1

2

3

4

5

UNITED STATES DISTRICT COURT

6 NORTHERN DISTRICT OF CALIFORNIA

7

8 ANDREW C. AND ROBERT C., Case No.: 17-CV-2072-YGR 9 Plaintiffs, ORDER GRANTING MOTION FOR JUDGMENT 10 IN FAVOR OF PLAINTIFF AND DENYING v. DEFENDANT’S CROSS-MOTION 11 ORACLE AMERICA INC. FLEXIBLE BENEFIT Dkt. Nos. 70, 71 12 PLAN, UNITEDHEALTHCARE INSURANCE COMPANY, et. al. 13 Defendants. 14 15 Presently before the Court are cross-motions for judgment on plaintiffs’ claim for health 16 benefits under a plan covered by the Employee Retirement Income Security Act (“ERISA”), 29 17 U.S.C. § 1001, et seq. filed by plaintiffs Andrew C. and Robert C. (Dkt. No. 70) and defendants 18 Oracle America Inc. Flexible Benefit Plan and United Healthcare Insurance Company (Dkt. No. 19 71). 20 Having considered the parties’ briefing and the complete administrative record,1 the Court 21 GRANTS plaintiffs’ motion for judgment pursuant to Rule 52 of the Federal Rules of Civil 22 Procedure on plaintiffs’ claim for health benefits and DENIES defendant’s cross-motion on that 23 claim.2 24 25

26 1 UHC amended the administrative record after resubmitting an appeal to an external reviewer. (See Dkt. No. 65, 66.) 27 2 The Court has reserved on plaintiffs’ breach of fiduciary duty claim. (See Scheduling and Sealing Order, Dkt. No. 36, at 1:12-13.) As set forth herein, the Court will seek the parties’ input 28 on proceedings as to that claim. 1 I. APPLICABLE STANDARD 2 Plaintiffs appeal a denial of healthcare plan benefits under ERISA, 29 U.S.C. section 3 1132(a)(1)(B). Beneficiaries and plan participants may sue in federal court “to recover benefits due 4 to [them] under the terms of [their] plan, to enforce [their] rights under the terms of the plan, or to 5 clarify [their] rights to future benefits under the terms of the plan.” 29 U.S.C. § 1132(a)(1)(B). 6 The Court has ruled previously that the standard applicable here is the de novo review 7 standard. (Dkt. No. 63.) On a de novo review, the court conducts a bench trial on the record, and 8 makes findings of fact and conclusions of law based upon that record. Kearney v. Standard Ins. 9 Co., 175 F.3d 1084, 1095 (9th Cir. 1999) (bench trial may “consist[] of no more than the trial judge 10 reading [the administrative record].”).3 Plaintiff bears the burden of establishing entitlement to 11 benefits during the claim period by a preponderance of the evidence, and the Court must evaluate 12 the persuasiveness of the conflicting evidence to make its determination. Id. at 1094-95; Eisner v. 13 The Prudential Ins. Co. of Am., 10 F.Supp.3d 1104, 1114 (N.D. Cal. 2014). 14 Under a de novo standard, a court does not give deference to an insurer’s determination to 15 deny benefits. Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 115 (1989). Instead the court 16 17 “determines in the first instance if the claimant has adequately established” entitlement to benefits 18 under the plan. Muniz v. Amec Constr. Mgmt. Inc., 623 F.3d 1290, 1295-96 (9th Cir. 2010). “In 19 conducting a de novo review, the Court gives no deference to the insurer’s interpretation of the plan 20 documents, its analysis of the medical record, or its conclusion regarding the merits of the 21 plaintiff’s benefits claim.” McDonnell v. First Unum Life Ins. Co., Case No. 10-cv-8140, 2013 WL 22 3975941, at *12 (S.D.N.Y. Aug. 5, 2013); Tedesco v. I.B.E.W. Local 1249 Ins. Fund, No. 14-CV- 23 3367 (KBF), 2017 WL 3608246, at *6 (S.D.N.Y. Aug. 21, 2017), aff'd, 729 F.App’x 136 (2d Cir. 24 2018) (citing Firestone, 489 U.S. at 112) (same). 25

