Summary
declining to authorize "batting average" statistical discovery
Summary of this case from Atkins v. UPMC Healthcare Benefits TrustOpinion
No. 12 C 2782
07-26-2013
DEBRA WARNER, Plaintiff, v. UNUM LIFE INSURANCE COMPANY OF AMERICA, Defendant.
Jeffrey T. Gilbert
Magistrate Judge
MEMORANDUM OPINION AND ORDER
This matter is before the Court on Plaintiff Debra Warner's Motion to Compel Discovery [Dkt.#39]. Warner filed this lawsuit against Defendant Unum Life Insurance Company of America seeking the reinstatement and payment of long-term disability benefits pursuant to the Employee Retirement Income Security Act of 1974 ("ERISA"), 29 U.S.C. § 502(a)(1)(B). In this motion, Warner seeks to compel discovery related to whether Unum's alleged structural conflict of interest as the simultaneous plan administrator and payer of benefits biased its decision to deny Warner's application for long-term disability benefits. The parties agree that the standard of review on the merits of this case is the arbitrary and capricious standard, and that the issue raised in this motion is what discovery, if any, should be permitted under such a standard. For the reasons stated herein, the Motion to Compel Discovery is granted in part and denied in part.
I. BACKGROUND
Plaintiff last worked as a Nurse Manager at Tyson Foods Inc. on January 19, 2011, at which time she alleges that she became disabled and unable to perform her job due to fibromyalgia, chronic fatigue syndrome, chronic back pain, seizure disorder and other impairments. At one point, Warner's primary physician diagnosed her as suffering from Lupus and Lyme disease, but those diagnoses later were ruled out. When she stopped working, Warner was insured under a group long-term disability insurance plan, Policy No. 112539, provided by her employer, which was both underwritten and administered by Unum.
Warner also was insured under a group policy of life insurance, Policy No. 112392, which provided for a continuation of coverage under a waiver of premium in the event she should become totally disabled which the policy defines as "unable to perform any work for compensation or profit as a result of injury or sickness." Comp.[Dkt.#1], at ¶9.
Warner applied for long-term disability benefits on April 22, 2001. Unum denied her application for benefits on July 29, 2011. Warner then appealed Unum's decision and submitted a functional capacity evaluation from her physical therapist. Unum reaffirmed its denial of benefits in March 2012. Warner then filed this lawsuit seeking the reinstatement and payment of her long-term disability benefits from May 5, 2011 to the present.
II. DISCUSSION
A. Limited Discovery May Be Permitted In ERISA Cases Under The Arbitrary And Capricious Standard In Appropriate Cases
In ERISA cases governed by the arbitrary and capricious standard, discovery generally has not been permitted, and judicial review is limited to the administrative record. Krolnik v. Prudential Ins. Co. of America, 570 F.3d 841 (7th Cir. 2009); Gutta v. Standard Select Trust Ins. Plans, 530 F.3d 614, 619 (7th Cir. 2008). In "exceptional cases," however, limited discovery beyond the claim file may be permitted when "a prima facie showing of impropriety has been made... ." Semien v. Life Ins. Co. of America, 436 F.3d 805, 814-15 (7th Cir. 2006). As discussed below, though, the Seventh Circuit recently walked back the standard it set in Semien, in light of the United States Supreme Court's decision in Metro Life Ins. Co. v. Glenn, 554 U.S. 105 (2008), "softening" the threshold showing a plaintiff must make to obtain discovery. See Dennison v. MONY Life Retirement Income Security Plan for Retirees, 710 F.3d 741, 747 (7th Cir. 2013).
In Semien, the Seventh Circuit articulated clear rules for district courts to apply in determining whether conflict of interest discovery is appropriate. First, a plaintiff "must identify a specific conflict of interest or instance of misconduct." Id. at 815. Second, the plaintiff "must make a prima facie showing that there is good cause to believe limited discovery will reveal a procedural defect in the Plan administrator's determination." Id.
