E.S. v. Regence BlueShield

Opinion

HONORABLE RICHARD A. JONES

WESTERN DISTRICT OF WASHINGTON

E.S., et al., Plaintiffs, Case No. 2:17-cv-01609-RAJ v. ORDER Regence BlueShield, et al., Defendants. I. INTRODUCTION This matter comes before the Court on Defendants’ Motion to Dismiss Amended Complaint. Dkt. # 37. Having considered the submissions of the parties, the relevant portions of the record, and the applicable law, the Court finds that oral argument is unnecessary. For the reasons below, the motion is GRANTED. II. BACKGROUND Plaintiffs E.S. and Jodi Sternoff (“Plaintiffs”) are both diagnosed with hearing loss. Dkt. # 32 ¶¶ 6, 7, 95. Both are insured under health insurance policies issued by Defendants Regence BlueShield and Cambia Health Solutions (collectively, “Regence”). Id. ¶ 3. Both were denied coverage for their hearing loss treatments. Id. ¶¶ 24, 86, 88, 90, 92. They were denied because Regence’s policy has an exclusion for certain types of hearing loss treatment, and Plaintiffs’ treatments fell under that exclusion. Id. On October 30, 2017, Plaintiffs sued Regence for discrimination. Dkt. # 1. After amending their complaint, they now allege that Regence violated the Affordable Care Act by designing the exclusion to exclude coverage for “insureds with disabling hearing loss,” a “form of intentional proxy discrimination. Dkt. # 32 ¶¶ 104, 121. Plaintiffs also assert a claim for breach of contract and violation of RCW 48.43.0128. Id. ¶¶ 125-29. Regence now moves to dismiss the amended complaint. Dkt. # 37. Plaintiffs oppose the motion to dismiss. Dkt. # 38. The matter is ripe and before the Court. Under Rule 12(b)(6) of the Federal Rules of Civil Procedure, a court may dismiss a complaint for failure to state a claim. The court must assume the truth of the complaint’s factual allegations and credit all reasonable inferences arising from those allegations. Sanders v. Brown, 504 F.3d 903, 910 (9th Cir. 2007). The court “need not accept as true conclusory allegations that are contradicted by documents referred to in the complaint.” Manzarek v. St. Paul Fire & Marine Ins. Co., 519 F.3d 1025, 1031 (9th Cir. 2008). Instead, the plaintiff must point to factual allegations that “state a claim to relief that is plausible on its face.” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 568 (2007). If the plaintiff succeeds, the complaint avoids dismissal if there is “any set of facts consistent with the allegations in the complaint” that would entitle the plaintiff to relief. Id. at 563; Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009). On a motion to dismiss, a court typically considers only the contents of the complaint. However, a court is permitted to take judicial notice of facts that are incorporated by reference in the complaint. United States v. Ritchie, 342 F.3d 903, 908 (9th Cir. 2003) (“A court may . . . consider certain materials documents attached to the complaint, documents incorporated by reference in the complaint.”); Mir v. Little Co. of Mary Hosp., 844 F.2d 646, 649 (9th Cir. 1988) (“[I]t is proper for the district court to ‘take judicial notice of matters of public record outside the pleadings’ and consider them for purposes of the motion to dismiss.”) (quoting MGIC Indem. Corp. v. Weisman, 803 F.2d 500, 504 (9th Cir. 1986)). Regence moves to dismiss both of Plaintiffs’ claims. Dkt. # 37. It moves to dismiss Plaintiffs’ first claim, a claim for discriminatory plan benefit design under Section 1557 of the Patient Protection and Affordable Care Act (“ACA”), 42 U.S.C. § 18116. Regence argues that Plaintiffs have failed to plead a “proxy discrimination” theory. Regence also moves to dismiss Plaintiffs’ second claim for “breach of contract and violation of RCW 48.43.0128.” The Court addresses each claim in turn. A. Violation of Section 1557 of the ACA (Count 1) This case is much like another pending in this district. In Schmitt v. Kaiser Foundation Health Plan, another insurer, Kaiser, is being sued for substantially the same reasons Regence is being sued here. There, plaintiffs are suing Kaiser because its policies, like Regence’s, “exclude coverage for all programs or treatments for hearing loss or hearing care with the exception of cochlear implants” and thus discriminate on the basis of hearing loss disability. Schmitt v. Kaiser Found. Health Plan of Washington, No. 2:17-cv-01611-RSL, 2018 WL 4385858, at *1 (W.D. Wash. Sept. 14, 2018), aff’d in part, rev’d in part and remanded, 965 F.3d 945 (9th Cir. 2020). The Schmitt plaintiffs, like Plaintiffs here, bring their claims under Section 1557 of the ACA. Id. Schmitt is significant because, after the plaintiffs there appealed the district court’s dismissal order, the Ninth Circuit recognized—for the first time—that insureds may state a claim for “discriminatory benefit design” directly under section 1557. Schmitt, 965 F.3d at 955. The Ninth Circuit also identified a theory that plaintiffs could use to plead such a claim, a “proxy discrimination” theory. Id. at 