26 3 The Court finds that the administrative record here suffices and a trial with live witness testimony is not necessary to decide the claim for benefits. The competing administrative records 27 submitted by the parties at Docket Nos. 39 and 45 were superseded by the complete record filed at Docket No. 64. The Court refers to the administrative record by reference to the sequential page 28 numbers, herein denoted as “AR #” (denoted in original as UHC#). 1 II. FINDINGS OF FACT 2 Plaintiff Andrew C. was sixteen years old when he was admitted to residential mental health 3 treatment at Change Academy Lake of the Ozarks (“CALO”). Andrew remained at CALO for a 4 little over one year, from January 29, 2014 through February 20, 2015. Robert C., Andrew’s father 5 and a participant in the Plan, made a claim for benefits for Andrew’s treatment at CALO. 6 Ultimately, after several appeals, UHC denied benefits for all but the first 30 days of treatment at 7 CALO, giving rise to the instant action. 8 A. Andrew’s History 9 Andrew is a child of “African American/Black Cuban” descent. (AR 780-81.) He was born 10 to a mother who had used alcohol and drugs, including heroin, during her first trimester of her 11 pregnancy with Andrew and was incarcerated during the second two trimesters. (AR 1248.) 12 Andrew was placed in foster care from the time he was born. After being placed in a series of 13 foster homes, at thirteen months old Andrew came to live with and eventually be adopted by Robert 14 C. and his wife, a “Caucasian” couple. (AR 781, 1248.) 15 At the time he was placed with his adoptive parents, Andrew was diagnosed with mild 16 17 cerebral palsy and was required to wear leg casts to correct an ankle defect from age 2 to 4. He had 18 weakness on the left side of his body, including trouble eating and swallowing as a small child, 19 requiring physical therapy up until he was in fifth grade. He also had a moderate speech delay 20 requiring speech therapy. (AR 1253.) He continues to have some left side weakness and difficulty 21 with fine motor skills. (Id.) 22 When he was elementary school-age, Andrew had trouble managing his emotions, 23 sometimes erupting in angry outbursts or tearing his room apart, and occasionally having 24 disciplinary issues at school. Andrew began taking medication in the third grade for attention 25 deficit hyperactivity disorder (ADHD) to help him control his fits of anger or rage. In fourth grade, 26 his parents pulled him out of public school in favor of homeschooling due to bullying and racial 27 discrimination at school. (AR 1251.) 28 1 Andrew started therapy in fifth grade for anger issues, attending for about a year. (AR 2 1254.) A year later, after stealing from his parents, Andrew began individual and family therapy 3 with Robert Brennan, MFT, to deal with his emotional and behavioral issues. Brennan diagnosed 4 Andrews as having Reactive Attachment Disorder with difficulty regulating emotion. (Id.) In 5 addition to therapy, Andrew continued to take medication for ADHD, later adding another 6 medication to stabilize his mood and help address his anger control issues. (Id.) 7 As he entered adolescence, Andrew’s behavior began to be more concerning. He engaged 8 in illicit behavior in his social groups and other programs, causing him to be asked to leave. (AR 9 1249.) He would erupt in anger, punching other children or punching and kicking walls and doors, 10 and would remain agitated, pacing and yelling after an outburst. 11 The incident immediately preceding his admission to CALO involved a heated argument in 12 December 2013 between Andrew and his parents which ended in Andrew shoving his mother and 13 punching his father in the face, and his parents calling the police. In addition to breaking his own 14 hand, Andrew broke his father’s nose and eye socket, injuries which required his father to undergo 15 facial surgery. (AR 1247.)4 Andrew was taken to a juvenile detention facility and remained in 16 17 detention until he was transferred to the residential treatment program at CALO as a term of his 18 probation. (AR 775, 1248.)5 19 Andrew’s therapist, Robert Brennan, submitted an opinion letter in the juvenile proceedings 20 stating: I have provided professional family counseling services to the [ ] family 21 since August 3, 2012. I met with the family on a weekly basis since the start date and for the last six months have met twice per week. 22 Over the last six months I have observed a slow increase in the escalation 23 of arguments within the family. The escalation occurs in the level of anger and rage expressed by Andrew [C.]. The most recent violent episode resulting in 24 incarceration represents a family dynamic that is physically unsafe. 25 It is my judgment that if the [ ] family resume living together at this time 26 4 This incident was the third time in six months that Andrew had broken his right hand as a 27 result of punching other people or hard objects. (AR 1249, 1254.) 5 The Court notes that “[d]efendants do not dispute that the medical records show Andrew’s 28 ongoing chronic issues.” (Oppo. and Cross-Motion at 3.) there will be a continuation of escalation and violence. Until there is a greater self 1 control that can be gained for Andrew [C.] it will not be in his best interest to return him to the family home. 2 I recommend to the court that a one[-]year inpatient treatment program be 3 ordered for Andrew [C.]. The treatment program must include psychiatric and psychological interventions to facilitate Andrew [C.’s] rehabilitation. 4 (AR 914.) 5 B. Andrew’s Treatment at CALO 6 Andrew was admitted to CALO on January 20, 2014. The psychosocial assessment done by 7 CALO’s staff at intake indicates that his diagnoses were Oppositional Defiant Disorder, Reactive 8 Attachment Disorder, ADHD, dyslexia, and dysgraphia, based on his prior outpatient treatment 9 with Dr. Charles Montgomery, Dr. Alice Del Rosario, and his therapist Mr. Brennan. (AR 236-37.) 10 His treatment plan identified problems with attachment, mood regulation, and social skills. (AR 11 12 245.) The planned treatment modalities at intake included weekly, therapist-led group therapy, 13 weekly individual therapy, and weekly family process therapy, in addition to medication 14 management. (Id.) Records from CALO, including therapy notes and treatment team summaries, 15 indicate that Andrew participated in all these forms of treatment throughout his time at CALO, 16 though at times refusing to participate when he was upset with his parents or therapist. (See AR 17 245-482, AR 929.) The treatment plans and treatment team summaries reflect that Andrew 18 generally participated in his individual and group therapy sessions—albeit with varying degrees of 19 engagement—and after initially refusing to participate in family therapy sessions, within his second 20 month at CALO, also began participating in those sessions consistently. (Id. at 247, 468-82 21 [Treatment Team Summaries].) Over the course of his stay, Andrew improved somewhat in his 22 interactions with peers at CALO and would accept coaching from CALO staff without as much 23 pushback as when he first was admitted. (UHC 468-82.) CALO staff consistently noted his need 24 to improve his temper and control his frustration. (Id.) 25 On May 4, 2014, Todd Odell, LPS, Andrew’s primary therapist at CALO, provided excerpts 26 of notes from his therapy sessions with Andrew. (AR 929-930.) Mr. Odell noted that “Andrew 27 lives in extreme fear everyday” which was “caused by his developmental trauma.” (AR 930.) 28 1 Odell opined that “Andrew needs specialized residential care or he will become over[-]stimulated, 2 confused, highly reactive and un-safe to his family and community members.” (Id.) 3 Andrew was treated at CALO until his discharge on February 20, 2015. On May 4, 2015, 4 Odell provided a summary of incident reports of Andrew’s aggressive or inappropriate behavior 5 while at CALO. (AR 931- 41.) He also provided excerpts of some of the daily records from 6 Andrew’s treatment, noting that Andrew showed a “consistent struggle with managing his 7 emotions, his impulsivity, his need for separation from his peers and his multiple physical 8 confrontations.” (AR 934.) Odell opined that “[t]hese examples provide very specific details as to 9 why Andrew required and still requires an acute level of care at a residential facility.” (AR 934.) 10 Subsequent to his treatment at CALO, on March 25, 2015, Dr. Kevin O’Keefe examined 11 Andrew and provided a psychological assessment report. (AR 942.) Dr. O’Keefe interviewed 12 Andrew and obtained a complete history from his parents, in addition to conducting a battery of 13 psychological tests. (Id.) After setting forth a summary of Andrew’s psychosocial, medical, 14 developmental, and mental health history, and his testing results, Dr. O’Keefe concluded that 15 Andrew met the criteria for Reactive Attachment Disorder, Persistent Depressive Disorder, and 16 17 Unspecified Anxiety Disorder, ADHD, as well as a strong possibility of a developmental disorder. 18 (AR 954, 956-57, 959.) Dr. O’Keefe’s treatment recommendations included attending further 19 residential treatment to address the issues he noted in the report. (AR 957.) He opined that “[i]f he 20 does not receive additional treatment, the potential for these problems to become worse is 21 significant. This is particularly true given the fact Andrew has been resistant to engaging in therapy 22 in the past.” (Id.) 23 C. Plan Terms 24 The Plan’s Certificate of Coverage defines “Covered Health Services” as follows: 25 Covered Health Services are those health services provided for the purpose of 26 preventing, diagnosing or treating a sickness, injury, mental illness, substance use disorder, or their symptoms. 27 28 (AR 34.) The Plan covers “Mental Health Services” including “the following services provided on 1 an inpatient basis: Partial Hospitalization/Day Treatment; and services at a Residential Treatment 2 Facility.” (AR 48-49.) With respect to inpatient mental health treatment, the Plan provides that:

3 Coverage for inpatient treatment of mental disorders is provided when the facility is a licensed crisis stabilization unit or a licensed residential treatment center, or 4 when the facility meets all of the following conditions: 5 1. Mainly provides a program for the diagnosis, evaluation and effective treatment of mental disorders. Effective treatment describes a program that is: 6 (a) Prescribed and supervised by a physician, and 7 (b) Is for a disorder that can be favorably changed. 2. Makes charges for services. 8 3. Meets licensing standards. 9 4. Is not mainly a school or a custodial, recreational or training institution. 5. Provides infirmary-level medical services. Also, it provides or arranges with a 10 hospital in the area for any other medical services that may be required. 11 6. Is supervised full-time by a psychiatrist who is responsible for patient care and who is on-site regularly. 12 7. Is staffed by psychiatric physicians involved in care and treatment. A psychiatrist is a physician who: 13 (a) specializes in psychiatry, or 14 (b) has the training and experience to do the required evaluation and treatment of mental illness. 15 8. Has a psychiatric physician present during the whole treatment day. 16 9. Provides, at all times, psychiatric social work and nursing services. 10. Provides, at all times, skilled nursing care by licensed nurses who are 17 supervised by a full-time RN. 18 11. Prepares and maintains a written plan of treatment for each patient based on medical, psychological and social needs. A psychiatric physician must supervise 19 the plan. 20 (AR 51-52, emphasis supplied.) The Plan states that inpatient mental health services require 21 authorization from United Behavioral Health, UHC’s administrator for processing claims for 22 mental health services. (AR 48.) 23 Under “What’s Not Covered,” the Plan defines “Mental Health and Substance Abuse 24 Exclusion” as: 25 o Services performed in connection with conditions not classified in the 26 current edition of the Diagnostic and Statistical Manual of the American 27 Psychiatric Association; o Services or supplies for the diagnosis or treatment of Mental Illness, 28 alcoholism or substance use disorders that, in the reasonable judgment of the Mental Health/Substance Use Disorder Administrator, are any of the 1 following: 2  not consistent with generally accepted standards of medical practice for the treatment of such conditions; 3  not consistent with services backed by credible research soundly demonstrating that the services or supplies will have a measurable 4 and beneficial health outcome, and therefore considered 5 experimental;  not consistent with the Mental Health/Substance Use Disorder 6 Administrator’s level of care guidelines or best practices as modified from time to time; or 7  not clinically appropriate for the patient’s Mental Illness, substance 8 use order or condition based on generally accepted standards of medical practice and benchmarks. 9 o Mental Health Services as treatments for V-code conditions6 as listed within the current edition of the Diagnostic and Statistical Manual of the 10 American Psychiatric Association; 11 o Mental Health Services as treatment for a primary diagnosis of insomnia other sleep disorders, sexual dysfunction disorders, feeding disorders, 12 neurological disorders and other disorders with a known physical basis; 13 o treatments for the primary diagnoses of learning disabilities, conduct and impulse control disorders, personality disorders and paraphilias (sexual 14 behavior that is considered deviant or abnormal); o educational/behavioral services that are focused on primarily building 15 skills and capabilities in communication, social interaction and learning; 16 o tuition for or services that are school-based for children and adolescents under the Individuals with Disabilities Education Act; 17 o learning, motor skills and primary communication disorders as defined in the current edition of the Diagnostic and Statistical Manual of the 18 American Psychiatric Association; 19 o mental retardation as a primary diagnosis defined in the current edition of the Diagnostic and Statistical Manual of the American Psychiatric 20 Association; o methadone treatment as maintenance, L.A.A.M. (1-Alpha-Acetyl- 21 Methadol), Cyclazocine, or their equivalents for drug addiction; 22 o any treatments or other specialized services designed for Autism Spectrum Disorder that are not backed by credible research demonstrating that the 23 services or supplies have a measurable and beneficial health outcome and therefore considered Experimental or Investigational or Unproven 24 Services. 25 (AR 63, emphasis supplied.) The Plan further lists “Custodial Care” as a general exclusion. 26

