 WL 66932 (S.D. Ca. Jan. 7, 2008), vacated and remanded in part, 2008 WL 63810 (S.D. Ca. Mar. 5, 2008), appeals dismissed, 327 Fed.Apx. 877, 2008 WL 6400775 (Fed. Cir. Aug. 18, 2008), on remand, 2010 WL 1336937 (S.D. Ca. Apr. 2, 2010), partially exonerating trial counsel after their careers were devastated by an earlier sanctions decision, illustrates the need for caution in analyzing ESI-related errors.

Software Glitches

Software glitches are a fact of contemporary life. C. Cavas, “F-35s resume flight ops after software glitch,” Navy Times (Jun. 24, 2011); ; J. Scully, “Software glitch delays launch,” Lompac Record (Jun. 9, 2011)(NASA rocket); C. Sorrel, “RIM Recalls PlayBooks Thanks to Software Glitch,” Gadget Lab (May 17, 2011); O. Wright, “Software glitch costs tax office millions in lost revenue,” (The Independent Feb. 1, 2011); Associated Press, “Nissan to fix software glitch in electric car,” (A.P. Apr. 18, 2011);  P. de Selding, “Software Glitch Blamed for Turning Satellite Into Space Zombie,” (Space.com Jan. 13, 2011).

Many cases illustrate software glitches. I.B.E.W. v. Limited Brands, Inc., 2011 WL 1238308, at *15 (S.D. Oh. 2011)(“CW4 and CW6 represented that the distribution center was experiencing software glitches throughout the class period. . . .”); Shah v. Washington County Assessor, 2011 WL 2651582, at *4 (Or. Tax Magistrate Div. 2011)(“Huffman admitted at trial that due to a ‘software glitch’ his analysis of Plaintiffs’ comparable properties gave them a downward adjustment for extra bathrooms when the analysis should have given an upward adjustment to those same properties.”);  Central Parking Systems of Tennessee, Inc. v. Nashville Downtown Platinum, LLC, 2011 WL 1344633, at *1 (Tenn.Ct.App. 2011)(“Due to a “software glitch[,]” however, the amount was double posted, and a check was sent to NDP for twice that amount: $53,761.78.”); In re Toyota Motor Corp. Unintended Acceleration Marketing, Sales Practices and Products Liability Litigation, 754 F.Supp.2d 1208, 1221 (C.D. Cal. 2010)(“Plaintiff identifies three design defects that cause or contribute to SUA events, including: . . . software glitches. . . .”);  Beverly Enterprises v. Mississippi Div. of Medicaid,  808 So.2d 939, 943 (Miss. 2002)(“It is unrefuted that a computer software glitch in a newly installed system caused Beverly to miscalculate the correct amount of funds necessary to properly reimburse them for services actually rendered to approximately 712 patients.”).

Sensient Colors, Datel Holdings, and Mt. Hawley

Three recent cases addressed software glitches in the ESI context.  In two, courts recognized the complexity, and failures were excused. In the other, a different result was reached.

Sensient Colors

Faced with determining the adequacy of counsel’s efforts to protect disclosed, privileged information, one court wrote: “Plaintiff’s production was made in the midst of a commendable effort to employ a sophisticated computer program to conduct its privilege review. Unfortunately, mistakes occurred. Plaintiff should not be unduly punished for occasional mistakes that occurred while it started to use new software to organize and sort its documents.”  U.S. v. Sensient Colors, Inc., 2009 WL 2905474, at *4 (D.N.J. Sep. 9, 2009).  According to the court, plaintiff “prepared a database of relevant fields of each of its documents,” conducted oral and written training for the reviewers, and employed pre-production, computer-assisted quality control.  Nevertheless, 214 documents were inadvertently produced out of a 45,000 document production.