In her opening motion and brief, Warner argued that the United States Supreme Court's decision in Metro. Life Ins. Co. v. Glenn, superceded Semien. The Supreme Court in Glenn, however, addressed only the standard of review appropriate in cases like Warner's; it did not address the scope of discovery available to a plaintiff such as Warner. In Glenn, the Supreme Court held that a conflict of interest is one of many factors to be considered in determining if an administrator's decision was made arbitrarily, because "any one factor will act as a tiebreaker when the other factors are closely balanced, the degree of closeness necessarily depending upon the tiebreaking factor's inherent or case-specific importance." 544 U.S. at 117. The Supreme Court reaffirmed that an administrator's structural conflict of interest as insurer and claims administrator does not change the standard of judicial review from arbitrary and capricious to de novo. Id. at 115-16.
While Warner's motion was being briefed, the Seventh Circuit, in Dennison v. MONY Life Retirement Income Security Plan for Retirees, addressed specifically the scope of discovery available in a case to be decided under the arbitrary and capricious standard of review in light of both Semien and Glenn. In the wake of Glenn but before the Seventh Circuit's decision in Dennison, district courts in the Seventh Circuit were split on the issue of how much discovery should be permitted in cases in which a structural conflict of interest is alleged and to what extent, if any, the Supreme Court's decision in Glenn modified or abrogated the Seventh Circuit's decision in Semien. The Seventh Circuit's decision in Dennison resolved that question.
See, e.g., Hoffman v. Sara Lee Corp., 2012 WL 404496 (N.D. Ill. 2012) (holding that the plaintiff should be allowed to explore, in a limited fashion, whether any structural conflict motivated the plan administrator's decision); Baxter v. Sun Life Assurance Co. of Canada, 713 F. Supp. 2d 766 (N.D. Ill 2010) (permitting limited discovery beyond the administrative record on the alleged structural conflict of interest); Gessling v. Group Long Term Disability Plan For Employees of Sprint/United Management Company, 639 F. Supp. 2d 947, 948 (S.D. Ind. 2009); Parrilli v. Hartford Life and Accident Ins. Co., 2009 WL 7447751 (N.D.Ill. 2009) (granting discovery beyond the claim file based on Glenn); Marszalek v. Marszalek & Marszalek Plan, 2008 WL 4006765 (N.D.Ill. 2008) (finding Semien still applies).
In Dennison, the Seventh Circuit made clear that the Supreme Court's decision in Glenn "suggests a softening, but not a rejection" of the requirement established in Semien that a plaintiff must make a prima facie showing in order to be entitled to discovery in a case to be decided under the arbitrary and capricious standard. 710 F.3d at 747. The Court in Dennison reaffirmed that Semien remains good law in the Seventh Circuit after Glenn albeit with a "softening" of the prima facie showing required as a prerequisite to obtaining discovery. Accordingly, the Seventh Circuit seems willing to see what discovery the district courts will permit under a "softened" Semien standard before it lays down firm markers in that regard. Id.
While the Seventh Circuit expressly declined to delineate "the contours of permissible discovery," it also implied that there are cases in which discovery may not be necessary. Id. Therefore, in light of Semien, as "softened" by Dennison, discovery still is not permitted in the run-of-the-mill case in the Seventh Circuit, and the two-part test established in Semien remains instructive. That means that to obtain discovery beyond the claim file in a case governed by the arbitrary and capricious standard, a plaintiff still must identify a specific conflict or instance of misconduct and make a prima facie showing that there is good cause to believe that limited discovery will reveal a procedural defect. Semien, 436 F.3d at 813-814. In light of Glenn, however, and given the softening of the Semien standard heralded by Dennison, a plaintiff's burden in making this showing is not onerous.
