Full opinion text
OPINION HOCHBERG, District Judge. Having reached a settlement in this case of unprecedented intensity and duration, the parties now move for (1) final approval of (a) the proposed Settlement Agreement between Plaintiffs and Defendant Health Net and (b) the proposed Plan of Allocation; and (2) an award of attorneys’ fees, expenses, and payment of Plaintiff incentive awards. The Court held a preliminary fairness hearing on April 24, 2008. The final fairness hearing was held on July 24, 2008 and an order approving the Settlement was entered on July 25, 2008. This opinion sets forth the Court’s reasons for entering that order. I. BACKGROUND Health Net is a health care insurer that offers several kinds of plans, one of which is a “point of service plan.” As the Court previously explained, pointrof-service plans permit the subscriber to use in-network or out-of-network providers. An out-of-network (or non-participating) provider is a provider who is not part of Health Net’s network and does not have a contracted-for rate with Health Net. If a subscriber decides to go to an out-of-network provider, the subscriber is subject to deductible, coinsurance, allowable amounts, reasonable and customary amounts, and/or usual, customary, and reasonable charge limitations. Wachtel v. Guardian Life Ins. Co., 223 F.R.D. 196, 199 (D.N.J.2004). These cases arose from Health Net’s treatment of out-of-network (“ONET”) claims submitted by insureds with point-of-service plans. Plaintiffs’ Complaints challenged Health Net’s policies for determining the usual, customary, and reasonable (“UCR”) charge limitations used to calculate the amount of an insured’s ONET claim that Health Net will reimburse. As the Court previously explained, Health Net’s plan contracts do not cover an entire fee charged by an out-of-network provider. Rather, [Health Net] pays a percentage of a certain allowed charge.... [The allowed charge] is most often defined as the Usual, Customary, and Reasonable charge for the service provided. The beneficiary pays the remaining percent of the UCR charge and is responsible for the rest [i]f a medical bill ... exceeds the UCR charge.... Thus, the coverage for out-of-network treatment that a beneficiary receives depends heavily on how UCR is defined. Wachtel, 223 F.R.D. at 200. In order to determine UCR charges for ONET claims, Health Net relied on databases licensed from a third-party vendor, Ingenix. Plaintiffs allege that Health Net’s rebanee on these databases (the “In-genix databases”) was improper and in violation of ERISA, 29 U.S.C. § 1001 et seq. Specifically, Plaintiffs alleged that the Ingenix databases are inherently flawed and that each time Health Net determined an ONET reimbursement in reliance upon an Ingenix database, Health Net failed to pay all the benefits due to its insureds and thereby violated ERISA. The Scharfman Plaintiffs also brought a claim pursuant to the Racketeering Influenced and Corrupt Organizations Act (“RICO”) arising from the same facts. A. Procedural History The instant Settlement comes after seven years of extraordinarily contentious litigation. The Wachtel action was removed to this Court in August 2001; the McCoy complaint was filed as a “related case” in April 2003, followed by the Scharfman complaint in January 2005. The seven years of litigation in these cases defy simple summary. As the Court noted in 2006, The Wachtel and McCoy cases are two of the oldest on this Court’s docket. The litigation has been fierce and without respite, through several changes of defense counsel.... In sum, it gives new meaning to the term “scorched earth” litigation tactics. Wachtel v. Health Net, Inc., 239 F.R.D. 81, 84 (D.N.J.2006). The Wachtel docket sheet is now 115 pages long, with 141 motions, 283 briefs, 316 other applications, 44 hearings, 11 conferences, and 5 appeals to the Third Circuit. The age of these cases and their extraordinary procedural histories are not solely attributable to the complexity and importance of the databases used to calculate UCR reimbursements. Class notice for the Wachtel and McCoy classes was complete as of July 6, 2007. Following the Court’s preliminary approval of the Settlement Agreement and preliminary approval of the Scharfman RICO and ERISA classes on April 24, 2008, “smart notice” was mailed to all class members in Wachtel, McCoy, and Scharfman between May 19, 2008 and June 11, 2008. Those Wachtel and McCoy class members who had not previously opted out were given an additional opportunity to opt out. Those Wachtel and McCoy class members who had previously opted out were given an opportunity to waive their opt out and again become Class Members. Class members were given the opportunity to opt out or object by June 23, 2008. B. Mediation and Settlement The parties engaged in settlement negotiations before several mediators over the past four years. The first mediation took place in 2004 with a retired state court judge. The second negotiation took place in Florida in January 2005 with a retired federal judge. The third and fourth medi-ations took place with a private and renowned mediator. In August 2007, the parties informed the Court that they had reached a settlement in principle. This settlement, however, included only the sum of money to fund direct reimbursements; equitable relief and other crucial reforms of the UCR calculation process took many months longer to negotiate. Between August 9, 2008 and April 17, 2008, the parties continued to negotiate the remaining provisions of this Settlement, including important business practice changes. The Court held several phone and in-person conferences to monitor the parties’ progress on the remaining terms. Prior to the preliminary fairness hearing, the Court held a “tutorial” hearing to more fully explore the Ingenix database. The parties submitted a final Settlement Agreement to the Court on April 17, 2008. C. Settlement Agreement The present Settlement is among the largest ERISA health insurance settlements on record. The Settlement entails a substantial cash component as well as business practice changes that will have a lasting impact on the way Health Net reimburses its subscribers for out-of-network medical services. It also raises a clarion call for greater disclosure about the databases used for health care coverage. The UCR database flaws are discussed in section III.B.2(g)(l). Health Net will provide a cash fund of $215 million, divided into three categories. First, Health Net will provide a $40 million prove-up fund for Class Members who were Balance Billed by their ONET providers after Health Net reimbursed the ONET provider at a rate lower than the provider’s billed charge. Second, Health Net will provide a $160 million cash fund from which all Class Members are entitled to receive a pro rata reimbursement for claims subject to an erroneous ONET determination. Finally, $15 million of the fund will be paid to the New Jersey Department of Banking and Insurance (“DOBI”) and used in the discretion of that state agency to reimburse members of the New Jersey small employer plan Class (the Wachtel Class). The Settlement Agreement divides Class Members who have been Balance Billed by their ONET providers into two groups for purposes of allocating the prove-up fund. Group B consists of those Class Members who received a Balance Bill from their ONET provider after the beginning of their class period and can demonstrate that they