The Sensient Colors court cited Heriot v. Byrne, 257 F.R.D. 645, 659 (N.D. Ill. 2009)(citations omitted), for the proposition that “where discovery is extensive, mistakes are inevitable. . . .” The court noted that “no disclosure would have occurred but for the [ESI] Vendor’s error. . . .” Id. at 661.   It concluded that “Plaintiffs relied, and should be able to rely, on their Vendor to faithfully carry out the instructions it had been given.”  Id. at 660.   The court deemed it unfair to penalize a litigant for a mistake it did not cause or anticipate. Id.

Thus, the Sensient Colors court reasoned: “It appears that plaintiff’s implementation of a new computer application was largely to blame for the errors that occurred. This is not unexpected. The use of sophisticated analytical software should be encouraged. Obviously, however, given plaintiff’s experience thus far, future errors will not be treated generously.”  Id. at n. 11.

Datel Holdings

In Datel Holdings Ltd. v. Microsoft Corporation, 2011 WL 866993 (N.D. Cal. Mar. 11, 2011), subsequent opinion on other issues, 2011 WL 2437265 (N.D. Cal. Jun 17, 2011), the defendant inadvertently produced privileged documents due to a software glitch, much like Sensient Colors.  In a nutshell, a litigation support program truncated the portion of an email chain that showed the entire chain was subject to an assertion of privilege. Therefore, reviewers missed it, failed to assert privilege, and six assertedly-privileged documents were erroneously included in the 119,000-document production.

As in Sensient Colors, defendant had a review protocol described by the court as “fairly robust.”  It included a first-pass review of potentially responsive documents, followed by a quality control team review of potentially privileged documents, followed by a privilege team review of the output.  Reviewers received written instructions and a tutorial from litigation counsel. Litigation counsel also conducted quality control checks.  Faced with a waiver argument, the Datel Holdings court wrote:

Specifically, Defendant explains that after potentially responsive documents were collected from custodians, they were loaded into a computerized document processing system known as “Clearwell.” Clearwell extracted metadata from each document and converted the documents into a format that allowed for text searching. Once the documents were processed through Clearwell, they were entered into an online platform, where they were reviewed by attorneys. For reasons still unknown to Defendant, Clearwell truncated some “Re-auth” documents during processing. [internal citations omitted].

The court wrote:

Here, although Defendant’s team of lawyers carefully reviewed documents to identify privileged communications, a computer glitch truncated the documents, removing the portion conveying the request from counsel to conduct a factual investigation. The technical glitch was a mistake, which occurred accidentally and unintentionally, and prevented Defendant’s team of lawyers from recognizing the privileged nature of the email chain. Mistaken production due to an unexpected software glitch that occurred despite the use of standard discovery software falls squarely on the inadvertent side of the divide between intentional disclosure under Rule 502(a) and unintentional disclosure under Rule 502(b). Under these circumstances, production of these six documents was inadvertent. [footnote omitted]. . . .  Here, Defendant used a computerized document processing system to organize its documents which, unbeknownst to Defendant, suffered a software failure.

The court concluded:

In relatively large productions of electronic information under a relatively short time table, perfection or anything close based on the clairvoyance of hindsight cannot be the standard; otherwise, the time and expense required to avoid mistakes to safeguard against waiver would be exorbitant, and complex cases could take years to ready for trial.

As in Sensient Colors, these actions were held to be reasonable and, therefore, sufficient.

“[F]airly robust” procedures were good enough in light of what the Datel Holdings court called a relatively large production “under a relatively short time table. . . .”  2011 WL 866993, at *4.

Mt. Hawley

In Mt. Hawley Insurance Co. v. Felman Production, Inc., 271 F.R.D. 125 (S.D.W.Va.  2010), objections overruled, 2010 WL 2944777 (S.D.W.Va. Jul. 23, 2010), the court was also faced with a software glitch “in the context of a massive production of e-discovery.” In brief summary, the plaintiff had produced an email between its Human Resources Manager and attorneys and, upon learning of the error, subsequently argued non-waiver of the privilege.  Additionally, 377 other documents were at issue.

Opinion of the Magistrate Judge

The May 14 email had been listed on plaintiff’s privilege log.  The court noted that plaintiff “and its counsel used various search protocols, software and vendors’ services to search for and produce documents.”  It recited that production took place in an accelerated, five-month period.  It wrote that defendants identified almost 980 attorney-client communications, of which plaintiff sought to claw back 377, although the initial decision turned primarily on the single email.