958-60. That is exactly the claim and theory that Plaintiffs assert here in their amended complaint. Dkt. # 32 ¶¶ 113-24. Given the novelty of Schmitt and its importance to Plaintiffs’ claim, the Court summarizes the decision and then determines whether Plaintiffs have adequately stated a discriminatory benefit design claim. i. Schmitt v. Kaiser Foundation Health Plan Section 1557 of the ACA prohibits certain types of discrimination in health care, including disability discrimination. Schmitt, 965 at 950. It does so by incorporating other nondiscrimination statutes and prohibiting discrimination “on those grounds in the health care system,” and specifically in “health insurance contracts.” Id. at 951. One such nondiscrimination statute is Section 504 of the Rehabilitation Act of 1973, which prohibits disability discrimination in general. Id. at 950-51. Generally, “the case law construing the Rehabilitation Act . . . applies to claims under section 1557 for disability discrimination by a health care insurer.” Id. at 954. Put differently, section 1557 disability discrimination claims under the ACA may often follow the same rubric as section 504 claims under the Rehabilitation Act. See id. But in Schmitt the Ninth Circuit recognized—for the first time—that section 1557 permits plaintiffs to sue covered health insurers for “discrimination in the design of plan benefits” independent from section 504. Id. at 954-55. Despite the similarities between the ACA and the Rehabilitation Act, the Schmitt court explained that they differ in key ways. Id. Most significant is how the statutes treat a claim for “discriminatory plan benefit design.” Id. Historically, the Rehabilitation Act “d[id] not cover discriminatory plan benefit design.” Id. at 955. That much was clear given the Supreme Court’s decision in Alexander v. Choate, 469 U.S. 287 (1985). Id. But, the Ninth Circuit said, the same is not true for the ACA, which in fact provides for such a claim. Id. In holding that the Rehabilitation Act does not cover discriminatory plan benefit design, the Supreme Court rejected a group of Medicaid recipients’ attempt to define the benefit at issue as “the amorphous objective of ‘adequate health care.’” Choate, 469 U.S. at 303, 105 S.Ct. 712. . . . [The Supreme Court explained that] [t]he Rehabilitation Act does not impose a general requirement on “each recipient of federal funds first to evaluate the effect on [disabled people] of every proposed action that might touch [their] interests ..., and then to consider alternatives for achieving the same objectives with less severe disadvantage to [them].” Id. at 298, 307, 105 S.Ct. 712. The ACA, in contrast, does almost all of this. While it does not guarantee individually tailored health care plans, it attempts to provide adequate health care to as many individuals as possible by requiring insurers to provide essential health benefits. And it imposes an affirmative obligation not to discriminate in the provision of health care—in particular, to consider the needs of disabled people and not design plan benefits in ways that discriminate against them. Id. (emphasis added). In the end, the Ninth Circuit held that “the ACA allows a claim for discriminatory benefit design notwithstanding that, under Choate, the Rehabilitation Act does not.” Id. (emphasis added).1 The origin of the claim explained, the Court now determines whether Plaintiffs have successfully pled such a claim. ii. Discriminatory Design of Plan Benefits The ACA “imposes an affirmative obligation not to discriminate in the provision of health care.” Schmitt, 965 at 955. Health care insurers must not “design plan benefits in ways that discriminate against [disabled people].” Id. Under Section 1557 of the ACA, plaintiffs may plead a discriminatory design benefits claim and may do so using a “proxy discrimination” theory. See id. at 958-60. (1) Proxy Discrimination Proxy discrimination is “when the defendant enacts a law or policy that treats individuals differently on the basis of seemingly neutral criteria that are so closely associated with the disfavored group that discrimination on the basis of such criteria is, constructively, facial discrimination against the disfavored group.” Id. at 958 (quoting Davis v. Guam, 932 F.3d 822, 837 (9th Cir. 2019)); see also Pac. Shores Props., LLC v. City of Newport Beach, 730 F.3d 1142, 1160 n.23 (9th Cir. 2013). Take for example a 1 Throughout their response, Plaintiffs cite cases analyzing ACA claims brought under a “meaningful access” theory of liability. Dkt. # 38 at 16-17, 24. But Plaintiffs did not plead a meaningful access theory; they pled a proxy discrimination theory. Dkt. # 32 ¶¶ 113-124. Thus, the Court agrees with Regence that such cases do not apply. Dkt. # 39 at 9. policy discriminating against an “individual with gray hair.” Schmitt, 965 F.3d at 958. Gray hair, in that case, would be a proxy for age because the “fit” between age and gray hair is “sufficiently close.” Id. (quoting Davis, 932 F.3d at 