paid some or all of the Balance Bill on or before April 24, 2008, the date of the Preliminary Fairness Hearing in this matter. Group C consists of those claims for covered services received by a Class Member on or after May 6, 2005 for which the Class Member received, but did not pay, a Balance Bill from their ONET provider on or before April 24, 2008. The closing date for both groups is established by the requirement that Health Net have in its claims system a check payment date for reimbursement to the ONET provider on or before July 31, 2007 (the date set by the Scharfman class definition). The third ground, Group A, consists of all remaining claims, including, for example, those claims for which Health Net reimbursed the ONET provider less than their full billed amount, but where the ONET provider did not Balance Bill the Class Member. The prove-up fund will be allocated first. Group B Class Members — those who have paid all or part of a Balance Bill — will be reimbursed first and they are entitled to receive up to 100% of the amount they paid out of pocket. If any portion of the $40 million prove-up fund remains after all Group B claims are satisfied in full, those funds will be used to satisfy Group C claims on a first-come, first-served basis. Any portion of the $40 million prove-up fund that remains after all Group B and C claims are satisfied reverts to Health Net. Any Group B or C claims that remain after the prove-up fund has been fully allocated become Group A claims for purposes of further distribution. All Group A claims will then be reimbursed pro rata from the $160 million pool. The $160 million pool is non-reversionary. Attorney’s fees and expenses will also be drawn from the non-reversionary $160 million cash fund. Health Net has also agreed to highly significant business practice changes, which it will perform until the four year anniversary of the Effective Date of the Agreement. First, and most importantly, Health Net has agreed to cease using the Ingenix database for determining UCR charges for ONET services or supplies. Indeed, Health Net has agreed to eliminate the use of UCR altogether for determining payments for its subscribers’ ONET services or supplies. Pursuant to the agreement, Health Net has agreed to eliminate the use of the Ingenix databases and UCR as soon after the Effective Date of the Settlement Agreement as possible, except where required by law or approved by regulators, or where specifically requested by a plan sponsor. The new methodology that Health Net develops to replace the Ingenix database and UCR will be fully disclosed in its health care plans. The specifics of Health Net’s new methodology are not presently before the Court and the Settlement Agreement does not purport to release Health Net from any claims that may arise from its use of new methodologies. The Settlement Agreement also requires Health Net to make business practices changes until Health Net can implement its new methodology. Health Net agrees that it will determine its reimbursement to ONET providers by reference to the current Ingenix database plus 14.5% (up to the billed charge). This will be referred to as the “Adjusted Allowable Amount.” According to the parties, this add-on will result in 80% of ONET claims being covered at 90% or more of the billed charge. Health Net will also institute a special appeals process for those subscribers who still have large outstanding balances even after the 14.5% add-on described above. A subscriber may take advantage of the special appeals process if the subscriber’s Adjusted Allowable Amount is less than 80% of the ONET provider’s billed charge and there is a difference of at least $4,500 between the adjusted allowable charge and the provider’s billed charge. These special appeals will be sent to an independent arbitrator, who will apply the “UCR Review Factors” set forth in the Settlement Agreement. The arbitrator’s decision will be binding on both Health Net and the subscriber. It is not mandatory that a subscriber use this special appeals process, and the decision to engage in binding arbitration is the choice of the Health Net subscriber. The cost of the special appeals process will be paid for by interest generated on the $160 million non-rever-sionary fund, which was deposited in escrow in January 2008. Those subscribers not eligible for the special appeals process, or who elect not to use it, may still use Health Net’s regular appeals process. Health Net will provide prompt and clear notice to beneficiaries who are eligible for the special appeals process. Health Net will also: (1) establish a “cost estimator process” so that subscribers can obtain accurate UCR amounts in advance of a procedure; (2) comply with the terms of its pre-authorizations and advance UCR determinations; (3) negotiate with the ONET provider in advance of non-emergent surgeries and if Health Net and the ONET provider reach a consensual fee arrangement, the subscriber will not be financially responsible for any amount over the applicable coinsurance and deductible amounts; (4) establish a web portal to keep subscribers appraised of ONET policies and procedures; (5) establish training programs and other procedures so that Health Net employees can provide members with more accurate information; (6) review the use of “adjustment codes” on subscribers’ Explanation of Benefits form to ensure that those codes convey meaningful information to the subscriber; (7) provide greater transparency for benefits related to, inter alia, multiple surgical procedures, assistant and co-surgeons. Health Net has estimated the business practices changes are worth between $26 and $38 million, and Plaintiffs concur. II. ANALYSIS A. Class Certification of the Scharfman classes The Court certified the Wachtel and McCoy classes in an order dated August 5, 2004. See Wachtel DKT# 139. On June 30, 2006 the Third Circuit vacated the Court’s August 5, 2004 class certification order and remanded to this Court “for a definition of the claims, issues, or defenses to be treated on a class basis.” Wachtel v. Guardian Life Ins. Co. of Am., 453 F.3d 179, 190 (3d Cir.2006). The Court issued a new class certification order on September 25, 2006 in which it set forth the claims, issued, or defenses to be treated on a class basis, but did not alter the class definitions given in the August 5, 2004 opinion. Wachtel DKT# 551. The Court certified the Wachtel and McCoy classes as follows: • Wachtel class: All persons in the United States who are, or were, from July 1, 1995 to August 31, 2004, subscribers or beneficiaries of any New Jersey small employer plan, who received medical services from an out-of-network provider and for whom Defendants made reimbursement determinations less than the providers’ actual charge. • McCoy class: All persons in the United States who are, or were, from April 1, 1997 to August 31, 2004, subscribers or beneficiaries of any large or small employer plan, other than in a New Jersey small employer plan, who received medical services or supplies (including, inter alia, surgery, anesthesia, and the like) from an out-of-network provider and for whom Defendants made reimbursement determinations less than the providers’ actual charge. See McCoy DKT# 89. The Scharfman ERISA and RICO classes were preliminarily certified at the Preliminary Fairness hearing held on April 24, 2008. The Scharfman classes are defined as follows: • Scharfman ERISA