27 6 “V-codes” refers to diagnostic codes used for substance abuse or life transition issues. See K.M. v. Regence Blueshield, No. C13-1214 RAJ, 2014 WL 801204, at *12 (W.D. Wash. Feb. 27, 28 2014). 1 (AR 66.)7 2 The Plan section entitled “Filing a Claim for Benefits” provides that:

3 When you receive a Covered Health Service from a non-Network provider, you are responsible for requesting payment from UnitedHealthcare. You must file the 4 claim in a format that contains all of the information described below. You must submit a request for payment of benefits within one year after the 5 date of the service. If you don’t provide this information to UnitedHealthcare 6 within one year of the date of service, benefits for that health service will be denied or reduced, at Oracle’s or UnitedHealthcare’s discretion. The time 7 limit does not apply if you are legally incapacitated. If your claim relates to an inpatient stay, the date of service is the date your inpatient stay ends. 8 9 (AR 70, emphasis in original.) The Plan requires that a denial of benefits provide the specific 10 reasons for the denial and a reference to the Plan provisions on which the denial is based. (AR 72.) 11 If the denial is based on medical necessity or a similar exclusion, the Plan requires “either an 12 explanation of the scientific or clinical judgment for the determination, applying the terms of the 13 Plan to your medical circumstances, or a statement that such explanation will be provided to you 14 free of charge upon request.” (Id.) 15 The Plan provides an appeal process, which permits the member to submit additional 16 documents, records, and other information related to the claim, and shall be conducted without 17 deference to the original denial. (AR 73.) After the initial appeal, the Plan further provides for a 18 second level appeal from UHC’s decision (AR 74) and an additional External Review Program if 19 UHC’s denial is based upon medical judgments concerning, for example, “medical necessity, 20 appropriateness, health care setting, level of care, or effectiveness” (AR 76). The independent 21 external review requires UHC’s submission of all relevant medical records, all documents relied 22 upon by UHC in making the decision, and all information or evidence submitted by the member to 23

24 7 While noting the exclusion of Custodial Care at several places in the Plan for several different kinds of care, the terms “custodial” and “Custodial Care” are not defined in the Plan or 25 distinguished from inpatient hospitalization or inpatient residential treatment. (AR 45, 46, 52, 58, 66.) The nearest the Plan itself comes to defining “custodial” care is under the “Skilled Nursing 26 Facility/Inpatient Rehabilitation Facility” definition, which states that “Benefits are NOT available for custodial, domiciliary or maintenance care (including administration of enteral feeds) which, 27 even if it is ordered by a physician, is primarily for the purpose of meeting personal needs of the covered person or maintaining a level of function, as opposed to improving that function to an 28 extent that might allow for a more independent existence.” (AR 58, emphasis in original.) 1 UHC. (AR 76-77.) 2 D. Optum Level of Care Guidelines 3 As indicated above, the Plan defines mental health services not covered, in part, by 4 reference to “the Mental Health/Substance Use Disorder Administrator’s level of care guidelines or 5 best practices as modified from time to time.” (AR 63.) Here, the level of care guidelines at issue 6 are the Optum8 Level of Care Guidelines (“LOC Guidelines”),9 and the Optum Coverage 7 Determination Guideline for Treatment of Oppositional Defiant Disorder (“ODD Guideline”). (AR 8 139-152 [2014 LOC Guidelines], AR 153-169 [2015 LOC Guidelines]); AR 170-190 [ODD 9 Guideline].) The LOC Guidelines indicate that they are to be used “when making medical 10 necessity determinations and as guidance when providing referral assistance.” (AR 142.) They 11 define “medically necessary” as follows: 12 . . . [care] provided for the purpose of preventing, evaluating, diagnosing or 13 treating a mental illness or substance use disorder, or its symptoms that are all of the following as determined by us or our designee, within our sole discretion: 14 1. In accordance with Generally Accepted Standards of Medical Practice. 15 2. Clinically appropriate, in terms of type, frequency, extent, site and duration, 16 and considered effective for the mental illness, substance use disorder, or its symptoms. 17 3. Not mainly for the member's convenience or that of the member's doctor or other health care provider. 18 4. Not more costly than an alternative drug, service or supply that is at least as 19 likely to produce equivalent therapeutic or diagnostic results as to the diagnosis or treatment of the member’s mental illness, substance use disorder, or its symptoms. 20 21 (Id.)10 22

23 8 “Optum is a brand used by United Behavioral Health and its affiliates.” (AR 139.) 9 UBH did not include the text of the guidelines or enclose copies of the guidelines with 24 denial letters, nor did UBH specify in its correspondence which annual version of the Guidelines it 25 applied. (See AR 1091-1099.) The Court has considered all the Optum Guidelines submitted in the administrative record. 26 10 The Optum Level of Care Guidelines note that they are to be “used flexibly, and [are] intended to augment - but not replace - sound clinical judgment.” (AR 143.) “Use is informed by 27 the unique aspects of the case, the member's benefit plan, services the provider can offer to meet the member's immediate needs and preferences, alternatives that exist in the service system to meet 28 those needs, and the member's broader recovery, resiliency and wellbeing goals.” (Id.) 1 The “Common Criteria” for admission in the LOC Guidelines are:

2 (1) eligibility for benefits; 3 (2) the current condition cannot be “safely, efficiently and effectively . . . treated in a less intensive setting due to acute changes in the members’ signs and 4 symptoms and/or psychosocial and environmental factors;” (3) the condition and proposed services are covered by the Plan; 5 (4) the services are within the scope of the provider’s professional training and 6 licensure; 7 (5) services are consistent with generally accepted standards of practice, clinically appropriate, and consistent with Optum’s best practice guidelines; 8 (6) it is reasonably expected that the services will improve the member’s presenting problems within a reasonable period of time; and 9 (7) “[t]reatment is not primarily for the purpose of providing social, custodial, 10 recreational, or respite care.”