The plaintiff, defending its production as inadvertent, “points to its ‘carefully selected privilege search terms,’ document-by-document review of potentially privileged documents, and a second electronic search of remaining documents, as evidence of its reasonable steps. It notes that the May 14 email was listed on its privilege log. After learning of the production of apparently privileged materials, Felman investigated and determined that certain documents were not tagged for attorney review due to an undetermined software error.” [internal citations omitted].

The court framed the dispositive issue as whether plaintiff took reasonable steps to  prevent disclosure. Id. at 133.  It described a sophisticated series of steps taken by counsel to obtain and process ESI, including retaining an “ESI collection vendor,” collecting 1,638 GB, selecting search terms, testing the search terms, refining the search terms after testing, tagging documents for privilege review, and performing a document-by-document review of potentially privileged materials. The ESI had been loaded in multiple Concordance databases and, after learning of the inadvertent disclosure of the privileged email, investigation revealed the one of the databases “inexplicably built an incomplete index of potentially privileged materials.”  Id. at 136.  The court wrote: “The manufacturer of the Concordance software, Lexis-Nexis, has not been able to explain why the index was incomplete.” Id. It was then determined that 328 of the 377 documents had come from the mis-indexed database.

The court found that the production of the 377 documents was not “solely attributable to the problem with the fourth Concordance database file.” It faulted plaintiff for failing to perform critical quality control sampling and found a failure to perform simple key word searches to locate copies of the key email. It concluded that “the precautions taken to prevent inadvertent disclosure were not reasonable.”

Opinion of the District Judge

Felman filed objections to the Magistrate Judge’s decision.  In overruling the objections, the District Judge wrote that discovery in the case had “taken on a life of its own” and “been especially adversarial. . . .” 2010 WL 294477, at * 1.  The court noted that 30% of Felman’s production had contained irrelevant materials and it had “produced nearly a thousand communications subject to attorney-client privilege,” a fraction of which it sought to claw back. It wrote:

Although inadvertently, Felman produced a massive amount of irrelevant documents and thousands of privileged communications. In light of this fact, the Court need not consider the details of Felman’s preproduction document review process – or the unexplained technical glitch that led to the fourth volume of Felman’s e-discovery being not properly screened for potentially privileged materials – to find that Felman’s screening precautions were not reasonable.  The ridiculously high number of irrelevant materials and the large volume of privileged communications produced demonstrate a lack of reasonableness.

The Mt. Hawley court may have been concerned by the allegations of a “data dump” and the assertion that at least 14.3 GB of “junk documents” had been produced.  On the other hand, however, the pre-production steps, detailed at 271 F.R.D. at 135, are impressive precautions.

What Can We Learn From Dr. Nolen?

After his surgical mistake during the appendectomy, Dr. Nolen concluded: “If [the patient] had died, I think I would have given up surgery for good.” Careers were ruined in Qualcomm.  Z. Elinson, “Judge Lifts Sanctions Over Qualcommm Discovery Scandal” (Law.com Apr. 6, 2010).

Dr. Nolen’s book demonstrates that error is a part of complex human endeavors.  In 1968, Dr. Nolen wrote that, “[t]o stick with the appendicitis example, it’s estimated that surgeons are wrong about 15% of the time” in their diagnosis of a “simple” condition like appendicitis. Id. at 58.  And, of course, “[e]ven if he [or she] makes the proper diagnosis the surgeon isn’t home free. There are often two, three or more ways to treat a patient with a specific disease. . . .  Every step of the way, you have to choose between options.” Id.

The Sensient Colors court wrote that a litigant “should not be unduly punished for occasional mistakes that occurred while it started to use new software to organize and sort its documents.”  Sensient Colors, 2009 WL 2905474, at *4. Just as Dr. Nolen was not able to find the landmarks for the initial incision on his “victim” for his first surgery, the preservation, collection, processing, review, and production of ESI is far from a science in this, the fifth year after the “ESI rules.”