class: All person in the United States who are, or were, from September 1, 2004 through July 31, 2007 members of any large or small employer plan insured or administered by Health Net, and subject to ERISA, who received medical services or supplies (including, inter alia, surgery, anesthesia, and the like) from an out-of-network provider and received reimbursement less than the provider’s billed charge. • Scharfman RICO class: All persons in the United States who are, or were, from September 1, 2004 through July 31, 2007 members in any large or small plan insured or administered by Health Net, who received medical services or supplies (including, inter alia, surgery, anesthesia, and the like) from an out-of-network provider and received reimbursement less than the provider’s billed charge that was determined by Health Net, Guardian or a Third Party Vendor applying Health Net’s ONET Claims Practices, including the use of Ingenix data. Since preliminarily certifying the Scharfman classes on April 24, 2008, the Court has received no objections to the Scharfman classes as to Rule 23’s requirements of numerosity, typicality, adequacy, predominance or superiority. The Court now enters final certification of the Scharfman ERISA and RICO classes for the following reasons. 1. Numerosity Rule 23(a)(1) provides that the Court may certify a class only if “the class is so numerous that joinder of all members is impracticable.” The Third Circuit has held that although “[n]o minimum number of plaintiffs is required to maintain a suit as a class action, ... generally if the named plaintiff demonstrates that the potential number of plaintiffs exceeds 40, the first prong of Rule 23(a) has been met.” Stewart v. Abraham, 275 F.3d 220, 226-27 (3d Cir.2001). Joinder need not be impossible. “When dealing with a class that numbers in the hundreds, joinder will most often be impracticable.” Lenahan v. Sears, Roebuck and Co., No. 02-0045, 2006 WL 2085282, at *7 (D.N.J. July 24, 2006). The instant classes easily satisfy Rule 23(a)(l)’s numerosity requirement. There are well over 2 million Class Members in the Wachtel, McCoy, and Scharfman classes. The parties have stated that class notice was sent to over 2.5 million individuals who are or were insured by Health Net. The Court finds that Rule 23(a)(l)’s numerosity requirement is satisfied. 2. Commonality and Predominance Rule 23(a)(2) requires that there be “questions of law or fact common to the class.” The “predominance” requirement of Rule 23(b)(3) requires that the common questions predominate over individual issues. As the Court explained when it certified the McCoy and Wachtel classes: The commonality prong of [Rule] 23(a) overlaps with the predominance requirement of 23(b)(3). Therefore, courts frequently examine the two requirements together. Commonality under 23(a)(2) is satisfied if the named plaintiff shares common questions of law or fact with the grievances of the prospective class. Predominance, however, requires more than the existence of common issues of law and fact. The common issues must be numerically and qualitatively substantial in relation to individual issues. Wachtel, 223 F.R.D. at 213 (internal citations omitted). The Third Circuit has instructed that “[t]he commonality requirement will be satisfied if the named plaintiffs share at least one question of fact or law with the grievances of the prospective class.” Stewart, 275 F.3d at 227 (emphasis removed). Commonality is easily met by the Scharfman ERISA class. The Scharfman ERISA class brings the same claims as the McCoy and Wachtel ERISA classes, but for a later time period. Consequently, the claims of all three classes are based on common operative facts and questions of law. As the Court explained when it certified the Wachtel and McCoy classes: The Plaintiffs allege a systematic course of conduct in interpreting contracts of insurance in an improper, undisclosed, and self-serving way in contravention of the plans and of Health Net’s fiduciary duty to beneficiaries who chose to use out-of-network providers. In both the McCoy class and the Wachtel class, if each member of the potential class were to bring an individual action, each would be required to prove that Health Net’s UCR and other policies violated ERISA. The issues of law and fact relating to whether Health Net fully disclosed and properly applied its reimbursement mechanisms for out-of-network provider services are common to the class members and predominate over individual questions. Wachtel, 223 F.R.D. at 213 (internal citation omitted). The Scharfman RICO class also satisfies the commonality and predominance requirements of Rule 23(a)(2) and (b)(3). The Scharfman RICO class alleged an association-in-fact between Health Net and Ingenix in which Health Net and Ingenix carried out a scheme to underpay benefits to Health Net subscribers. See Scharfman DKT# 54 (Second Amended Complaint and RICO Case Statement). The Class’s RICO claims raise numerous common questions of law and fact concerning the existence of an association-in-fact, the performance of predicate acts including mail and wire fraud, and the existence of a common plan or scheme to underpay Health Net’s subscribers and to improperly shift the cost of medical care to Health Net subscribers. The issues of law and fact raised by the RICO claims are common to the Scharfman RICO class and predominate over any individual issues that may be present among RICO Class Members. The Court finds that Rule 23(a)(2)’s commonality requirement and Rule 23(b)(3)’s predominance requirement are satisfied. 3. Typicality Rule 23(a)(3) requires that “the claims or defenses of the representative parties are typical of the claims or defenses of the class.” As the Third Circuit has noted, “[t]he typicality inquiry centers on whether the interests of the named plaintiffs align with the interests of the absent members.” Stewart, 275 F.3d at 227. “[C]ases challenging the same unlawful conduct which affects both the named plaintiffs and the putative class usually satisfy the typicality requirement irrespective of the varying fact patterns underlying the individual claims.” Id. “Factual differences will not render a claim atypical if the claim arises from the same event or practice or course of conduct that gives rise to the claims of the [absent] class members, and if it is based on the same legal theory.” Id. at 227-28. Both Scharfman classes satisfy the typicality requirement of Rule 23(a)(3) because both classes’ claims arise from Health Net’s reliance on the Ingenix database to calculate UCR charges. Although facts of individual class members’ claims differ in some ways — the ONET treatments they received or whether they have exhausted the appeals process, for example — those minor difference do not alter the fact that all Plaintiffs rely on the same legal theories in both the ERISA class and the RICO class. The Court finds that the typicality requirement of Rule 23(a)(3) is satisfied. 