11 (AR 144-46.) The Common Criteria in the LOC Guidelines further state that continued treatment is 12 appropriate when the admission criteria are still met and “active treatment” is being delivered, that 13 is “supervised and evaluated” by the provider; provided in an individualized treatment plan focused 14 on presenting admission factors and using clinical best practices; “reasonably expected to stabilize 15 the member’s condition” within a reasonable period of time; and engaging the member’s family 16 17 and other resources as clinically indicated. (Id.) The LOC Guidelines’ common criteria indicate 18 discharge is appropriate when: (1) the factors leading to admission “have been addressed to the 19 extent that the member can be safely transitioned to a less intensive level of care or no longer 20 requires treatment;” (2) the factors that led to admission cannot be addressed and the member 21 requires a more intensive level of care; (3) the member is unwilling or unable to participate in 22 treatment and involuntary treatment is not being pursued; or (4) “the member requires care that is 23 primarily social, custodial, recreational, or respite.” (AR 144-45.) 24 Particularly with respect to residential treatment centers, the 2014 LOC Guidelines state that 25 continued treatment cannot be “primarily for the purpose of providing custodial care” which is 26 specified as involving “services that don’t seek to cure, are provided when the member’s condition 27 is unchanging, are not required to maintain stabilization, or don’t have to be delivered by trained 28 clinical personnel.” (AR 150.) The 2015 LOC Guidelines for residential treatment centers define 1 custodial care more particularly, requiring that continued services are “not primarily for the purpose 2 of providing custodial care” and specifying that:

3 Services are custodial when they are any of the following: 2.2.1. Non-health-related services, such as assistance in activities of daily living 4 (examples include feeding, dressing, bathing, transferring and ambulating). 2.2.2. Health-related services that are provided for the primary purpose of meeting 5 the personal needs of the patient or maintaining a level of function (even if 6 the specific services are considered to be skilled services), as opposed to improving that function to an extent that might allow for a more 7 independent existence. 2.2.3. Services that do not require continued administration by trained medical 8 personnel in order to be delivered safely and effectively. 9 (AR 167-68.) 10 Optum and its reviewers also use more specific Guidelines tailored to treatment of particular 11 mental disorders. Relevant here, the ODD Guideline reflects “Optum’s understanding of current 12 best practices in care,” including standards for diagnosis and assessment of the member, parameters 13 for treatment planning, and criteria for determining the most appropriate level of care for treatment 14 for Oppositional Defiant Disorder. (AR 171; see generally AR 170-190.) The ODD Guideline lists 15 the admission criteria for inpatient residential treatment (AR 185), in part, as follows: 16  The member is experiencing a disturbance in mood, affect or cognition 17 resulting in behavior that cannot be safely managed in a less restrictive setting. 18 - OR -  There is an imminent risk that severe, multiple and/or complex psychosocial 19 stressors will produce significant enough distress or impairment in psychological, social, occupational/educational, or other important areas of 20 functioning to undermine treatment in a lower level of care. - OR - 21  The member has a co-occurring medical disorder or substance use disorder which complicates treatment of the presenting mental health condition to the 22 extent that treatment in a Residential Treatment Center is necessary.

23 (AR 185, emphasis supplied.) By contrast, the criteria for an intensive outpatient program 24 admission include that the member’s symptoms and impairment of “psychological, social” or other 25 functioning are “moderate,” and that the member and his family are able to “comply with the 26 27 requirements” of such an outpatient program. (AR 183-84.) Similarly, the least intensive treatment 28 level—“outpatient care”—requires that “[t]he member exhibits adequate behavioral control to be 1 treated in this setting.” (AR 182-83.) Each of these levels of care requires that the member “is not 2 at imminent risk of serious harm to self or others.” (AR 183-185.) 3 E. Claim Review History 4 1. Initial Claim 5 Between April and December of 2014, UHC received several claims for benefits from 6 CALO for the treatment Andrew received there. (See AR 1016-1036 [Member Claims History].) 7 Plaintiffs were sent “explanations of benefits” (EOBs) denying coverage for the bills submitted for 8 treatment at CALO. (AR 727-748 [Member Explanations of Benefits]; AR 749-771 [Provider 9 Explanations of Benefits].) UHC requested records from CALO so it could assess whether the 10 treatment was covered under the Plan. 11 On August 20, 2014, UHC Care Advocate Alexander Thomas conducted a “retrospective 12 review” encompassing the “Medical Record / Linx [internal case notes] / [and] Claim Information.” 13 (AR 1047.) Thomas noted that Andrew had been admitted to CALO for treatment of ADHD, 14 ODD, and reactive attachment disorder. (Id.) He noted that Andrew previously was treated in 15 outpatient individual and family therapy. (Id.) His notes state that Andrew was “engaged and 16 17 partic[i]pates in treatment protocols, works on interpersonal skills, developing better coping skills 18 and insight.” (Id.) He noted that the records referenced “periodic emotional/behavioral 19 dysregulation requiring staff intervention and at which point [he] calms down and regains his 20 composure.” (Id.) Thomas referred the claim for an administrative review by Dr. Jeffrey Uy, MD. 21 (Id.) 22 On August 21, 2014 UHC issued a denial letter, signed by Jeffrey C. Uy, MD, stating that 23 “a reconsideration of the previously issued benefit determination has been completed” and 24 “[f]ollowing a discussion with Change Academy of the Ozarks on 8/21/2014,”11 I have determined 25 11 Dr. Uy’s case notes contradict this statement in the letter. Dr. Uy’s notes do not show 26 any conversation with CALO, nor do the notes of care advocate Alexander Thomas. (AR 1047, 1048.) According to the internal case notes, Dr. Uy and Mr. Thomas reviewed only the medical 27 records and internal file notes. (AR 1047 [A. Thomas: “Review Based on Medical Record / Linx / Claim Information”]; AR 1048 [Dr. Uy: “avail documentation reviewed, case d/w CA-A Thomas, 28 see note dated 8/20/2014”].) 1 that benefit coverage is not available. (AR 776.) The letter stated:

2 Specifically, upon review of the clinical information presented, it appears that your child would be most appropriate for a custodial level of care; which is not a 3 covered benefit[.] Custodial care is defined as a domiciliary setting that provides a protected, controlled environment for the primary purpose of assuring the 4 patient’s safety and/or providing services necessary to assure the patient's competent functioning. It is not expected that the care provided will significantly 5 reduce the disability to the extent necessary for the patient to function outside the 6 controlled environment. This includes those patients for whom there is little expectation of improvement in spite of multiple repeated treatment attempts 7 and/or patients who have been repeatedly non-adherent with treatment recommendations and have therefore demonstrated the inability to function 8 outside of a controlled environment.

9 (Id.) Dr. Uy authorized the alternative service of “Mental Health Outpatient Services/ individual 10 psychotherapy and medication management.” (AR 777.) The letter stated that the decision was 11 based upon “Optum Health’s Coverage Determination Guideline for Oppositional Defiant Disorder 12 for Level of Care: Mental Health Residential Treatment Center” as well as clinical guidelines from 13 the American Academy of Child and Adolescent Psychiatry (AACAP) for the Assessment and 14 Treatment of Children and Adolescents with Oppositional Defiant Disorder. (Id.) 15 2. Appeal of Dr. Uy’s Determination and Denial by Dr. Sane 16 On February 8, 2015, plaintiffs submitted an appeal of UHC’s denial of benefits, including 17 information from CALO (i.e., the admission application, Client information, Parent information, 18 Referral information, Intervention history, Psychological history, treatment plans, treatment team 19 summaries); a letter from Charles Montgomery, M.D. (Andrew’s prescribing psychiatrist) that 20 recommended residential treatment; a letter from Robert Brennan, M.A., M.F.T., (Andrew’s 21 outpatient therapist) recommending residential treatment; and a portion of an order of probation 22 23 from Alameda County’s juvenile court listing one year of residential treatment at CALO as a 24 condition of probation. (AR 773 -970.) 25 On March 13, 2015, UHC partially reversed its prior denial, authorizing payment for the 26 first 30 days of plaintiff’s treatment at CALO and denying the remainder. (See AR 1014 [member 27 28 1 Authorization History]; AR 1091-99 [letter from Dr. Natasha Sane, Associate Medical Director].)12 2 The letter indicated Dr. Sane had reviewed the appeal request received February 12, 2015, as well 3 as the “case records, claims database, [and] medical records.” (AR 1091.) The letter stated:

4 Per review of the available clinical information, I find that benefit coverage was available for treatment at the Mental Health Residential Level of Care from 5 1/29/2014 through 2/27/2014 for acute stabilization of aggressive behavior & 6 impulsivity, monitoring, and working towards safe return to home/community environment. However, as of 2/28/2014, your child was not reported to be at 7 immediate risk of harm to himself or others. Though he continued to exhibit intermittent impulsivity and emotional reactivity, there was no indication that he 8 required continued stabilization in a residential setting. He was medically stable. There was no reasonable expectation for improvement with further 24 hour 9 treatment, nor was there significant change in his behavior as treatment continued. It appeared that your child was at a baseline level of functioning and in need of 10 custodial care, which is not a benefit under his insurance plan.