837-38). The “crucial question” in a proxy discrimination analysis is whether a proxy’s “fit” to a protected group is “sufficiently close” to make a discriminatory inference plausible. Id. Given Schmitt’s novelty, authority on what constitutes adequate “fit” is slim. Still, the Court has several guiding principles. If an “overinclusive” proxy “primarily” or “predominately” affects disabled people, then that would support an inference of discrimination. Schmitt, 965 F.3d at 959 n.8. On the other hand, if a proxy is “underinclusive,” because a policy for example can meet the needs of similarly disabled people, then that would undermine an inference of discrimination. Id. at 959. Courts may also consider “historical facts” about a policy’s enactment. See Davis, 932 F.3d at 838-41. For example, in Davis, the Ninth Circuit reviewed a Guam law restricting voting to “Native Inhabitants of Guam.” Id. at 824. The Davis court held that “Native Inhabitants of Guam,” a classification based on ancestry, was in fact a racial proxy for “Chamorro,” and therefore the Guam law violated the Fifteenth Amendment. Id. at 839, 843. In concluding that “Native Inhabitants of Guam” was a racial proxy, the court relied on several “historical facts” and “terms of the classification.” Id. at 838. For example, it determined that the Guam law intended to discriminate on racial grounds by looking to the law’s “immediate predecessors.” Id. at 839. Those preceding laws “were not shy about using an express racial classification” for they explicitly relied on the racial classification of Chamorro. Id. at 839-43 (“That Guam applies nearly identical definitions to the terms ‘Chamorro,’ a racial category, and ‘Native Inhabitants of Guam’ indicates that these terms are interchangeable.”); see also Pac. Shores Props., 730 F.3d at 1162, 1164 (relying on “[a]ll of the circumstances surrounding the enactment of the [o]rdinance” to determine the ordinance’s “discriminatory purpose” in harming a disabled group). What is more, a court may also find helpful allegations of “specifically targeted . . . enforcement.” Pac. Shores Props., 730 F.3d at 1162. In Pacific Shores, the Ninth Circuit, reviewing a spate of city zoning policies intended to eliminate “group homes” for the disabled, held that there was enough evidence to raise an issue of fact regarding the policies’ discriminatory intent. Id. at 1164. In reaching that conclusion, the Pacific Shores court noted how the city specifically targeted the disabled. Id. at 1162. For example, it explained that the city created a task force that “locate[d] group homes,” “surveil[ed] . . . them,” and “enforce[d] the zoning code strictly against them.” Id. Meanwhile, the city did not enforce the zoning policies against other facilities—such as “dormitories,” “sororities,” and “private residential clubs”—even though those facilities also came under the policies’ ambit. Id. at 1164-65. Taken together, then, to determine whether a proxy’s “fit” is “sufficiently close” to state a claim for discriminatory plan design, the Court may look to a given policy’s disproportionate effect on disabled insureds (overinclusion), ability to service the needs of similar disabled insureds (underinclusion), historical enactment, or targeted enforcement. The Court begins its analysis by identifying the “proxy” in this case. Then, using the principles above, the Court answers the “crucial question”—whether the proxy here is “sufficiently close” to a protected group to make a discriminatory inference plausible. (2) Proxy – Excluded Insureds with Disabling Hearing Loss Explained more below, Plaintiffs’ proxy discrimination theory relies on several premises: Hearing loss ranges in severity. The most severe forms of hearing loss are treated with cochlear implants. Regence’s policy covers only cochlear implant treatment and excludes all other forms of hearing loss treatment. As a result, the policy creates a category of excluded insureds. That category includes some disabled insureds—those whose hearing loss is severe enough to require treatment and to be considered a “disability” under antidiscrimination statutes, yet not severe enough to require cochlear implants. The category also includes non-disabled insureds. That would be both those who suffer hearing loss but not to a debilitating degree and those who do not suffer any hearing loss but who desire routine hearing examinations. According to Plaintiffs, this category of excluded insureds is a proxy. It is a proxy to discriminate against excluded insureds with “disabling hearing loss.” Those insureds are disabled but need treatment other than cochlear implants, such as hearing aids and outpatient audiologist visits. A proxy exists because “non-disabled insureds,” though they too are excluded, rarely seek treatment for hearing loss, and thus those who are primarily affected by the exclusion are disabled insureds. Hence, the “fit” between the proxy and the protected group is “sufficiently close” to make a discriminatory