4. Adequacy of Representation Rule 23(a)(4) requires that the “representative parties fairly and adequately protect the interests of the class.” The Third Circuit has held that “[a]dequate representation depends on two factors: (a) the plaintiffs attorney must be qualified, experienced, and generally able to conduct the proposed litigation, and (b) the plaintiff must not have interests antagonistic to those of the class.” New Directions Treatment Servs. v. City of Reading, 490 F.3d 293, 313 (3d Cir.2007). “In analyzing this criteria, the court must determine whether the representatives’ interests conflict with those of the class and whether the class attorney is capable of representing the class.” Johnston v. HBO Film Mgmt., Inc., 265 F.3d 178, 185 (3d Cir.2001). The Court found the adequacy requirement met under Rule 23(a)(4) in McCoy and Wachtel, and specifically noted that “Counsel for McCoy and the Wachtels are well-seasoned and have demonstrated adequacy and tenacity during the protracted proceedings that have already occurred in this case.” Wachtel, 223 F.R.D. at 216. As evidenced by the lengthy procedural history outlined above, in the years since the Court certified the Wachtel and McCoy classes, Class Counsel have proved themselves more than adequate to face the challenges posed by this litigation. At no time has there been any suggestion that the Scharfman representative Plaintiffs’ interests in any way conflict with the interests of the class. Consequently, the Court finds that the adequacy requirement of Rule 23(a)(4) has been met as to the Scharfman classes. 5. Superiority In addition to the “predominance” requirement, Rule 23(b)(3) also requires “that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” The rule provides the Court with four non-exhaustive factors to consider. The matters pertinent to these findings include: (A) the class members’ interests in individually controlling the prosecution or defense of separate actions; (B) the extent and nature of any litigation concerning the controversy already begun by or against class members; (C) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; and (D) the likely difficulties in managing a class action. Fed.R.CivP. 23(b)(3). Superiority is also easily satisfied in these cases for the reason given at the time of the Wachtel and McCoy class certifications. As the Court explained Class action is the superior form of litigation in this case because it ensures that potentially meritorious claims will be addressed efficiently and without waste of judicial resources. There is no indication that individual members of the class[es] have a compelling interest in controlling the prosecution of separate actions. Indeed, the high cost of prosecuting this complex case on an individual basis suggests that the opposite is true. This Court is not aware of pending litigation against Health Net by any proposed class members that would undermine the suitability of class litigation. The high cost of this complex litigation also suggests that meritorious claims may go unaddressed unless the Plaintiffs are permitted to proceed as a class. Relitigating the same issues and presenting similar evidence regarding Health Net’s policies and practices and non-disclosures for out-of-network charges would be inefficient and wasteful of judicial resources. Joinder or wholesale intervention would result in a multiplicity of repetitive actions. Wachtel, 223 F.R.D. at 217 (internal citations omitted); see also Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 620, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997) (“Confronted with a request for settlement-only class certification, a district court need not inquire whether the case, if tried, would present intractable management problems.”). Given that the facts that underlie the claims of both Scharfman classes are the same as those alleged by the Wachtel and McCoy classes, the same analysis applies here. B. Final Approval of Class Action Settlement and Plan of Allocation “The law favors settlement, particularly in class actions and other complex cases where substantial judicial resources can be conserved by avoiding formal litigation.” In re General Motors Corp. Pick-Up Truck Fuel Tank Prods. Liability Litig., 55 F.3d 768, 784 (3d Cir.1995). Class action settlements must receive approval from the Court under Rule 23(e), which requires that “[t]he claims, issues, or defenses of a certified class may be settled, voluntarily dismissed, or compromised only with the court’s approval.” Fed.R.Civ.P. 23(e). Before the Court may approve a settlement under Rule 23(e), the Court must be satisfied that it is fair, reasonable, and adequate. See In re General Motors, 55 F.3d at 785. 1. Presumption of Fairness The Third Circuit has “directed a district court to apply an initial presumption of fairness when reviewing a proposed settlement where: ‘(1) the settlement negotiations occurred at arm’s length; (2) there was sufficient discovery; (3) the proponents of the settlement are experienced in similar litigation; and (4) only a small fraction of the class objected.’ ” In re Warfarin Sodium Antitrust Litig., 391 F.3d 516, 535 (3d Cir.2004) (quoting In re Cendant Corp. Litig., 264 F.3d 201, 232 n. 18 (3d Cir.2001)). Each of the factors set forth in Warfarin counsels in favor of extending such a presumption to the instant Settlement agreement. There can be no doubt that the negotiations that led to this Settlement were undertaken at arm’s length. The Settlement Agreement was reached during the course of the fourth mediation attempt by the parties and after seven highly contentious years of litigation. “[T]he litigation has been fierce and without respite, through several changes of defense counsel” and Health Net’s litigation strategy in this case “gives new meaning to the term ‘scorched earth’ litigation tactics.” Wachtel, 239 F.R.D. at 84. For similar reasons, it is beyond dispute that there has been sufficient discovery in this case. The parties have produced tens of thousands of pages of documents and fought tooth and nail throughout the course of discovery, and two special masters have been assigned to serve in roles designed to ensure that proper discovery is made. The parties to this litigation have been represented by seasoned class action litigators throughout. As the Court noted when it certified the classes in McCoy and Wachtel, “Counsel for McCoy and the Wachtels are well-seasoned and have demonstrated adequacy and tenacity during the protracted proceedings that have already occurred in this case.” Wachtel, 223 F.R.D. at 216. The Court made this statement nearly four years ago. Since that time, Plaintiffs counsel have continued to zealously represent the classes and have admirably dealt with Defendants’ discovery abuses. Class Counsel has proven themselves capable litigators. Defendants’ present counsel is also an experienced class action litigator and he defended this action in a hard fought, persistent and determined manner. Notice of the instant Settlement was sent to over 2,572,342 subscribers and former subscribers, and notice was posted in sixty regional and national publications. See Rosenbaum Aff. Ex. B. Class Counsel has received 601 opt-outs in addition to the 1,107 timely opt outs received following certification and mailing of notice to the McCoy and Wachtel classes, representing a combined total of approximately 0.066% of all three classes. Class Counsel received nine objections. This small percentage of opt-outs and objections qualifies for the presumption of fairness under In re Warfarin Sodium. Cf. First State Orthopaedics v. Concentra, Inc., 534 F.Supp.2d 500, 516 (E.D.Pa.2007) (finding presumption of fairness met when “only 0.16% of the class opted out or objected to the settlement”); In re Remeron End-Payor Antitrust Litig., No. 02-2007, 2005 WL 2230314, at *16 (D.N.J. Sept. 13, 2005) (finding that 70 opt-outs and 8 objections from class of 850,000 qualified for presumption of fairness). Each of the factors set forth in In re Warfarin Sodium have been met and the Court will therefore extend an initial presumption of fairness to this Settlement Agreement. 