11 Under the terms of his Summary Plan Description (SPD), treatment at a Mental Health Residential level of care starting 2/28/14 would not be considered 12 consistent with the Mental Health/Substance Use Disorder Administrator’s level of care guidelines or best practice guidelines. Under the terms of the SPD, 13 custodial care is excluded. 14 (AR 1091-92). The letter indicated it was based on clinical guidelines from the AACAP. (AR 15 1092.) It did not indicate that it relied on any particular LOC Guidelines or the ODD Guidelines. 16 3. Denial Letter Issued by William Suhay 17 On July 16, 2015, UHC issued a denial letter signed by William Suhay, MFT. (AR 965- 18 67.)13 The letter stated that services for March 2, 2014 through February 20, 2015 at CALO were 19 denied. (AR 965.) The denial letter stated: “This review included an examination of the following 20 information: medical records.” (AR 965.)14 The stated basis for the denial was that: 21 Your child was not reported to be at risk of harm to himself or others. Though he 22 continued to exhibit intermittent impulsivity and emotional problems, there was 23 12 UHC’s “Member Claims History” records confusingly categorize plaintiff’s treatment at 24 CALO variously as “outpatient,” “acute inpatient,” “professional services,” and “residential” in these records and identify only ADHD as the relevant diagnosis, making these records less than 25 reliable as evidence here. (AR 1024-1029.) 26 13 UHC’s internal file notes indicate that the review was conducted by Dr. Svetlana Libus. (AR 1053-55.) 27 14 UHC’s internal file notes state that the review was not limited to the “medical records” but also included “Letters from OP [outpatient] providers, Linx electronic clinical notes, and 28 coverage determination guideline,” which the notes identified as the ODD Guideline. (AR 1054.) no indication that he required continued stabilization in [a] residential setting. He 1 was medically stable. There was no reasonable expectation for improvement of his behavior with further 24[-]hour treatment, nor was there significant change in 2 his behavior as treatment continued. It appeared that your child was at a baseline level of functioning and in need of custodial care, which is not a benefit under his 3 insurance plan. 4 (AR 965.) The reviewer found that coverage would be inconsistent with UBH’s “level of care 5 guidelines or best practice guidelines” and that the reviewer clinical guidelines from the American 6 Academy of Child and Adolescent Psychiatry. (Id.) The findings concluded by indicating that the 7 “alternate recommended level of care” was “Mental Health Outpatient Services.” (Id.) 8 4. November 2015 Request and Independent External Reviews 9 On November 16, 2015, plaintiffs submitted a request for an independent external review. 10 (AR 973.)15 Plaintiffs’ request letter detailed concerns with UHC’s July 2015 review of the claim, 11 including that the reviewer did not have expertise in Reactive Attachment Disorder and the denial 12 relied on Andrew’s “medical stability” and “lack of imminent harm to self or others”—both of 13 which are basic criteria for any level of care less intensive than inpatient hospitalization. (AR 974.) 14 The letter further detailed Andrew’s symptoms and treatment history before and after CALO and 15 provided treatment recommendations from the psychological assessment of Andrew undertaken by 16 17 Dr. Kevin O’Keefe. (AR 974, 976, 978-985.) It also summarized the medical records from CALO, 18 listing the various interventions Andrew’s care had required. (AR 985-996.) 19 Further, and for the first time in the review process, the request letter provided a copy of the 20 detailed evaluation done by Dr. Penelope Russell in connection with the court proceedings arising 21 from his juvenile detention prior to Andrew’s admission to CALO. (AR 1245-1261.) Dr. Russell 22 opined that Andrew would benefit from individual and group therapy. (AR 1260.) In assessing the 23 appropriate treatment setting, she noted that “removal from the home appears contraindicated for a 24 15 The Court notes that UBH’s initial response to the November 16, 2015 request for an 25 external review was not sent until April 1, 2016, though the request letter had been received by UBH on February 16, 2016. (AR 1005.) The record does not indicate why there was such a 26 significant lag between the request and UBH’s agreement to forward the request to the independent external review step, other than that UBH requested Andrew’s parents complete a release of 27 information. (AR 999.) Further, based on the record and the procedural history of this case, UBH’s representation that it was forwarding the complete request to the outside reviewer 28 apparently was in error. (See AR 3068.) 1 child with an attachment disorder” and recommended that he receive treatment and support services 2 in his home. She further noted that Andrew’s parents were unwilling to take him home because 3 they do not feel safe, but that “Andrew’s charges do not qualify him for out-of-home placement 4 through the Court.” (AR 1260.) 5 a. First External Review 6 On May 10, 2016, the independent external reviewer issued a decision denying benefits. 7 (AR 1007-12.) The review decision stated, in part: 8 In this case, the patient had chronic and ongoing difficulties with mood lability, 9 lack of self control, interpersonal conflict, and temper outbursts. However, these difficulties were of a mild to moderate severity, and did not warrant long-term 10 treatment in a 24 hour setting. It is noted that there was one episode of greater concern on 4/25/14, when the patient took his sister for a walk in the woods 11 without permission. When the staff tried to contain the patient, he responded by attempting to lash out physically and was placed in a personal hold. Although 12 concerning, it is noted that this level of severity was isolated without any consistency of other severe problems. . . . Overall, the patient did not require 24 13 hour treatment, and therefore, the plan LOC Guidelines were not met. 14 (AR 1010.) The external review decision indicated that it was based only on AACAP principles 15 and the Optum 2015 LOC Guidelines. (AR 1011-12.) The review did not indicate that it relied on 16 the Optum ODD Guideline, or that the Optum ODD Guideline was provided to the external 17 reviewer as part of the review. (AR 1009.) 18 b. Second External Review 19 Due to a UHC error that caused some of the medical records not to be transmitted to the 20 first external reviewer, a second external independent review was initiated on May 2, 2019. (AR 21 3068, 3076.) The decision stated: 22 Per review of the available clinical information, I find that benefit coverage was 23 not available for treatment at the mental health residential level of care from 3/2/14 forward. Your child was not reported to be at risk of harm to himself or 24 others. Though he continued to exhibit intermittent impulsivity and emotional 25 problems, there was no indication that he required continued stabilization in an [sic] residential setting. He was medically stable. There was no reasonable 26 expectation for improvement of his behavior with further 24 hour treatment, nor was there significant change in his behavior as treatment continued. It appeared 27 that your child was at a baseline level of functioning and in need of custodial care, which is not a benefit under his insurance plan. 28 (AR 3076.) The decision’s Patient Clinical History (Summary) provides a lengthier statement of 1 Andrew’s history and his treatment at CALO. (AR 3078-82.) The reviewer noted that Andrew 2 “had a history of doing therapy, but started to refuse to go in May 2013 and then stopped going in 3 November 2013,” just prior to the December 2013 altercation with his parents that precipitated his 4 juvenile detention and transfer to CALO. 5 With respect to his treatment at CALO, the clinical summary provides a chronology of 6 Andrew’s aggressive or violent incidents, attempts to run away or “elope,” and other inappropriate 7 behavior, while indicating throughout that Andrew was “going to treatment.” (AR 3078-3082.) All 8 told, the decision enumerates some seventeen episodes of violent and aggressive episodes or 9 attempts to run away, averaging roughly one episode every two weeks in Andrew’s first four 10 months at CALO, tapering off to one or fewer incidents in the next seven months, and escalating 11 again in the last month of his treatment there, at which point he was “emergently placed” in another 12 program. (Id.) 13 Relying on the general LOC Guidelines, the reviewer found that benefits were properly 14 denied because “[t]his level of care was in excess of the patient’s needs.” (AR 3082). Despite 15 detailing multiple incidents of choking, punching, and physical altercations with peers that required 16 17 physical holds, the decision concluded that Andrew was “not in imminent or current risk of harm to 18 self, others and/or property.” (AR 3083.) While finding that there was a “reasonable expectation 19 that services will improve the member's presenting problems within a reasonable period of time” 20 and that CALO was providing services “within the scope of the provider's professional training and 21 licensure,” the second external review concluded that Andrew “could have been treated safely and 22 effectively at a lower level of care.” (AR 3082.) The reviewer found that “[t]he requested level of 23 care was for the convenience of the family and the patient to give him a safe and structured 24 environment and treatment could have been addressed at a lower level of care.” (AR 3083.) 25 Again, this external reviewer was not provided with the Optum Guideline for Oppositional Defiant 26 Disorder on which some of UHC’s internal reviewers had relied. (AR 3078.) 27 // 28 // 1 III. ANALYSIS AND FINDINGS 2 The Court finds that the preponderance of the evidence in the administrative record 3 demonstrates plaintiff was entitled to coverage under the Plan. Based upon a thorough review of 4 the record, the Court concludes that Andrew met the criteria for, and was provided, residential 5 treatment at CALO, a covered benefit under the Plan. 6 A. The Plan Incorporates the Guidelines 7 As a preliminary matter, plaintiffs argue that UHC relied on improper standards that are not 8 part of the Plan and inconsistent with generally accepted standards of care, namely UBH’s Optum 9 Guidelines. Plaintiffs argue that the Court should find that the Guidelines used by UHC are 10 improper and fall below generally accepted standards of care in the medical community, and that 11 the Plan’s denial of benefits should be overturned on this basis. 12 Several recent decisions have found that the Optum Guidelines are not consistent with any 13 generally accepted standards of medical practice, including a sweeping, comprehensive review by 14 the court in the Wit class action challenging the Optum Guidelines. See Wit v. United Behavioral 15 Health, No. 14- CV-02346-JCS, 2019 WL 1033730 (N.D. Cal. Mar. 5, 2019); see also S.B. v. 16 17 Oxford Health Ins., Inc., No. 3:17-CV-1485 (MPS), 2019 WL 5726901, at *12-13 (D. Conn. Nov. 18 5, 2019); Bain v. Oxford Health Ins. Inc., No. 15-CV-03305-EMC, 2020 WL 808236, at *10 (N.D. 19 Cal. Feb. 14, 2020). Following a ten-day bench trial that included testimony from several mental 20 health experts, the Wit court found under the more deferential abuse of discretion standard that the 21 Optum Guidelines were inconsistent with “generally accepted standards of medical practice” in the 22 respective plans because they are focused on managing acuity rather than providing effective 23 treatment. Wit, 2019 WL 1033730 at 14-17, 55; see also L.B. ex rel. Brock v. United Behavioral 24 Health Wells Fargo & Co. Health Plan, 47 F.Supp.3d 349, 360 (W.D.N.C. 2014) (court found 25 denial of benefits unreasonable, noting “unprincipled and unreasonable claims review by UBH in 26 applying these [Optum] Guidelines does not appear to be isolated,” citing Pacific Shores Hosp. v. 27 United Behavioral Health, 764 F.3d 1030 (9th Cir.2014)). 28 Defendants argue that the Court cannot rely on Wit for several reasons. First, they argue 1 that the court’s findings there are hearsay and inadmissible. Further, they contend the decision in 2 Wit has no preclusive effect here, that is, the decision in Wit is not final and is not identical to the 3 issues and parties here. The Court agrees that the findings in Wit are not evidence, nor have 4 plaintiffs established that a final judgment between the same parties or their privies has been 5 entered there. See New Hampshire v. Maine, 532 U.S. 742, 748–49, 121 S. Ct. 1808, 1814, 149 L. 6 Ed. 2d 968 (2001) (issue and claim preclusion require a prior judgment and the same parties or 7 parties in privity); Syverson v. Int'l Bus. Machines Corp., 472 F.3d 1072, 1078 (9th Cir. 2007) 8 (offensive nonmutual issue preclusion appropriate only if there was identity of issues that were 9 actually litigated and decided in a final judgment, and party against whom issue preclusion is 10 asserted was a party or in privity with a party to the prior action). 11 Further, the Court does not find it appropriate to reach the question of whether the Optum 12 Guidelines are inconsistent with generally accepted medical standards. The terms of the Plan here, 13 as part of the description what services are excluded from coverage, specifically incorporates the 14 Mental Health Administrator’s “level of care guidelines or best practices” as a criterion for denying 15 benefits. (AR 63-64.) Unlike the court in Wit, this Court does not have before it the kind of expert 16 17 testimony and other evidence necessary to decide whether these Guidelines, in the context of this 18 Plan, are improper. More importantly, however, the Court need not reach the question of whether 19 UHC used the Guidelines improperly to deny benefits since the Court finds Andrew was entitled to 20 benefits for the period in dispute even under those Guidelines. 21 B. Andrew Met the Criteria for Residential Treatment Under the ODD Guideline 22 While defendants variously refer to both the Optum LOC Guidelines and the Optum ODD 23 Guideline as the appropriate standards for determining entitlement to benefits, it is the ODD 24 Guideline that, on its face, sets forth UHC’s more specific “understanding of current best practices 25 in care” and factors for determining the appropriate level of care with to individual whose chief 26 diagnosis is Oppositional Defiant Disorder, like Andrew. Thus, the Court looks primarily to the 27 ODD Guideline in determining whether plaintiffs are entitled to benefits for Andrew’s care for the 28 disputed period at CALO. 1 The ODD Guideline requires, at a minimum, that the member not be at imminent risk of 2 serious harm to self or others in order to be admitted to any of the treatment levels described 3 therein: outpatient, intensive outpatient, partial hospital/day treatment, or residential treatment. (AR 4 182-85.) The ODD Guideline states that inpatient residential treatment is appropriate when a 5 member is “experiencing a disturbance in mood, affect or cognition resulting in behavior that 6 cannot be safely managed in a less restrictive setting.” (AR 185.) It defines a residential treatment 7 center as a program “that provides overnight services to members who do not require 24-hour 8 nursing care and monitoring offered in an acute inpatient setting but who do require 24-hour 9 structure.” (AR 188, emphasis supplied.) The preponderance of the evidence in the record here 10 establishes that Andrew experienced a disturbance in his mood, affect, or cognition which resulted 11 in aggressive, impulsive behavior that could not be managed safely at home, and that he required 12 the structure of a residential treatment center to engage in therapeutic interventions to treat this 13 disturbance. 14 1. Opinions of Treating and Examining Mental Health Professionals 15 Here, the vast majority of the mental health professionals who actually examined or treated 16 Andrew found that his symptoms required inpatient residential treatment. Andrew’s regular 17 therapist, who had been treating him and his family since August of 2012, opined that the 18 escalation of Andrew’s rage in his interactions with his family made the dynamic at home unsafe, 19 and that Andrew needed residential treatment to gain greater self-control before he returned home. 20 (AR 914.) Dr. Montgomery, his psychiatrist, recommended residential treatment again based on 21 Andrew’s aggression and lack of insight into his behavior. (AR 913.) Likewise, the opinions of 22 the professionals who treated Andrew at CALO, and who evaluated him after his stay at there, 23 found that he needed residential treatment to reduce his reactivity and develop the tools to manage 24 his rage in a healthier way. (AR 930, 957.)16 25 16 The Court further notes that, in his March 2015 evaluation, Andrew reported to Dr. 26 O’Keefe that “[b]efore I went to an RTC [residential treatment center], I couldn’t control my anger” and would refuse to attend outpatient therapy sessions or “not pay attention” when he was 27 there. (AR 947-48.) Andrew told Dr. O’Keefe that he “learned at [CALO] to control [his anger] better” and he believed CALO was helpful in learning to regulate himself and communicate with 28 his parents. (Id.) 1 Only Dr. Russell, the psychologist who performed the evaluation for the juvenile court, 2 recommended against residential treatment. Dr. Russell’s opinion about residential treatment was 3 based upon two factors: (1) Andrew’s co-existing attachment disorder would make separation from 4 his parents traumatic; and (2) the juvenile court would not order out-of-home placement given the 5 relatively low severity of the charges against him. (AR 1259-60.) However, Dr. Russell 6 acknowledged that Andrew had demonstrated a “moderate to high” potential for violence toward 7 others when he lost control of his emotions and expressed concerns that his parents felt unsafe with 8 him at home. (AR 1258-59.) Moreover, none of UHC’s internal reviewers considered Dr. 9 Russell’s evaluation in reaching their decisions to deny benefits, since her evaluation was not 10 provided to UHC until the second external review.17 11 In sum, these opinions all support a conclusion that Andrew’s condition at the time of his 12 treatment at CALO resulted “in behavior that cannot be safely managed in a less restrictive setting” 13 as set forth in the ODD criteria for residential treatment. Courts generally give greater weight to 14 doctors who have examined the claimant versus those who only review the file. Holmgren v. Sun 15 Life & Health Ins. Co., 354 F.Supp.3d 1018, 1030 (N.D. Cal. 2018) (citing Salomaa v. Honda Long 16 Term Disability Plan, 642 F.3d 666, 676 (9th Cir. 2011); Heinrich v. Prudential Ins. Co. of Am., 17 No. C 04-02943 JF, 2005 WL 1868179, at *8 (N.D. Cal. July 29, 2005); Cooper v. Life Ins. Co. of 18 N. Am., 486 F.3d 157, 167 (6th Cir. 2007)). 19 Moreover, the records of Andrew’s treatment at CALO show that, while he improved 20 somewhat over the course of his treatment there, he continued to experience episodes of aggression 21 and volatility that he was unable to control, supporting the conclusion that he continued to need this 22 level of care throughout the treatment period. UHC either ignored this evidence or failed to explain 23 how it could reach its conclusion in light thereof. 24 2. UHC’s Reviewers’ Opinions Are Entitled to Little Weight 25 UHC’s internal and external reviewers all determined that Andrew was not entitled to 26