inference plausible. (a) Degrees of Hearing Loss The severity of hearing loss ranges. Dkt. # 32 ¶ 46. On one end, there is no hearing loss. Id. ¶ 25. Generally, the standard for “normal hearing” is a threshold of 25 decibels. Id. ¶ 44. That means if a tone must be louder than 25 decibels for it to be audible to a listener, then the listener’s hearing is “worse-than-normal.” Id. Contrast the other end. If it takes more than 90 decibels for a tone to be audible to a listener, then the listener has “profound” hearing loss. Id. ¶ 46. Those who suffer profound hearing loss “will hear almost nothing.” Id. Between no hearing loss and profound hearing loss are three categories: “mild,” “moderate,” and “severe.” Id. Arranged from least to most severe, the five categories of hearing loss would thus be none, mild, moderate, severe, and profound. Id. Those five categories are descriptive; they are not legal. Two legal categories are also relevant here: disabled and non-disabled. Under the Americans with Disabilities Act (“ADA”), a “disability” is in part a “a physical or mental impairment that substantially limits one or more major life activities,” including “hearing.” 42 U.S.C. § 12102. Along the hearing loss spectrum, some people with mild hearing loss would likely not be “disabled” under the ADA, while some people with profound hearing loss would almost certainly be.2 See Dkt. # 32 ¶¶ 4, 24, 34, 40, 46, 54. These two legal categories are significant because only those with a disability are protected by the ACA and other antidiscrimination laws. (b) Policy Exclusion Plaintiffs’ health insurance policy with Regence has several exclusions. See, e.g., Dkt. # 32-1 at 50-53. Among them are exclusions for “Hearing Aids, and Other Hearing Devices” and for “Routine Hearing Examination[s]” (together, the “Exclusion”). Id. Under the Exclusion, “[h]earing aids (externally worn or surgically implanted), and other hearing devices are excluded” from the insurance policy. Id. at 51. So too are routine hearing examinations. Id. at 52. Cochlear implants, however, are not excluded. Id. at 51. (c) Proxy Discrimination Theory According to Plaintiffs, the Exclusion creates a category of excluded insureds, and that category is a proxy to discriminate against disabled insureds. The theory goes: Cochlear implant treatment addresses the needs of “only a very small fraction of the total population of people with hearing loss.” Dkt. # 32 ¶ 72. Such treatment is reserved only for those with “severe to profound hearing loss” who “cannot be adequately treated with hearing aids.” Id. ¶ 68. Plaintiffs estimate that only about 5.6% of “people under 65 with self-reported hearing loss[]” would be potentially eligible for cochlear implant treatment. Id. ¶ 69. All other people with hearing loss, Plaintiffs allege, are excluded and underserved

2 Because the Court is testing the legal sufficiency of this claim, it assumes that this is true. Throughout their complaint, Plaintiffs implicitly concede that many people with hearing loss would not be “disabled” under the ADA. See Dkt. # 32 ¶¶ 46, 56, 58, 106. The Schmitt court made the same assumption. 965 F.3d at 958 (“But since not all hearing loss is substantial, at least some—and potentially most—individuals with that condition are not deemed disabled.”). by Regence’s policy. Id. ¶¶ 4, 72. For example, the policy does not cover less severe hearing loss treatment, such as “outpatient office visits” and “durable medical equipment or prosthetic devices” like hearing aids. Id. ¶¶ 14, 24. Take Plaintiffs here. Both suffer from disabling hearing loss and cannot receive coverage for treatment that they need. Id. ¶ 24. E.S. was born without an outer ear or properly formed middle ear. Id. ¶ 74. Regence denied her coverage for a bone-anchored hearing aid, or BAHA, a device that “bypasses the damaged middle-ear structures and transmits sound directly to the cochlea and hair cells.” Id. ¶ 49. Because a BAHA is different than a cochlear implant, Regence denied E.S. coverage for the device. Id. ¶¶ 49, 74-93. Ms. Sternoff, on the other hand, has moderate to severe hearing loss in her left hear. Id. ¶ 91. To treat her condition, she uses a “cross hearing aid” to transmit sound from her left side to her hearing right ear. Id. She too was denied coverage for her hearing aid under the Exclusion. Id. ¶ 92. Disabled insureds aside, the Exclusion also excludes non-disabled insureds. Some insureds may have hearing loss, but their hearing loss may not be severe enough to qualify them as “disabled” under the ADA. See supra note 2. Further, some insureds may have no hearing loss at all, and thus also not “disabled” under the ADA, but still desire routine hearing examinations. Like their disabled counterparts, these non-disabled insureds have no coverage under Regence’s policy. Dkt. # 32-1 at 50-53. Plaintiffs’ discrimination theory is that, because disabled insureds are affected most by the Exclusion, this is a form of “proxy discrimination.” Dkt. # 32 ¶¶ 104, 118, 120. According to Plaintiffs, the Exclusion creates a category of excluded insureds, which is but a proxy to discriminate against disabled insureds. See id. Plaintiffs concede that the proxy contains both disabled and non-disabled insureds alike. See supra note 2. Yet, because “[n]on-disabled insureds rarely seek treatment for hearing loss,” they are not affected by the Exclusion as much as disabled insureds are. See Dkt. # 32 ¶¶ 56, 60, 106, 118. In effect, then, the Exclusion primarily affects disabled insureds and prevents them from getting the non-cochlear-implant treatment—such as hearing aids and audiologist visits—that they need. Id. ¶¶ 24, 108. The proxy now defined, the Court turns to the crucial question in this case: whether the proxy’s fit is sufficiently close to disabled hearing loss to make a discriminatory inference plausible. See supra Section IV.A.ii.1. The Court analyzes the Exclusion’s historical enactment and targeted enforcement, its disproportionate effect on disabled persons (overinclusion), and its ability to service the needs of similar disabled insureds (underinclusion). (3) Enactment and Targeted Enforcement To start, Plaintiffs assert no historical facts about the Exclusion’s enactment.3 They do not allege, for example, any facts about how Regence, in previous versions of its policies, explicitly discriminated against insureds with disabling hearing loss. Nor do they allege any facts about the policy drafter’s intent. For their part, Plaintiffs allege that “Regence designed the Exclusion intentionally to deny services to insureds with disabling hearing loss.” Dkt. # 32 ¶¶ 61-63. Yet this allegation of intent, not grounded in fact, is conclusory, and the Court disregards it. What is more, Plaintiffs also allege no facts about Regence’s targeted enforcement. Nowhere in the complaint is an allegation that Regence only enforces this Exclusion against disabled insureds and not non-disabled insureds. For example, Plaintiffs do not allege that Regence waives this exclusion for those with non-disabling hearing loss or those merely seeking routine hearing examinations, while simultaneously enforcing it against disabled insureds seeking hearing aids. In all, neither of these indicia of “fit”—circumstances surrounding the enactment of the policy or selective enforcement of the Exclusion—raise a plausible inference of

3 The proxy discrimination cases addressed legislation, not insurance policies. See supra Section IV.A.ii1. But the same principles apply—relevant to a proxy discrimination fit analysis is how a policy (whether one of law or contract) came to be. discrimination. (4) Overinclusion and Underinclusion To support their fit analysis, Plaintiffs offer two allegations. First, they reason that non-disabled insureds do not seek coverage under Regence’s policy to begin with. Second, they provide a statistical breakdown of the hearing population to show that few disabled insureds are covered by the policy, while many disabled insureds are excluded. The Court takes each in turn. First, Plaintiffs allege that the proxy is overinclusive because many non-disabled insureds do not seek coverage at all, and therefore they are not denied coverage as often as disabled insureds. Plaintiffs posit that non-disabled insureds with hearing loss “do not generally seek formal treatment from medical professionals, and rarely, if ever, seek hearing instruments.” Dkt. # 32 ¶ 55. And, even if non-disabled insureds did indeed seek treatment, they would be denied coverage because such treatment would not be a “medical necessity” and Regence’s policy excludes treatments that are not medically necessary. Id. ¶¶ 54-58. These allegations are conclusory, and the Court need not take them as true. They are assertions devoid of “underlying facts.” Eclectic Properties E., LLC v. Marcus & Millichap Co., 751 F.3d 990, 996 (9th Cir. 2014); Blantz v. California Dep’t of Corr. & Rehab., Div. of Corr. Health Care Servs., 727 F.3d 917, 927 (9th Cir. 2013) (holding that an allegation that a defendant “directed” other defendants to take actions was “conclusory” and “insufficient” to defeat a motion to dismiss because “no factual assertions support[ed] th[e] allegation”). Here, Plaintiffs do not allege any facts to support their allegation that non-disabled insureds rarely, if ever, seek treatment. Dkt. # 32 ¶ 55. Also unsupported is their allegation that any such treatment would be unnecessary and “only for the convenience of the patient.” Id. ¶ 57. These conclusory allegations “require[] further explanation to be plausible.” Schmitt, 965 F.3d at 959 (finding insufficient the allegation that “few, if any, non-disabled insureds had claims denied under the Hearing Loss Exclusion”). Second, Plaintiffs offer several hearing loss statistics to show that the proxy here is overinclusive. Many of their statistics come from different sources. Dkt. # 32 ¶¶ 44, 47, 51, 69, 70. Some come from the U.S. Census Bureau. Id. ¶ 47 (estimating that “just under 