2. Standard for Approval under Rule 23(e)(2) The Court’s analysis does not end with the factors set forth in In re Warfarin Sodium. The Court must next consider the factors discussed by the Third Circuit in Girsh v. Jepson, 521 F.2d 153 (3d Cir.1975). Girsh instructs the court to consider the following factors: (1) the complexity, expense and likely duration of the litigation; (2) the reaction of the class to the settlement; (3) the stage of the proceedings and the amount of discovery completed; (4) the risks of establishing liability; (5) the risks of establishing damages; (6) the risks of maintaining a class action through the trial; (7) the ability of defendants to withstand a greater judgment; (8) the range of reasonableness of the settlement fund in light of the best recovery; and (9) the range of reasonableness of the settlement fund to a possible recovery in light of all the attendant risks of litigation. In re AT & T Corp., 455 F.3d 160, 164-65 (3d Cir.2006) (quoting In re Rite Aid Corp. Secs. Litig., 396 F.3d 294, 301 n. 9 (3d Cir.2005) and Girsh, 521 F.2d at 157). The Third Circuit also recently noted that “[t]he Girsh factors do not provide an exhaustive list of factors to be considered when reviewing a proposed settlement.” Id. Rather, because of a ‘sea-change in the nature of class actions’ since Girsh was decided in 1975, district courts should also consider other potentially relevant and appropriate factors, including, among others: [T]he maturity of the underlying substantive issues, as measured by the experience in adjudicating individual actions, the development of scientific knowledge, the extent of discovery on the merits, and other factors that bear on the ability to assess the probable outcome of a trial on the merits of liability and individual damages; the existence and probable outcome of claims by other classes and subclasses; the comparison between the results achieved by the settlement for individual class or subclass members and the results achieved — or likely to be achieved — for other claimants; whether class or subclass members are accorded the right to opt out of the settlement; whether any provisions for attorneys’ fees are reasonable; and whether the procedure for processing individual claims under the settlement is fair and reasonable. Id. (quoting In re Prudential Ins. Co. Am. Sales Practice Litig. Agent Actions, 148 F.3d 283, 323 (3d Cir.1998)). The Court will now turn to all of the factors set forth by the Third Circuit. a. Complexity, Expense and Likely Duration of the Litigation This first factor “captures ‘the probable costs, in both time and money, of continued litigation.’ ” In re Cendant Corp. Litig., 264 F.3d 201, 233 (3d Cir.2001). “Courts must balance a proposed settlement against the enormous time and expense of achieving a potentially more favorable result through further litigation.” In re Remeron Direct Purchaser Antitrust Litig., No. 03-0085, 2005 WL 3008808, at *4 (D.N.J. Nov. 9, 2005). The complexity, expense and likely duration of further litigation in this case strongly counsels in favor of settlement. As described in Part I above, this case already has a long and often tortured procedural history. It took the parties seven years and well over a hundred motions merely to reach the point at which productive settlement talks could begin. Prior to trial, the parties would, among other things, be required to further examine the Ingenix database and analyze claims data for millions of subscribers. This would undoubtedly involve numerous experts and further pretrial motions, and thousands more hours billed to this matter. Further, the likelihood of appeal from any decision on the merits counsels in favor of approving the Settlement. See, e.g., In re Ikon Office Solutions, Inc., Secs. Litig., 194 F.R.D. 166, 179 (E.D.Pa.2000) (“Finally, the extremely large sums of money at issue almost guarantee that any outcome, whether by summary judgment or trial, would be appealed. This factor thus weighs in favor of the proposed settlement.”). Finally, “even if a trial resulted in a judgment for Plaintiffs, such judgment might not equal the amount of the Settlement, while Plaintiffs would have incurred additional expense and delay, as well as the risk of non-recovery based on a verdict for Defendants or reversal of a verdict for Plaintiffs on appeal. Therefore, this factor weighs in favor of approving the Settlement.” In re Remeron Direct, 2005 WL 3008808, at *5. The Court can safely conclude that this factor favors settlement because continuing litigation through trial would have required additional discovery, extensive pretrial motions addressing complex factual and legal questions, and ultimately a complicated, lengthy trial. Moreover, it was inevitable that post-trial motions and appeals would not only further prolong the litigation but also reduce the value of any recovery to the class. In a class action of this magnitude, which seeks to provide recovery for [Health Net subscribers], the time and expense leading up to trial would have been significant. In re Warfarin Sodium Antitrust Litig., 391 F.3d at 536. The Court concludes that this factor counsels in favor of approving the Settlement. b. The Reaction of the Class to the Settlement The Court must next consider the reaction of the Class to the Settlement Agreement. Notice was sent to 2,572,342 current and former Health Net subscribers, and as noted above, Class Counsel has received 601 opt-outs in addition to the 1,107 timely opt outs received following certification and notice in McCoy and Wachtel. These 1,708 individuals represent approximately 0.066% of the total class. Class Counsel has received only nine objections. In a class of this size, this can only be considered an overwhelmingly favorable response. Cf. Stoetzner v. U.S. Steel Corp., 897 F.2d 115, 118-19 (3d Cir.1990) (holding that “only” 29 objections in 281 member class “strongly favors settlement”); see also In re Prudential, 148 F.3d at 318 (affirming conclusion that class reaction was favorable where 19,000 policyholders out of 8 million opted out and 300 objected). “Such acceptance of the Settlement on the part of the Class is convincing evidence of the Settlement’s fairness and adequacy.” In re Remeron Direct, 2005 WL 3008808, at *6. c. The Stage of Proceedings and Amount of Discovery Completed The Court must next consider the stage of the proceedings and the amount of discovery completed to date. The Third Circuit has explained that this factor is intended to ensure “that a proposed settlement is the product of informed negotiations” and that “[t]he parties ... have an ‘adequate appreciation of the merits of the case before negotiating.’ ” In re Prudential Ins. Co. Am. Sales Practice Litig. Agent Actions, 148 F.3d 283, 319 (3d Cir.1998). The procedural history reflected in the 115 page docket sheet of this litigation counsels in favor of the Settlement. The parties engaged in five years of hard-fought discovery, during which time Defendants were sanctioned for discovery violations more than once. The parties made 141 motions, most of which were related to the contentious discovery process. “Given this vast amount of discovery obtained, and the volume of motion practice that enabled Plaintiffs’ Counsel to preview some of the defenses that Defendants would advance, Plaintiffs’ Counsel had a valid basis to negotiate a settlement.” In re Remeron Direct, 2005 WL 3008808, at *6. After years of discovery and motion practice before this Court, “the parties certainly [had] a clear view of the strengths and weaknesses of their cases.” Bonett v. Educ. Debt Serv., Inc., 2003 WL 21658267, at *6 (E.D.Pa. May 9, 2003) (quoting In re Warner Commc’ns Sec. Litig., 618 F.Supp. 735, 745 (S.D.N.Y.1985)). The Settlement presently before the Court “occurred only after the parties and the Court were able to assess its fairness adequately.” In re Remeron Direct, 2005 WL 3008808, at *7. The Court finds that the stage of the proceedings and the amount of discovery completed counsels in favor of the instant Settlement. d. The Risks of Establishing Liability and Damages “The fourth and fifth Girsh factors consider the risk of establishing liability at trial in order to balance the parties’ relative likelihood of success against the immediate benefits derived from a settlement.” Lenahan, 2006 WL 2085282, at *14 (citing In re Prudential, 148 F.3d at 319); see also In re Remeron Direct, 2005 WL 3008808, at *7. Because establishing damages will be contingent on establishing liability, the same concerns animate both of these elements of the Girsh test. The size of the present Settlement is some indication of the strength of the Plaintiffs’ case. Nevertheless, these cases present many difficult questions of law, the presence of which create a risk that Plaintiffs would not be able to establish liability at trial or on summary judgment. For example, there are genuine issues as to the liability of the parent company, Health Net, Inc., the availability of injunctive relief under ERISA, and the appropriate standard of review to be applied to Health Net’s ONET determinations. Further, Plaintiffs faced a substantial hurdle with regard to Health Net’s defense that New Jersey required the use of the Ingenix database. Plaintiffs would have sought to establish that New Jersey was unaware of the defects in the database, and also that, in any event, the state requirement did not obviate Health Net’s duties under ERISA. Further, with regard to the Ingenix database, Plaintiffs’ arguments rely on expert testimony, which has inherent risks. See, e.g., In re Prudential Ins. Co. of Am. Sales Practices Litig., 962 F.Supp. 450, 539 (D.N.J.1997) (“another potential risk may be plaintiffs’ necessary reliance on expert testimony to establish liability and damages; a jury’s acceptance of expert testimony is far from certain, regardless of the expert’s credentials.”); see also In re Vicuron Pharms., Inc. Secs. Litig., 512 F.Supp.2d 279, 285 (E.D.Pa.2007) (“Compelled to choose between experts, it is far from certain that a jury would have found for the class, much less awarded it damages on the order of the settlement agreement.”). Finally, if Plaintiffs established liability, the extent of damages would flow from that determination. The parties have insisted on vastly different methodologies for determining damages. In short, determining damages “is a complicated and uncertain process” that counsels in favor of settlement. See In re Remeron Direct, 2005 WL 3008808, at *8. e. Risk of Maintaining Class Action through Trial Because “ ‘the prospects for obtaining certification have a great impact on the range of recovery one can expect to reap from the [class] action,’ this factor measures the likelihood of obtaining and keeping a class certification if the action were to proceed to trial.” In re Warfarin Sodium, 391 F.3d at 537 (3d Cir.2004) (quoting General Motors, 55 F.3d at 817). The McCoy and Wachtel classes were certified by the Court in 2004. Health Net appealed and the Third Circuit vacated and remanded to the Court “for a definition of the claims, issues, or defenses to be treated on a class basis.” Wachtel, 453 F.3d at 190. The Court issued a new class certification order on September 25, 2006, Wachtel DKT# 551, noting that the Third Circuit “made clear that this Court’s certification order ‘easily meets the requirements of Rule 23(c)(1)(B) with respect to the definition of the class itself.’ ” See DKT# 551 at 2 (quoting Wachtel, 453 F.3d at 188). Health Net again appealed and the Third Circuit denied Health Net’s second Rule 23(f) appeal. As a result, it is unlikely that Plaintiffs face a substantial risk of decertification of the Wachtel or McCoy classes. Because the Scharfman ERISA class is materially the same as the McCoy ERISA class, it is equally unlikely that the Scharfman ERISA class would face decertification. The Scharfman RICO class brings claims that have not yet been tested in a motion to dismiss, however. On balance, the Court finds that the risk of maintaining a class action neither favors nor disfavors the Settlement. f. The Ability of Defendants to Withstand a Greater Judgment The Court cannot say with certainty to what extent Health Net could afford a greater settlement amount. Even assuming that Health Net could, however, that is no basis for rejecting the instant Settlement. As this Court has noted, “many settlements have been approved where a settling defendant has had the ability to pay greater amounts.” In re Remeron Direct, 2005 WL 3008808, at *9 (citing In re Warfarin Sodium, 391 F.3d at 538 (“[T]he fact that DuPont could afford to pay more does not mean that it is obligated to pay any more than what the ... class members are entitled to under the theories of liability that existed at the time the settlement was reached.”)); Oh v. AT & T Corp., 225 F.R.D. 142, 150-51 (D.N.J.2004); In re Linerboard Antitrust Litig., 321 F.Supp.2d 619, 632 (E.D.Pa.2004); Erie County Retirees Assoc. v. County of Erie, 192 F.Supp.2d 369, 376 (W.D.Pa.2002); Lazy Oil Co. v. Witco Corp., 95 F.Supp.2d 290, 318 (W.D.Pa.1997). This factor neither favors nor disfavors the Settlement. g. The Range of Reasonableness of the Settlement Fund in Light of the Best Recovery and all the Attendant Risks of Litigation “An assessment of the reasonableness [in light of the best recovery] of a proposed settlement seeking monetary relief requires analysis of the present value of the damages a plaintiff would likely recover if successful, appropriately discounted for the risk of not prevailing.” In re Remeron Direct, 2005 WL 3008808, at *9 (citing In re Prudential, 148 F.3d at 322). When analyzing the recovery in light of the attendant risks of litigation, “the Court [must] examine the terms of settlement from a ‘slightly different vantage point[]’ than reasonableness in light of the best recovery.” In re Remeron Direct, 2005 WL 3008808, at *10 (quoting In re General Motors, 55 F.3d at 806). The Court must consider “whether the settlement represents a good value for a weak case or a poor value for a strong case.” In re Remeron End-Payor, 2005 WL 2230314, at *23 (quoting In re Warfarin Sodium, 391 F.3d at 538). In these cases, the parties took differing approaches to determining total potential damages. Defendants maintained that only those subscribers who could show that they had been billed by their ONET provider were entitled to recover. On this definition, Defendants believed damages would amount to less than $100 million. Plaintiffs, by contrast, believed that any damage assessment must encompass individuals whose ONET claims were subject to Health Net’s