Farmers Ins. Co., Inc. v. Peterson
81 P.3d 659
Okla.
2003
Check Treatment

MEMORANDUM OPINION

SUMMERS, J.

1 1 Original jurisdiction is assumed. Okla. Const. Art. 7 § 4. Real Parties in Interest (Third-Party Plaintiffs or Plaintiffs) were granted, in part, a motion to compel discovery. The order of the District Court requires Defendant and Petitioner (Farmers Insurance Co.) to search claim files for the years 2000, 2001 and 2002 to determine those that contain complaints from Oklahoma insureds on med-pay claims. Farmers seeks to prohibit enforcement of that order.

1 2 Evidence submitted by Farmers shows that compliance with the order would require Farmers to manually examine approximately 600,000 closed files located in a Kansas storage facility. Farmers estimated that it would take a staff of thirty people working two months to review these files. Farmers currently employs a retrieval staff of six people for all of the retrieval needs of its business. After October 2001 Farmers used an electronic method for storing files, and provided to Plaintiffs a list of approximately 3300-3400 claims (approximately 70 pages with approximately 48 claims per page) that arose from accidents occurring in Oklahoma prior to January 1, 2008. The evidence shows that compliance with the order requires an employee to examine all 3800-3400 electronic files, and this task may take thirty minutes for each file.

T3 Discovery may be limited or denied when discoverable material is sought in an excessively burdensome manner. YWCA of Oklahoma City v. Melson, 1997 OK 81, ¶ 25, 944 P.2d 304, 312. We agree that requiring an examination of all paper and electronic files for the three-year period would be unduly burdensome. However, we also agree with Plaintiffs that Farmers' unilateral decision on how it stores information cannot, by itself, be a sufficient reason for placing discoverable matter outside the seope of discovery. See, eg., Fagan v. District of Columbia, 136 F.R.D. 5, 7 (D.D.C.1991), ("Plaintiffs should not suffer if the information is not easily accessible because defendants have an inefficient filing system."); Koglowski v. Sears, Roebuck & Co., 73 F.R.D. 73, 76 (D.Mass.1976), ("defendant may not excuse itself from compliance with Rule 34, Fed.R.Civ.P., by utilizing a system of record-keeping which conceals rather than discloses relevant records, or makes it unduly difficult to identify or locate them, thus rendering the production of the documents an excessively burdensome and costly expedition").

T4 Plaintiffs asserts that the material is sought for the purpose of their theory that *661 Farmers engaged in a pattern of conduct that was wrongful as it relates to med-pay claims. Statistical sampling is a common technique used to determine a pattern of conduct.

Statistical protocols represent standards for determining whether and under what cireumstances it is reasonable to extrapolate from a known universe to an unknown one. See Kaye & Freedman, Reference Guide on Statistics, Federal Judicial Center's Reference Manual on Scientific Evidence 382 (1994). They draw on probability theory to determine whether the observed variations likely depend upon chance, or whether they likely represent a pattern of intentional conduct.

U.S. v. Skodnek, 933 F.Supp. 1108, 1117 (D.Mass.1996), (emphasis added). See also Federal Judicial Center, Reference Manual on Scientific Evidence, 88-178 (2d ed.2000), (Reference Guide on Statistics).

For example, in the context of the federal government seeking recoupment of Medicare overpayments involving a large number of patient files, one court said that "courts have routinely permitted the use of statistical sampling to determine whether there has been a pattern of overpayments spanning a large number of claims where case-by-case review would be too costly." Chaves County Home Health Service, Inc. v. Sullivan, 289 U.S.App.D.C. 276, 931 F.2d 914, 919, (D.C.Cir.1991), cert. denied, 502 U.S. 1091, 112 S.Ct. 1160, 117 L.Ed.2d 408 (1992). Statistical sampling as discovery is not novel in the courts of law. See Rosado v. Wyman, 322 F.Supp. 1173, 1180 (E.D.N.Y.), aff'd, 437 F.2d 619 (2d Cir.1970), aff'd, 402 U.S. 991, 91 S.Ct. 2169, 29 L.Ed.2d 157 (1971): "Sampling has long been considered an acceptable method of determining the characteristics of a large universe. . Such mathematical and statistical methods are well recognized by the courts as reliable and acceptable in determining adjudicative facts."

15 Permissible inferences from particular statistical methods depend upon the appropriateness of a particular statistical method in a given set of circumstances. For example, the High Court has said that: "We caution only that statistics are not irrefutable; they come in infinite variety and, like any other kind of evidence, they may be rebutted. In short, their usefulness depends on all of the surrounding facts and circumstances." International Broth. of Teamsters v. U.S., 431 U.S. 324, 340, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977). A particular statistical method selected by one party may be challenged as either factually or legally deficient by the opposing party. Watson v. Fort Worth Bank and Trust, 487 U.S. 977, 996-997, 108 S.Ct. 2777, 101 L.Ed.2d 827 (1988).

T6 In Christian v. Gray, 2003 OK 10, 65 P.3d 591, we issued a writ and addressed the general nature of evidence used to show causation for a particular type of tort, but we expressly declined to address in detail the specific evidence that each of the parties should offer on the reliability of their respective expert witnesses. Id., 2003 OK 10, at ¶ 12, 65 P.3d at 600. We noted the absence of a record on the issue and left it for the parties to litigate in the District Court. Id.

T7 The parties in the case today do not discuss either the varying methods of selecting particular files for sampling or the proper discoverable information in those files that would necessarily support the litigation objective of Plaintiffs. They may, or may not, agree to a particular method. No authority is cited in support of placing a burden on the trial court to create, sua sponte, a statistical sampling discovery technique for parties. We decline to hypothesize on the specific parameters of a discovery request that uses statistical sampling prior to the parties addressing the issue in the District Court.

T8 The trial court's order of June 10, 20083, required Farmers to comply with discovery. The trial court denied, by order dated July 11, 2008, Farmers' motion for a protective order. A trial court's abuse of discretion in denying a discovery protective order may be corrected by an extraordinary writ issued by this Court. Inhofe v. Wiseman, 1989 OK 41, 772 P.2d 389, 393.

19 We hereby direct that the District Court's orders of June 10, 2003, and July 11, 2008, in Cause CS-2002-4453, District Court of Tulsa County, shall not be enforced to the *662 extent that they require Farmers to search all of its files for the three-year period and identify those in which a complaint by an Oklahoma insured was made concerning a med-pay claim. The extent to which Plaintiffs may obtain information from Farmers' files by discovery, using a workable statistical sampling technique which would meet the trial court's muster for integrity of the process and protect both litigants from distor-tive effects, is left for the parties to litigate in the District Court. WRIT ISSUED.

10 OPALA, V.C.J., LAVENDER, HARGRAVE, BOUDREAU, JJ., CHAPEL, S.J., sitting in lieu of KAUGER, J., who disqualified-Concur. 1 11 WATT, C.J., dissents and is joined by HODGES, J. "I would allow discovery of those claims that occurred within one year of Plaintiff's claim. 1 12 WINCHESTER, J,.-Dissents.