9.2 million Americans under age 65” self-reported as having “serious” difficulty hearing). Some come from medical journals. Id. ¶ 40 (one study estimating that “48 million Americans age 12 and over have impairing hearing loss in at least one ear”), ¶ 69 (another study estimating the “national prevalence of hearing loss by severity”). Extrapolating across these various statistics, Plaintiffs make their own deductions about the hearing loss population. For example, relying on two different data sets—a U.S. Census Bureau survey of people with self-reported “serious” hearing difficulty and a medical journal study of people with “severe” or “profound” hearing loss—Plaintiffs conclude that “just 5.6% of the 9.2 million people under 65 with self-reported hearing loss[]” would be eligible for cochlear implants.4 Id. ¶ 69. Like Plaintiffs’ conclusory allegations, the Court also need not take these statistical deductions as true.5 They are conclusions based on different data sets and hence different data populations. They shed no light on “fit”: they do not identify the total hearing loss population, the proportion of that population that is disabled and eligible for cochlear implants, the proportion that is disabled but ineligible for cochlear implants, or the proportion that is not disabled. Their statistical deductions aside, Plaintiffs’ claim may still be viable. The complaint attaches and incorporates one of the medical journal studies discussed in the

4 Plaintiffs do not explain the difference, if any, between one study’s measurement of “serious” hearing difficulty and another study’s measurement of “severe” or “profound” hearing loss. 5 To be clear, on a motion to dismiss, the Court does not dispute the veracity of the underlying statistics. It takes those underlying statistics as true. Yet the Court need not accept Plaintiffs’ deductions from those statistics. body of the complaint. Dkt. # 32 ¶ 69; Dkt. # 32-2. That study does, in fact, have the data required for a fit analysis. Dkt. # 32-2. It “estimate[s] the number of people in the United States who have a hearing impairment by severity and age.” Id. at 1. Thus, it contains a single population and a breakdown of that population by hearing loss severity. The Court relies on this study, incorporated in the complaint, instead of Plaintiffs’ statistical deductions.6 Daniels-Hall v. Nat’l Educ. Ass’n, 629 F.3d 992, 998 (9th Cir. 2010) (“We are not . . . required to accept as true allegations that contradict exhibits attached to the Complaint . . . .”); Gonzalez v. Planned Parenthood of Los Angeles, 759 F.3d 1112, 1115 (9th Cir. 2014) (collecting cases). That study reveals the following statistics: In the United States, there are about 38.17 million people with “bilateral” hearing loss, that is hearing loss in both ears. Dkt. # 32-2 at 2. Of those 38.17 million, about 2.12 million (or 5.5%) have “profound” or “severe” hearing loss; about 10.66 million (or 27.9%) have “moderate” hearing loss; and about 25.39 million (or 66.5%) have “mild” hearing loss. Id. Though it is not clear which of these people with hearing loss would be “disabled” under the ADA, the Court makes several assumptions. The Court assumes that those with “severe” or “profound” hearing loss would be disabled and eligible for cochlear implant treatment. People in this category “hear almost no speech” or, in profound cases, “hear almost nothing.” Dkt. # 32 ¶ 46. And cochlear implant treatment is reserved for the “extreme cases” of hearing loss, which people in this category presumably have. Id. ¶¶ 48, 66. The Court further assumes that people in this category are disabled under the ADA, for their condition would substantially limit their major life activities. 42 U.S.C. § 12102. 6 By no means does the Court require, at the complaint stage, statistical allegations to plead a proxy discrimination claim. But, to the extent a plaintiff relies on statistics incorporated in the complaint, a court need not take the plaintiff’s deductions from those statistics as true on a motion to dismiss. Next, the Court assumes that every person with “moderate” hearing loss would be disabled, yet ineligible for cochlear implant treatment. People in this category “have difficulty understanding speech at normal levels.” Dkt. # 32 ¶ 46. Given that “hav[ing] difficulty understanding speech at normal levels” is vague, it is not clear that all 10.66 million people in this category would indeed qualify as disabled under the ADA because everyone’s hearing “difficulty” might not rise to a degree that substantially limits their major life activities. 