By necessity, ESI is a team effort. Similarly, Dr. Nolen wrote that:

Every operation is a team event.  A surgeon can no more operate on a patient by himself [or herself] then can a general fight a war alone.  An operation, like a battle, is a co-operative venture.

Id. at 217.  ESI attorneys rely on complex software and service providers as part of the team. They have no other choice. Wholesale manual review of terabytes of information is not possible. In Heriot, the court wrote that a litigant “relied, and should be able to rely, on their Vendor to faithfully carry out the instructions it had been given.”  257 F.R.D. at 660.

Proportionality, transparency, and cooperation are central concepts.  “The public has an interest in effective and efficient legal proceedings calculated to lead to fair and just results. The litigants have an interest in having their legal matters addressed fairly, promptly and economically.”  Gould v. Charlton Co., Inc.,  929 S.W.2d 734, 737-38 (Ky. 1996). Each case turns on its unique facts, however, as in Datel Holdings, “fairly robust” measures, even if imperfect, should generally suffice.


One issue in Mt. Hawley was that the recipient of the inadvertent production successfully argued that it had no duty to notify the producing party of its receipt of privileged information.  “Notice” should be a topic discussed in Rule 26(f) conferences. M. Berman, “The Rule 26(f) Conference of Parties,” in M. Berman, C. Barton, and P. Grimm, eds., “Managing E-Discovery and ESI: From Pre-Litigation Through Trial” (ABA forthcoming Aug. 2011), Chap. 19.

For a discussion of search methodology, see Chapters 21 and 22 of M. Berman, C. Barton, and P. Grimm, eds., “Managing E-Discovery and ESI: From Pre-Litigation Through Trial” (ABA forthcoming Aug. 2011).


R. Losey, “Winning isn’t everything, it’s the only thing” (“Perfection is not attainable, but if we chase perfection we can gain excellence.”)

“[T]he perfect is the enemy of the good,” P. Grimm, L. Bergstrom, and M. Kraeutter, “Federal Rule of Evidence 502: Has It Lived Up To Its Potential?,” 17 Rich.J.L. & Tech. 13, 41, quoting Voltaire Quotes, Famous Quotes.

Thorncreek Apartments III, LLC v. Village of Park Forest, 2011 WL 3489828 (N.D. Ill. Aug. 9, 2011)(documents tagged for privilege during document review were inadvertently produced in online database.  Court held that procedures were not reasonable. Court cited statements of counsel that “all” documents were produced, inadequate description of precautions taken, failure to timely notice the production, and failure to timely prepare a privilege log.  “Perhaps the most telling in assessing the adquacy of the safeguards is the abject failure of the Village’s processes to protect allegedly privileged documents.  The Village did not succeed in identifying and withholding from production even a single privileged document.”).

B. Kerschberg, “Privilege Waived? Federal Court Says Don’t Blame Your Electronic Discovery Vendor,” Forbes (Aug. 19, 2011)(“Thorncreek may produce a sigh of relief among e-discovery vendors, at least based on these facts. There are myriad scenarios that are equally plausible and would probably result in a different outcome, but the courts will address them if and when they arise. In the meantime, counsel has been warned: electronic discovery is not the bailiwick of third-party vendors alone. Absent active involvement in—and indeed control of—the e-discovery process, serious repercussions may ensue.”).

The Sedona Conference Best Practicies Commentary on the Use of Search and Information Retrieval Methods in E-Discovery, 8 Sed.Conf.L.J. 189, 204, 211 (2007)(“The discovery standard is, after all, reasonableness, not perfection. . . . [T]here is no requirement that ‘perfect’ searches will occur. . . . A standard of absolute perfection is and always has been unrealisitic, but now, with quantitative data available, we know perfection is  not only unrealistic, but also quite simply unachievable. . . . ‘[P]erfection should not be allowed to be an enemy of the attainable and reasonable goal of reasonableness.”).

This post was reproduced by Litigation Support Technology and News.