B. Warner Has Made A Threshold Showing Of A Conflict Of Interest Or Instance Of Misconduct Sufficient To Allow Limited Discovery Under The "Softened" Semien Standard
Warner argues that she has cited sufficient evidence of a specific conflict of interest or an instance of misconduct to warrant discovery beyond the claim file in this case. She points to Unum's alleged disregard of her physical therapist's functional capacity evaluation in December 2011 in connection with its decision to reaffirm its denial of benefits in March 2012; its reliance on file reviews by consulting physicians instead of in-person examinations; Unum's alleged historical hostility toward fibromyalgia claims; and its history of biased claims administration that sparked an investigation of its claims practices by the United States Department of Labor ("DOL") and dozens of state regulatory authorities, and culminated with Unum paying a multi-million dollar fine and hundreds of millions of dollars to past claimants. Pl's Reply [Dkt.#58], at 1.
In response, Unum argues that it did not disregard Warner's functional capacity evaluation but that it properly weighed it against other objective medical evidence in reaffirming its denial of benefits. Unum also argues that it is entirely appropriate for it to rely on claim file reviews and that one of its medical consultants even tried to contact Warner's primary treating physician by telephone without success. Finally, Unum says it made structural changes in its claims administration practices in the wake of its settlement with the DOL and state regulators and that Warner's allegations of its historical hostility to fibromyalgia claims are overblown.
The Court recognizes that Unum's settlement of claims by the DOL and the state insurance regulators alleging that a culture of bias resulted in Unum denying claims based on subjective symptoms, such as fibromyalgia, was nine years ago and seven years before it initially denied Warner's benefit claim. Unum paid a $15 million fine in 2004 and over at least the next four years paid out another $676 million in benefits to people whose claims had been denied. Unum's Response, at Exhibits K and L, [Dkt.#56-11 and 56-12]. In connection with that settlement, Unum was required "to completely restructure its claim handling procedures to ensure objectivity and fairness." Id. at Exhibit L. Unum asks the Court to take judicial notice of statements made by the New York State Department of Insurance in 2008 that Unum had complied with the settlement and made changes in its claims administration procedures required by that settlement. Id. Based on these statements, Unum apparently suggests that it should be afforded a clean bill of health with respect to claims administration.
Without determining when Unum should be liberated from the cloud of its previous conduct, Unum has enough of a documented "history of biased claims administration," Glenn, 554 U.S. at 117, that it was working to correct in close enough temporal proximity to its consideration of Warner's benefits claim in 2011 to warrant scrutiny in this case on the conflict of interest issue. In addition, Warner has raised legitimate issues with Unum's decision not to give much weight to the functional capacity evaluation submitted by her physical therapist in December 2011 and its reliance on physician file reviews only, in a case in which the medical reasons for Warner's symptoms were somewhat of a moving target, to justify discovery beyond the claims file under the "softened" Semien standard.
For example, Warner saw a physical therapist twelve times in 2011 but her therapist's functional capacity evaluation in December 2011 merited less than a two-line, cursory comment in Dr. Rodela's report in February 2012 that Unum used to support its denial of Warner's appeal. Unum's Response [DE#56-9] at Exhibit I, p. 4. In addition, although Unum reviewing physician, Dr. Folkening, apparently felt it was important to speak with Warner's family treating physician, Dr. Jeffries, before completing his file review, he did not do so. Unum's Response [DE#56] at pp. 3-5 and Exhibit B [DE#56-3]. Instead, he spoke with Dr. Sureka, a physiatrist, who had seen Warner only once as of his conversation with Dr. Folkening and who apparently assumed incorrectly (in connection with a later report) that Warner had been diagnosed with Lupus and Lyme disease. Id. and Exhibit D [DE#56-4].