42 U.S.C. § 12102. In any event, the Court assumes that all people with “moderate” hearing loss would be disabled under the ADA because the Court must construe all inferences in Plaintiffs’ favor. McAdory v. M.N.S. & Assocs., LLC, 952 F.3d 1089, 1092 (9th Cir. 2020), cert. denied sub nom. DNF Assocs., LLC v. McAdory, 141 S. Ct. 627 (U.S. 2020). Finally, the Court assumes that every person with “mild” hearing loss would not be disabled. People in this category may only have “some difficulty hearing softly voiced sounds.” Dkt. # 32 ¶ 46. Given these assumptions, the Court finds that Plaintiffs have failed to allege a “sufficiently close” fit between the proxy and disabled insureds. Based on the statistics incorporated in Plaintiffs’ complaint, the Exclusion does not “predominately affect disabled persons.” Schmitt, 965 F.3d at 959 n.8. It “predominately” or “primarily” affects non-disabled persons. Id. Of the entire hearing loss population, only about 27.9% would be disabled and excluded by Regence’s policy. Dkt. # 32-2 at 2. And, again, that is assuming that every person with “moderate” hearing loss would be disabled under the ADA. On the other hand, 66.5% of the hearing loss population—more than double—would not be disabled under the ADA and would also excluded by Regence’s policy. Id. Thus, in all, the Exclusion mostly affects non-disabled insureds. Indeed, this does not even consider those with no hearing loss. The study only surveyed those with hearing loss; it did not survey those with no hearing loss. Dkt. # 32- 2. Yet those without hearing loss would also be affected by the Exclusion given that many might seek preventative care such as routine hearing examinations, which are also excluded under Regence’s policy. Were the Court to consider this population as well, even more non-disabled insureds would be affected by the Exclusion. To be sure, the policy only covers a small minority of insureds. It presumably covers only those with “profound” or “severe” hearing loss, about 5.5% of the total population. Dkt. # 32-2 at 2. And, addressing underinclusion, the Schmitt court said that “[i]f cochlear implants serve the needs of most individuals with hearing disability, that fact would tend to undermine a claim of proxy discrimination.” 965 F.3d at 959. Here, because it appears that only a small proportion of the hearing loss population would be served by Regence’s policy, this would support a claim of proxy discrimination. Still, the Court finds that Plaintiffs have failed to state a claim. The proxy created by the Exclusion is wide. And only a relatively small proportion of that proxy is disabled. The non-disabled proportion of that proxy outnumbers the disabled proportion by at least 2-to-1. And that is ignoring the many others without hearing loss, which are also part of the proxy. Based on these statistics, the Court finds that Plaintiffs have failed to allege a “seemingly neutral criteria”—drawn “so closely” with a “disfavored group”— that is “constructively [] facial discrimination.” Schmitt, 965 F.3d at 958 (quoting Davis v. Guam, 932 F.3d 822, 837 (9th Cir. 2019)). Plaintiffs’ allegation of underinclusion slightly tips in their favor but not enough to tip this claim to the level of plausibility. Thus, the Court dismisses Plaintiffs’ section 1557 claim. (5) Leave to Amend To plead a proxy discrimination theory, insureds are not required to use statistics. That is simply the route Plaintiffs chose here. Nowhere in Schmitt did the Ninth Circuit say that the only way insureds could plead proxy discrimination would be by showing that disabled insureds numerically outnumber non-disabled insureds. In fact, the Schmitt court itself recognized the difficulty in relying on statistics before discovery. 965 F.3d at 959 n.8 (“We recognize that prior to discovery it may be difficult for [plaintiffs] to allege with statistical accuracy the number of policy claims by disabled persons relative to non- disabled persons that were denied under the hearing loss exclusion, as this information may reside exclusively with Kaiser.”). Statistics aside, Plaintiffs may also successfully plead a proxy discrimination claim “by alleging facts showing how the needs of hearing disabled persons differ from the needs of persons whose hearing is merely impaired such that the exclusion is likely to predominately affect disabled persons.” Id. Plaintiffs tried that approach but failed. They only asserted conclusory allegations that many non-disabled insureds do not seek coverage at all and that even if non-disabled insureds did indeed seek treatment, they would be denied coverage because such treatment would not be a “medical necessity.” Dkt. # 32 ¶¶ 54-58. These allegations, lacking underlying facts, were insufficient. Still, because Plaintiffs’ complaint may be cured “by the allegation of other facts,” the Court grants leave to amend. Lopez v. Smith, 203 F.3d 1122, 1127 (9th Cir. 2000) (en banc) (quoting Doe v. United States, 58 F.3d 494, 497 (9th Cir. 1995)). B. Breach of Contract (Count 2) Plaintiffs assert a second claim for “breach of contract and violation of RCW 48.43.0128.” Dkt. # 32 ¶¶ 125-29. According to Plaintiffs, “[a]ll Washington health plan [sic] incorporate the relevant requirements of the Insurance Code as additional terms.” Id. ¶ 126. RCW 48.43.0128, they say, is one such statute incorporated into Regence’s policy. See id. ¶¶ 125-29. That section, like