allegedly improper methodology regardless of whether they were Balance Billed by their ONET provider. On this assessment, Plaintiffs assessed damages at approximately $400 million. The negotiated recovery is reasonable both in light of the best possible recovery and in light of the attendant risks of litigation. The parties estimate the settlement’s total value between $249 and $261 million. That total amount includes a $175 million cash settlement fund and a $40 million prove up fund, plus important business practice reforms in the use of UCR as a method for reimbursing ONET claims. Moreover, the Court has reiterated its view that a reasonable settlement must address how UCR is calculated. As noted above, the parties value the business practice changes at between $26-$38 million. (1) The Importance and Value of the Business Practice Changes in the use of UCR to Calculate Reimbursement to Insureds Because of the importance of these business practices changes in evaluating the fairness of this Settlement, the Court held a “tutorial” on the Ingenix database on April 10, 2008, prior to the preliminary fairness hearing. The Court heard expert testimony from Dr. Bernard R. Siskin, a statistics expert. The Court heard testimony about the methods used by Ingenix to create its commercial databases, and the statistical flaws in those methods. The testimony shed significant light on the composition and use of the Ingenix databases. Any “fair, reasonable, and adequate” settlement must address the use of data that suffers from major flaws in its collection and use as a proxy for UCR. Although this Settlement addresses the In-genix database, the use of any database to calculate medical insurance coverage can be analyzed by inquiring into the same data collection and statistical manipulation issues addressed herein. Health Net states in its Explanation of Coverage forms and its Summary Plan Descriptions that out-of-network charges will be calculated based on UCR charge for a particular service. Health Net determines the UCR charge for a given procedure by consulting two Ingenix databases: the Prevailing Healthcare Charge System database (“PHCS database”) and the MDR/Medicode database. The question addressed during the Court’s April 10th hearing was whether these databases satisfied the two “core concepts” of UCR. That is, whether the Ingenix databases provided accurate data as to the reasonable charge for a particular service, and the geographical area where the service was performed. To assess a reasonable charge for a particular medical service, it is essential to know the actual charges billed by similar providers for reasonably similar services in a relevant geographic area. In order to determine the set of reasonably similar services, the database would need to contain information on those factors which would affect the cost of the services, such as: (I) significant differences in provider qualifications, (ii) significant differences in type of medical service provided, and (iii) significant differences in medical market area. Given this information, a database analysis could then determine which charges are reasonable and which are “too high.” The Ingenix database suffers from numerous errors that must be addressed if a settlement is to be deemed “fair, reasonable, and adequate” under Rule 23(e)(2). The Court will discuss these errors in three broad categories: (a) data collection/sampling errors; (b) database creation/editing errors; (c) data analysis errors. (a) Data Collection There are two serious flaws in Ingenix’s data collection methods: one relates to Ingenix’s data sources; the other relates to the number of data points collected for each medical procedure. The database is compiled from data submitted by several insurers pursuant to a purely voluntary “data contribution program.” See Davekos, P.C. v. Liberty Mut. Ins. Co., No. 10002, 2008 WL 241613, at *2 (Mass.App.Div. Jan. 24, 2008). Under this program, “some, but not all, of only those health insurers that are Ingenix clients submit information, on a purely voluntary basis, about the amounts they happen to have been billed by an undisclosed number of unidentified health care providers for specific CPT code services.” Id. As Judge Curtin succinctly stated in Davekos, “[a]t best, the Ingenix database includes the bills of an unspecified number of medical providers who, within a specific period of time, happened to have billed only those health insurers that were not only Ingenix clients, but also Ingenix clients that elected to participate in [Ingenix’s] voluntary data contribution program.” Id. at *4. This method of data collection is considered by statisticians to be a “convenience sample.” A convenience sample is the easiest way to collect the data, but it is haphazard. Convenience samples are chosen on the basis of expediency, cost, efficiency or other reasons not directly concerned with scientific sampling parameters. As a result, convenience samples are considered the “most suspect type of sample.” April 10, 2008 Ingenix Hearing Transcript (“Tr.”) at 16(22). A convenience sample is not necessarily invalid, but it must be subject to further testing to determine whether the data collected is in fact representative of what an insurer is trying to estimate. Ingenix does not test the voluntarily submitted data to see if the data constitutes an accurate representative sample of charges for a particular procedure in a particular geographical area. Moreover, the data collection methodology provides no reassurance that the raw data collected is representative of the actual charges billed for any given procedure. Companies that submit data receive a discount based on the amount of usable data submitted. This arrangement can encourage insurers to remove high charges before submitting their data, in order to ensure “that a lot of it’s not going to be knocked out” during the “data scrubbing process.” See Tr. at 22(3)-(4); see generally Tr. at 20(19) — 21(1), 21(25)-22(7). Because other insurance companies who themselves use the database are permitted to choose what data to submit, there is a built in incentive to submit low cost data that will produce a lower UCR database that the submitting insurance company will itself use to calculate a lower UCR for its own reimbursements to its insureds. Second, the database relies upon too few data points for each procedure. The database relies upon just four pieces of data for each submitted charge: date of service; 5-digit Current Procedural Terminology code (“CPT code”); the address where the procedure was performed; and the amount of the provider’s billed charge. Ingenix relies upon these four data points to facilitate comparison among similar procedures and geographical zones. In other words, these data points represent the sum total of the information that purportedly allows an insurer to compare similarly situated procedures. See Tr. at 28(2). These four data points exclude several factors that are critical to the “core concepts” of UCR. These four data points do not identify: (1) the provider’s licensure or qualifications; (2) the patient’s age or health status; (3) the type of facility where the procedure was performed. The database does not take into account whether a particular procedure was performed by a highly-skilled Board Certified specialist or a general practitioner or a paraprofessional or a nurse. It is a matter of common sense that these factors