In its Response to Warner's Motion to Compel, Unum argues that its decision to deny Warner benefits is supported, in part, by objective tests establishing that Warner did not have Lupus or Lyme disease (Unum's Response [DE#56] at 6). But Warner's treating physicians ruled out those diseases; her primary diagnosis was low back pain and osteoarthritis with a secondary diagnosis of fibromyalgia. Warner's Reply [DE#58-3] at Exhibit A. These diagnoses were confirmed by Warner's long-time physician, Dr. Couri, well-before Unum's decision to deny Warner benefits. Unum's Response [DE#56-5] at Exhibit E. Yet Unum continues to raise the withdrawn Lupus and Lyme disease diagnoses as a basis for its decision to deny Warner benefits. Unum's Response [DE#56] at 6. Warner has articulated a reasonable concern that there are circumstances at play here such that a conflict of interest would act as an important tiebreaker if the other factors relevant to a court's decision under the arbitrary and capricious standard are closely balanced. Glenn, 544 U.S. at 117.
Unum admits it was wrong for it to argue in its Response that it rejected Warner's physical therapist's functional capacity evaluation because he mistakenly said Warner had Lupus and Lyme disease. Unum's Response [DE#56] at 5. The physical therapist did not make that mistake; he correctly noted Warner's actual diagnoses in his report. While Unum has withdrawn that argument, the Court notes, parenthetically, that one of the premises of that argument -- that a physical therapist's mistake as to one of his patient's diagnoses is a good reason to reject the therapist's entire functional capacity evaluation — remains suspect. Unum also argues that it discounted the physical therapist's evaluation because his assessment of Warner's pain upon movement of her extremities conflicted with the assessment of another treater, Dr. Bobbrey. Id. These arguments advanced by Unum in support of its benefits determination lend weight to Warner's contention that Unum's reliance on file reviews only in a case like this, with shifting or conflicting diagnoses and assessments of a Claimant's subjectively reported symptoms, may indicate a conflict-induced bias against fibromyalgia claims. While the Court expresses no opinion on the ultimate merits of Warner's claim, these are relevant considerations under Semien in the context of deciding the scope of permissible discovery.
Accordingly, for all of these reasons, the Court finds that Warner has made a prima facie or threshold showing of a conflict of interest or an instance of misconduct, and that there is good cause to believe limited discovery will reveal a procedural defect in Unum's determination, sufficient to warrant limited discovery here under the "softened" Semien standard. No single factor is determinative, but all of this together suggests this may not be a run-of-the-mill case. Warner thus has tipped the scale enough to allow her to look behind the curtain to see what more, if anything, may be there on the conflict of interest issue.
C. The Scope Of Discovery To Which Warner Is Entitled On The Conflict Of Interest Issue
The issue then becomes what discovery is appropriate. Warner specifically seeks discovery regarding: (1) the professional qualifications of Unum's reviewing doctors - Dr. James Folkening, Dr. Susan Council and Dr. Laina Rodela; (2) the compensation these doctors received from Unum between 2007 and 2011; (3) the performance evaluations for those doctors from 2007 to 2011; (4) the number of claims reviewed by those doctors between 2007 and 2011 that resulted in Unum's denial of benefits; (5) the bonus structure and financial incentives in place for employees handling Unum's claims, appeals staff, and in-house medical and vocational staff; (6) the financial performance of the Policy; and (7) Unum's claims administration procedures, guidelines, and materials in effect at the time Warner's claim was denied. Pl's Br. [Dkt.#39], at 4.
In response to Warner's requests, Unum has agreed to produce (1) the medical qualifications and employment history of Dr. Folkening, Dr. Council and Dr. Rodela; (2) Dr. Rodela's 1099 forms for 2011 and 2012 showing her compensation from Unum; (3) internal company materials relating to compensation, bonus, and incentive programs applicable to the benefits personnel and medical consultants employed by Unum, including Drs. Folkening and Council; and (4) Unum's Benefits Center Claims Manual. Unum's Resp. [Dkt.#56], at 2. Unresolved is what additional information, if any, should be produced, specifically performance evaluations of doctors who performed the claim review, compensation information for Dr. Folkening and Dr. Council, statistical evidence regarding claims reviewed by the physicians, and information on the profitability of Unum's Policy.
The compensation materials include Unum's Management Incentive Compensation Plan, Stock Incentive Plan, Compensation Programs Summary and the Benefits Center Recognition Programs Brochure. Unum's Resp. [Dkt.#56], at 2.