27 17 Because Dr. Russell’s evaluation was not part of the records reviewed by any of UHC’s internal reviewers, defendants’ heavy reliance on her certain statements in that evaluation as 28 support for UHC’s denial is unfounded. 1 benefits for the treatment period at issue here because the Plan did not cover “custodial care.” For 2 the reasons stated herein, the Court finds UHC’s reviewers’ opinions that Andrew “was in need of 3 custodial care” are entitled to very little weight.18 4 While the Court is not required to give any particular weight to the opinions of medical 5 professionals who treated or personally evaluated the claimant, neither should it give deciding 6 weight to the opinions of a plan’s reviewers who “arbitrarily refuse to credit a claimant’s reliable 7 evidence, including the opinions of a treating physician.” Black & Decker Disability Plan v. Nord, 8 538 U.S. 822, 834 (2003). Here, although UHC’s internal reviewers purport to have considered the 9 medical records, therapy notes, and treatment plans from CALO, their decisions only minimally 10 acknowledge the contents of those records. CALO’s records indicate that, after his first 30 days in 11 the program, Andrew consistently participated in individual, family, and group therapy, conducted 12 by a licensed professional, multiple times per week. The records show a decrease in frequency of 13 Andrew’s aggressive or violent conduct and an increase in his engagement in therapy, none of 14 which the internal reviewers acknowledge. The record does not support a conclusion that 15 Andrew’s condition was unchanging and not likely to improve with treatment. 16 17 For instance, Dr. Uy’s review found that Andrew “would be most appropriate for a 18 custodial level of care,” which he described as a “protected, controlled environment for the primary 19 purpose of assuring the patient’s safety” provided to “patients for whom there is little expectation 20 of improvement in spite of multiple repeated treatment attempts and/or patients who have been re 21