may be fundamental to a comparison of charges. A procedure performed by a highly skilled physician is likely to be more expensive than one performed by a physician’s assistant or nurse practitioner, but the physician’s higher charge may nevertheless be the most valid comparator if an insured was treated by a physician of comparable skill and experience. Yet, by including every possible type of provider in the CPT Code Service, even a totally average bill from a skilled physician will be higher than the UCR yielded by the database. These excluded data points may be the most important factors in determining “reasonable” and “customary” costs. One might expect that it would cost significantly more to have a highly skilled, Board Certified heart specialist interpret an echocardiogram than it would to have a general practitioner do the same task. The database improperly assumes that these factors are all irrelevant for determining the usual and customary rate charged for particular procedures. Any accurate database would control for these additional factors. Ingenix’s failure to control for these factors means that the database is not actually comparing similarly situated procedures when it purportedly yields a “usual” and “customary” rate for that procedure. (b) Database creation/editing The method Ingenix uses to create a database from the collected data also undermines the “core concepts” of UCR. After Ingenix collects the data, the data is “scrubbed” to remove certain charges. This process is not per se improper; to ensure an accurate database, it is necessary to review the data and remove erroneous or incorrectly reported charges. Specifically, erroneous “outlier values” that are either too high or too low will skew the data and can be removed if done by a statistically fair and reliable method. The key flaw is the method by which the database determines which values are “outliers.” This database uses a “mean to median” test to scrub its data. By this process, this database eliminates all bills for a given CPT code if the mean to median ratio within that CPT code is above 2.5 for surgical CPT codes and' 1.5 for medical CPT codes. The more high fees contained in the data set, the more likely the mean to median ratio will remove these fees. For example, if a data provider submits 500 fee charges for a CPT code for an office visit for a new patient, there may be 200 $10 charges, 100 $30 charges, and 200 $100 charges. The disparity in pricing may accurately reflect the length of a patient’s visit — for example, a visit concerning a serious illness may take longer than one concerning a minor illness — or the skill of the provider — a visit with a physician’s assistant would likely cost less than one with a physician. In this data set, however, the mean is $50 and the median is $30. The mean to median ratio for this medical CPT code therefore exceeds 1.5 and as a result, Ingenix would delete the entire data set submitted by that data provider. This method of scrubbing would bias the resulting distribution downwards. The databases’s “data scrubbing” method is also considered by statisticians to be one of convenience. The database does not review outlier values to determine whether high values are accurate. Rather, the database simply removes all high fees without any evidence that those values represent data errors. For example, the high charge resulting from a skilled surgeon performing a difficult operation at an excellent medical facility may appear as a statistical “outlier” when compared to all charges for procedures in the same CPT code, without regard to the identity of the provider, procedure difficulty, or location of service. Under Ingenix’s method, this outlier data would not be reviewed individually to consider whether it was valid. It would simply be thrown out as too high, which would skew the data downwards. This method is compounded by the fact that, as discussed above, many of the data providers themselves “pre-scrub” the data they submit to Ingenix in the first place. As a result of this “pre-scrubbing,” before the data editing process even begins, the data contribution method provides an incentive for a downward bias. Ingenix applies this scrubbing process to groups of CPT codes in a way that further skews UCR rates downward. Each individual CPT code represents a different procedure and, as a result, the mean charge for each individual CPT code may differ. Ingenix attempts to account for the differences between CPT codes by giving a relative value for each CPT code. For example, if, in a group of two codes, one CPT code has a mean of $50 and a second code has a mean of $100, the first code will receive a relative value of 1 and the second a relative value of 2, reflecting the fact that the second code has a relative mean value twice the first code. The CPT code values are standardized (by applying the relative values) and combined into groups. The grouped values are then subjected to formulas to eliminate outlier data at the high and low end. “Standardizing” charges in this way without also taking into account the standard deviation within each CPT code will improperly skew the UCR rate downward because the method fails to account for differing distributions of charges within each CPT code. Charges within some CPT codes — those for routine procedures, for example — may be tightly grouped. Other CPT codes may have wide distributions, reflecting differences in the providers’ skills, patients’ conditions, or facilities. The methodology used to create the database rests on the assumption that the distribution of charges as to all CPT codes in the CPT code range is the same. The net result of this erroneous assumption is that high charges which are valid, usual, and customary are rejected as unreliable outliers, and are eliminated from the common data, thereby skewing downward the upper percentile values in the final reported data. (c) Data Analysis and Publication Data that survives the scrubbing process described above is then compiled by CPT code. If a CPT code has nine or more charges, those charges are reported as “actual data.” Actual data is reported for just 10% of all CPT codes. Ingenix “derives” data for the remaining 90% of CPT codes that have fewer than nine charges. The method for deriving data for these CPT codes involves combining CPT codes and standardizing the values so that the values can be compared across CPT codes. First, Ingenix groups together CPT codes into a bodily system. For example, Ingenix groups numerous CPT codes together as “upper digestive system.” This group of codes contains relatively simple procedures — such a lip repair (CPT 40490) — and extremely complex procedures — such as an esophagecto-my (CPT 43116). To standardize these disparate charges, Ingenix uses relative values provided by a company called Relative Value Studies, Inc. This process is similar to that used at the “scrubbing” stage, described above, but using different relative values. And this process suffers from the same fundamental flaw as the scrubbing process: the process assumes that each CPT code has the same distribution of values. Any accurate standardization method must account for differences in both relative values and relative standard deviations between CPT codes. Ingenix’s failure to account for standard deviation when it derives data means that almost any charge above the mean in the less common CPT codes with a higher relative