--------
The Supreme Court's decision in Glenn and subsequent Seventh Circuit cases, including Dennison, do not open up ERISA cases to the full discovery allowed in typical civil cases. The discovery must be tailored to what is necessary and reasonable to probe the conflict of interest allegations and the safeguards in place to protect the appeals process from bias. At the same time, the Court recognizes that Warner is looking at the proverbial "black box" and does not know what is inside that may bear on Unum's alleged structural conflict of interest. See Barrett v. Life Insurance Co. of North America, 2012 WL 3544839, at *4 (N.D. Ill. Aug. 16, 2012). Yet, even in cases in which discovery has been permitted, it generally has been very limited. See, e.g., Hoffman v. Sara Lee Corp., 2012 WL 404496 (N.D. Ill. 2012) (holding that the plaintiff should be allowed to explore, in a limited fashion, whether any structural conflict motivated the plan administrator's decision); Baxter v. Sun Life Assurance Co. of Canada, 713 F. Supp. 2d 766 (N.D. Ill. 2010) (permitting limited discovery beyond the administrative record on the alleged structural conflict of interest). Pointedly, in neither Hoffman nor Baxter did the courts allow the type of broad discovery into claims administration decisions about other claims that Warner seeks in this case.
Under the circumstances of this case, the Court finds that compensation and bonus information for the doctors employed by Unum (Drs. Folkening and Council) and Dr. Rodela is discoverable based on Warner's theory that it may reveal circumstantial evidence of structural bias. In the Court's view, Unum implicitly acknowledged that compensation information is relevant by agreeing to produce some compensation information for Dr. Rodela as well as materials relating to Unum's compensation, bonus, and incentive programs applicable to benefits personnel and medical consultants employed by Unum. It is not an unreasonable leap from there to the actual compensation and bonuses received by all the reviewing doctors with an appropriate confidentiality designation. Therefore, the Court concludes that Unum must produce such compensation and bonus information for Dr. Folkening, Dr. Council and Dr. Rodela for the five years preceding their determination on Warner's claim, which is the time frame requested by Warner. Other courts, in this circuit and elsewhere, have allowed discovery of this type of compensation information. See, e.g., Hall v. Life Ins. Co. of N. Am., 265 F.R.D. 356, 367 (N.D. Ind. 2010).
In addition, Unum must produce the performance evaluations for Dr. Folkening, Dr. Council and Dr. Rodela for the same five years. Compensation and bonus information in combination with performance evaluations for the reviewing physicians could provide circumstantial evidence of a structural conflict or bias. See Gessling v. Group Long Term Disability Plan For Employees of Sprint/United Management Company, 639 F. Supp. 2d 947, 948 (S.D. Ind. 2009).
As for the statistical information concerning other claims reviews, the Court is reluctant to go down the route of what other courts have characterized as "batting average" statistics for similar claims or to permit Warner to review the claims files for a discrete number of other benefits applicants, an alternative that Warner suggested in her briefs and at oral argument. The Seventh Circuit has warned, ''Glenn does not invite a 'batting average' approach, assessing conflict by comparing the number of benefits decisions affirmed and reversed in federal court. (The sampling problems with that approach would be daunting.)" Holstrom v. Metro. Life Ins. Co., 615 F.3d 758, 767-768 (7th Cir. 2010). Judge Feinerman in Garvey v. Piper RudnickLLP Long Term Disability Ins. Plan, 2012 WL 1079966 (N.D. Ill. Mach 30, 2012), also refused to use a similar batting average approach in assessing structural conflicts. "The Seventh Circuit has warned that sampling problems render useless this type of 'batting average approach,' in which prior adverse court decisions are used to infer bias on the part of a plan administrator." Id. at *9; see also Hoffman, 2012 WL 404496, at *5 (denying plaintiff's request for discovery into the disability claims of other insureds).