22 18 UHC’s reviewers never addressed the recommendation of Andrew’s his long-time therapist Robert Brennan or his psychiatrist Dr. Montgomery. Similarly, UHC did not address or 23 distinguish the opinions of the treating therapist at CALO, or the post-treatment opinion of examining psychologist Dr. Kevin O’Keefe. Having failed to do so, UHC arguably waived its 24 ability to challenge those opinions in these proceedings. Cf. Harlick v. Blue Shield of California, 686 F.3d 699, 720 (9th Cir. 2012) (ERISA undermined “’where plan administrators have available 25 sufficient information to assert a basis for denial of benefits, but choose to hold that basis in reserve rather than communicate it to the beneficiary.’” (quoting Glista v. Unum Life Ins. Co. of Am., 378 26 F.3d 113, 129 (1st Cir. 2004)); Mitchell v. CB Richard Ellis Long Term Disability Plan, 611 F.3d 1192, 1199 (9th Cir. 2010) (same); Nieves v. Prudential Ins. Co. of Am., 233 F. Supp. 3d 755, 764 27 (D. Ariz. 2017) (administrator did not raise argument during the appeal process, thereby “forfeited its ability to assert that defense in this litigation.”). However, given that UHC offers no substantive 28 basis to discount those opinions, the Court need not rest its decision on waiver. 1 repeatedly non-adherent with treatment recommendations and have therefore demonstrated the 2 inability to function outside of a controlled environment.” (AR 776, emphasis supplied.) Dr. Uy 3 cites to nothing in the records to explain that opinion.19 To the contrary, the records of Andrew’s 4 treatment at CALO demonstrate progressive improvement in his impulsivity and aggression up 5 until he declined in the month prior to his discharge. Indeed, Dr. Uy’s opinion internally 6 contradicts itself since, in that same evaluation, he recommends “Outpatient Services” (AR 777), a 7 level of treatment which would only be appropriate if he determined that Andrew “exhibits 8 adequate behavioral control to be treated in this setting.” (AR 182-83.) 9 UHC’s second reviewer, Dr. Sane, approved Andrew’s first 30 days at CALO as necessary 10 “acute stabilization of aggressive behavior [and] impulsivity.” However, for the period thereafter 11 she opined that benefits should be denied. “Though he continued to exhibit intermittent impulsivity 12 and emotional reactivity, there was no indication that he required continued stabilization in a 13 residential setting.” (AR 1091.) Dr. Sane concluded that Andrew was “at a baseline level of 14 functioning and in need of custodial care,” with “no reasonable expectation for improvement with 15 further 24-hour treatment” and “no significant change in his behavior.” (Id. at 1091-92, emphasis 16 17 supplied.) 18 Dr. Sane’s conclusion that Andrew needed custodial care is inconsistent with the record and 19 the ODD Guideline. First, Dr. Sane does not support her statement—in a review conducted after he 20 completed treatment at CALO—that Andrew had “no reasonable expectation for improvement” 21 with continued treatment there other than to make bald, circular conclusions. Indeed, Dr. Sane 22 implicitly acknowledged that Andrew did improve after the first 30 days at CALO when she 23 concluded his impulsivity and emotional reactivity were no longer as acute as they had been at his 24 admission. Moreover, the treatment records themselves do not support a conclusion that Andrew 25 showed no change and no improvement during the remainder of his treatment there, since they 26