the ACA, forbids healthcare insurers from discriminating in their benefit “design” or “implementation” because of an individual’s “present or predicted disability.” RCW 48.43.0128(1). Because Regence allegedly discriminated against disabled hearing insureds when it designed its policy, Plaintiffs claim that Regence violated the state statute. Dkt. # 32 ¶ 129. Moreover, because RCW 48.43.0128 is incorporated in Regence’s policy, Plaintiffs claim that “Regence’s use of the Exclusion to deny coverage is also a breach of contract.” Id. To plead a breach of contract claim under Washington law, a plaintiff must allege “that a valid agreement existed between the parties, the agreement was breached, and the plaintiff was damaged.” Univ. of Washington v. Gov’t Emps. Ins. Co., 404 P.3d 559, 566 (Wash. Ct. App. 2017). Here, Plaintiffs’ amended complaint fails to identify a contract term that Regence breached, and Plaintiffs have therefore failed to state a claim. Compliance with RCW 48.43.0128 is not part of the parties’ contract. There is no provision in Regence’s policy stating that Regence must comply with RCW 48.43.0128 to fulfill its performance. See Dkt. # 32-1. Indeed, the policy does not mention RCW 48.43.0128 or any other section of Washington’s insurance code at all. Id. Thus, compliance with RCW 48.43.0128 was not a contractual term, and Plaintiffs could not have breached the contract by violating that section. Because they do not allege another breached contract term, Plaintiffs’ claim fails. For their part, Plaintiffs offer a novel breach theory. They allege that every insurance contract incorporates all relevant portions of Washington’s insurance code and that any violation of the insurance code could thus serve as a basis for a breach of contract claim. Dkt. # 32 ¶¶ 125-29. The Court agrees that it is black letter law in Washington that “relevant statutes are read into insurance contracts” so that “a valid statute becomes a part of and should be read into the insurance policy.” McLaughlin v. Travelers Com. Ins. Co., 476 P.3d 1032, 1036-37 (Wash. 2020) (quoting Touchette v. Nw. Mut. Ins. Co., 494 P.2d 479 (Wash. 1972)) (collecting cases). Given that, courts often apply this legal doctrine by using insurance statutes either to interpret an undefined contract term or to strike down provisions that violate public policy. See, e.g., id. at 1035-36 (using RCW 48.22.005(11) to define the word “pedestrian,” an undefined term in an insurance policy); Hartford Acc. & Indem. Co. v. Novak, 520 P.2d 1368, 1371-72 (Wash. 1974) (using RCW 48.22.030 to void a policy that limited “hit-and-run” coverage only to when a hit-and-run driver made “physical contact” with the injured driver because RCW 48.22.030 intended to protect “an insured for injuries or damages proximately caused by a hit-and-run vehicle, irrespective of its actual physical contact with the vehicle of the insured”); Touchette, 494 P.2d at 484 (using RCW 48.22.030 “to declare a public policy overriding . . . exclusionary language”). But Plaintiffs are asking the legal doctrine to do much more. They are not asking the Court to use an insurance statute merely to define a term in Regence’s policy. Nor are they using RCW 48.43.0128 solely to void the Exclusion. Instead, they ask that the Court apply the doctrine to read a statute into an insurance policy and, from there, conclude that any violation of the statute is in and of itself a breach of contract. In effect, that would grant a private right of action for any violation of the insurance code—a novel and sweeping theory. In support of that theory, Plaintiffs cite no case. And the Court similarly has found none. Absent any authority dictating otherwise, the Court declines to extend the doctrine that far. Surely, Regence’s policy must comply with Washington’s insurance regulations. McLaughlin, 476 P.3d at 1036-37 (“There is no longer any judicial doubt that the state may regulate insurance, so closely is that industry affected with the public interest, and regulatory statutes become a part of the policy of insurance.” (quoting Touchette, 494 P.2d at 482-83)). But it does not follow that non-compliance results in a breach of contract and hence damages. The Court has encountered no authority for that conclusion. Rather, based on the case law, non-compliance ordinarily results in a declaration that a policy provision is void and unenforceable, something that Plaintiffs do not seek given that they do not assert a declaratory judgment claim. Hence, because the Court rejects Plaintiffs’ breach theory, and because Plaintiffs fail to allege a contract term that Regence breached, the Court dismisses Plaintiffs’ breach of contract claim. For the reasons stated above, the Court GRANTS Defendants’ Motion to Dismiss Amended Complaint. Dkt. # 37. The Court grants Plaintiffs leave to file an amended complaint within 21 days of the entry of this Order. DATED this 31st day of January, 2022. A

The Honorable Richard A. Jones United States District Judge