Although Warner is not requesting UNUM's won-loss record in court decisions, her request for the "up or down" statistics for Unum's reviewing physicians in other cases suffers from the same statistical and sampling errors as the won-loss record discussed by other courts. Such statistical evidence ignores the potentially determinative nuances in individual cases. Just because a doctor's assessment resulted in a denial or grant of benefits in another case does not make it more likely that the denial of benefits in this case — with its unique facts — was infected by bias. The number of claims handled by a particular medical reviewer, or the number of those claims that were denied or recommended for non-payment, without more information about those individual claims, is at best weak circumstantial evidence of bias — the search for which does not justify the burden and cost on Unum to allow Warner to pursue it.
Each claim is different, and any statistical evidence that Unum produced would have to be evaluated in the context of each such claim - or at least Unum would argue that it is entitled to the opportunity to do so. In order to evaluate the probative value of that evidence beyond its surface appeal or, alternatively, its prejudicial effect, the district court effectively would have to conduct mini-trials concerning each individual benefits decision. The Court finds that such information has minimal relevance and would be an overly burdensome frolic here given the limited discovery permitted in a case to be decided under the arbitrary and capricious standard even considering the "softened" Semien test.
As an alternative, Warner suggests that Unum be required to produce a more limited number of file reviews, specifically the five file reviews performed by the doctors before and after their review of Warner's case. But Warner's suggestion that she would be satisfied with this limited production of the actual file review reports is at least a tacit acknowledgment that, eventually, the specifics of each of those other cases could be fodder for discussion and argument in this case. Whether we are talking about ten cases, or dozens or hundreds of cases, the issue is the same. In the Court's view, this stretches too far the notion of limited discovery on the conflict of interest issue, at least in this case. Therefore, the Court denies Warner's Motion to Compel a statistical summary of the number of claims reviewed by Dr. Folkening, Dr. Council and Dr. Rodela between 2007 and 2011 and Warner's alternative request, raised during the parties' meet and confer and discussed on the record at the hearing held on July 2, 2013, for the five file reviews performed by Dr. Folkening, Dr. Council and Dr. Rodela before and after Warner's claim was denied.
The Court also declines to order that Unum produce information about the profitability of its Policy. Warner has failed to provide a sufficient reason as to how such information would be relevant to a structural conflict analysis.
D. The Communications Between Unum's Benefit Analysts And Its Legal Department Are Not Privileged
Finally, Warner has moved to compel Unum to produce documents that Unum has withheld on grounds of the attorney-client privilege. Warner challenges Unum's assertion of privilege over three communications between Unum's benefit analysts and its legal department on September 20, 26, and 28, 2001, arguing that the documents fall within the "fiduciary exception" to the privilege. "Under that exception, a fiduciary of an ERISA plan 'must make available to the beneficiary, upon request, any communications with an attorney that are intended to assist in the administration of the plan.'" Bland v. Fiatallis N. Amer. Inc., 401 F.3d 779, 878 (7th cir. 2005). The exception does not apply to "[d]ecisions relating to the plan's amendment or termination," which are "not fiduciary decisions." Id. at 788.
After an in camera review of the documents Unum has withheld, the Court finds that the documents relate to plan administration and do not contain any legal advice notwithstanding that they are identified as attorney client communications and are marked privileged and confidential. This lawsuit was not pending at the time these communications took place. Although Warner's counsel sent Unum a letter before these documents were created raising the possibility of an appeal of Unum's benefits decision and asking Unum to preserve all electronically stored information relevant to Warner's case, that is not enough to create or trigger an imminent threat of litigation or to categorize these communications as anything other than communications with an attorney to assist in the administration of the plan. Therefore, the Court finds that the documents are not privileged and shall be produced.
III. CONCLUSION
For all of the reasons discussed herein, Warner's Motion to Compel Discovery [Dkt.#39] is granted in part and denied in part with each party to bear its own costs and fees. The Court denies Warner's request for her attorneys' fees in filing this Motion.
It is so ordered.
_________________
Jeffrey T. Gilbert
United States Magistrate Judge