27 19 The Plan requires “an explanation of the scientific or clinical judgment for the determination, applying the terms of the Plan to your medical circumstances” when benefits are 28 denied. (AR 72.) 1 showed improvement in his emotional regulation and a decrease in aggressive episodes over that 2 whole period up until the last month of treatment. 3 Second, Dr. Sane’s opinion, like other UHC reviewers, noted that Andrew was: (1) 4 “medically stable” and (2) did not show an imminent risk of serious harm to self or others. (See 5 AR 1091.) Those findings offer no justification for the opinion that Andrew nevertheless needed 6 “custodial care.” Further, these two factors are prerequisites for all levels of care other than 7 inpatient hospitalization, including residential treatment. (AR 183-185.) Even assuming her “no 8 imminent risk of harm” opinion was consistent with the treatment records at CALO, it would not 9 explain denying coverage for residential treatment on the basis that Andrew required “custodial 10 care.” 11 Third, Dr. Sane’s decision is not consistent with the ODD Guideline since she did not 12 explain why the factors that had supported Andrew’s admission could be “safely, efficiently, and 13 effectively managed in a less intensive setting” after the initial 30 days. (AR 150-51.)20 The ODD 14 Guideline counsels that the “choice of the most appropriate treatment setting should take into 15 consideration” whether the level of care is “structured and intensive enough to safely and 16 17 adequately treat the member’s presenting problem and support the member’s recovery/resiliency.” 18 (AR 182, emphasis supplied.) The ODD Guideline states that a residential treatment program is 19 appropriate when the individual’s symptoms “cannot be safely managed in a less restrictive setting 20 and [there is] imminent risk that the individual’s psychosocial stressors will impair his ability to 21 function at a lower level of care.” (AR 185.) All levels of care below residential treatment—from 22 outpatient to partial hospitalization—require that the individual be able to comply the program 23 requirements in light of the lower level of structure and supervision they provide. Dr. Sane’s 24 opinion does not explain why Andrew would be appropriate for the unstructured setting she 25 recommended, outpatient treatment, in light of his treatment history. The medical records and 26 20 Dr. Sane’s opinion, unlike Dr. Uy’s, does not reference or state that it utilized the ODD 27 Guideline. (AR 1092; see also AR 1051 [internal case notes].) It did, however, reference American Academy of Child and Adolescent Psychiatry Practice Parameters for ODD, which are 28 not part of the record here. (AR 1092.) 1 treating/examining professionals’ opinions all show that, in a less structured environment, Andrew 2 had been inconsistent in treatment participation and unable to regulate his behavior to keep himself 3 and his parents safe.21 4 5 The external reviewers, while setting forth lengthier explanations for their decisions, quite 6 obviously ignored or disregarded parts of the CALO records. The first external reviewer’s decision 7 noted Andrew’s chronic difficulties in self-control and mood lability but described them as being 8 “of a mild to moderate severity.” (AR 1010.) The reviewer noted “one episode of greater concern” 9 in which Andrew tried to walk out of the program and physically lashed out when staff tried to 10 contain him. (AR 1010.) The review characterized this episode as “isolated without any 11 consistency of other severe problems.” (AR 1010.) However, the records themselves and the 12 summary provided by Andrew’s CALO therapist, Mr. Odell, indicate that Andrew repeatedly 13 engaged in physically aggressive or inappropriate behavior requiring staff intervention during his 14 time at CALO, more frequently at the beginning of his treatment and continuing to a lesser extent 15 in later months. (AR 931-41.)22 16 17 The second external review recited the entire chronology of Andrew’s aggressive episodes 18 at CALO that his therapist had provided. (Compare AR 931-41 with AR 3078-3081.) 19 Nevertheless, the second external reviewer concluded that these episodes were “intermittent” and 20 not an indication that Andrew needed residential treatment. (AR 3084.) The second reviewer 21 found treatment at CALO was “custodial to give him a safe and structured environment while 22 seeking treatment.” (Id.) Again, this review discounted the CALO treatment records without 23

21 Moreover, the ODD Guideline requires consideration of whether improvement can be 24 expected at the level of care “within a reasonable period of time.” (AR 182, emphasis supplied.) 25 Dr. Sane acknowledged that the impulsivity and aggressive behavior Andrew showed in the first 30 days of treatment at CALO continued “intermittently” over the next months. She never explained 26 why 30 days was a “reasonable period of time” to expect improvement or determine whether residential treatment would be ineffective. 27 22 Given that the first external reviewer did not consider all the records in reaching a conclusion, the Court gives that opinion only minimal consideration in this de novo review. (See 28

AR 3068, 3076.)

1 explanation. Further, the reviewer equated a “safe, structured environment for treatment” with 2 custodial care, although that description applies precisely to residential treatment as defined in the 3 ODD Guideline. (AR 185, 188.)23 Finally, this reviewer did not explain how, given Andrew’s 4 treatment history, a less structured program would provide treatment that was at least as safe and 5 effective as a residential treatment program. 6 3. Care Provided by the CALO Program Was Not Excluded from 7 Coverage As Custodial 8 In their briefing, defendants argue that the real question before the Court is whether the care 9 Andrew received at CALO was custodial and therefore not covered by the Plan. (See Cross-Motion 10 at 2, 3.) They contend that Andrew’s weekly individual and family therapy sessions and his weekly 11 group therapy sessions were all conducted by a “licensed professional counselor” and “the vast 12 majority of Andrew’s interactions with staff members at CALO were with staff members who were 13 not licensed mental health professionals.” (Cross-Motion at 16.) They argue that no psychiatrist 14 was “in any meaningful way supervising or adjusting Andrew’s treatment plan,” but only met 15 regularly with Andrew for medication management. (Id.) 16 None of these arguments was the basis the Plan provided for its denial of benefits. The 17 Plan’s reviewers found that Andrew was “in need of custodial care,” not that the program offered 18 by CALO was not covered or was merely custodial. Custodial Care is not defined in the Plan itself. 19 The general 2015 LOC Guidelines describe custodial care as assistance in activities of daily living, 20 such as feeding, dressing, or bathing; services that are not required to be performed by trained 21 personnel; or activities that are done for the purpose of meeting the member’s personal needs, or for 22 maintaining the member’s functioning. (AR 167-68.)24 The ODD Guideline does not mention nor 23 define custodial care. 24 25 23 Neither the first or second external reviewer addressed or considered the ODD Guideline 26 since UHC apparently did not provide it to those reviewers. (AR 1009, 3078.) This is further reason for the Court to give little weight to their opinions. 27 24 The 2014 LOC Guidelines similarly characterize custodial care as care provided by someone other than trained medical personnel, or that does not seek to cure or stabilize the 28 member’s condition. 1 Defendants’ denial letters said absolutely nothing about the professional qualifications of 2 the staff providing Andrew’s treatment or that the program itself was not covered. Nothing in the 3 record suggests that CALO was not a licensed residential treatment center. The denial letters do 4 not state that CALO’s program or treatment providers did not meet the requirements for a covered 5 facility. (Cf. AR 51-52 [Plan provides that coverage for inpatient treatment requires that the facility 6 be licensed as a residential treatment center or meet a list of identified criteria].) Likewise, the 7 denials provide no explanation of how a program that provided Andrew with therapy (individual, 8 family, and group) multiple times per week, as well as a structured environment and interventions 9 to address his aggression, impulsivity, and difficulty regulating his emotional responses would be 10 considered “custodial” care.25 To the contrary, the second external reviewer expressly stated the 11 requirements that “[s]ervices are within the scope of the provider’s professional training and 12 licensure” and that there was “a reasonable expectation that services will improve the member’s 13 presenting problems within a reasonable period of time” were both met. (AR 3082.) The record 14 provides no support for the contention that the care CALO provided Andrew was merely custodial 15 under the Plan or the LOC Guidelines’ definition. 16 17 In short, the contention that UHC properly denied benefits because CALO only provided 18 custodial care is without merit.26 19 // 20 // 21

22 25 The Court finds particularly puzzling defendants’ argument that the CALO program was a non-covered, custodial program considering that UHC approved Andrew’s treatment there for his 23 first 30 days of treatment (finding that it provided “acute stabilization”) when Andrew was not fully participating in therapy sessions other program requirements, but then denied coverage for the 24 treatment period thereafter when Andrew actually began participating fully the program. 25 26 Defendants offer the specious argument that none of Andrew’s providers ever opined his care at CALO was not custodial. (Cross-Motion at 15.) Andrew’s providers were not asked to or 26 required to provide what would essentially be a Plan interpretation—that the treatment at CALO did not amount to a “residential treatment program” but was instead “custodial.” The opinions of 27 the professionals who treated and evaluated Andrew described his treatment and provided their recommendations for further treatment. None of them recommended that Andrew be provided 28 custodial care rather than residential treatment. 1 |]IV. | CONCLUSION AND DISPOSITION 2 Upon de novo review of the record, the Court finds that plaintiff was entitled to coverage for 3 || the residential treatment provided from February 28, 2014 to February 12, 2015 in the CALO 4 program. Plaintiff’s motion for judgment is GRANTED and defendant’s cross-motion is DENIED. 5 The parties shall, within thirty (30) days of the date of this Order: (1) meet and confer to 6 regarding proceedings on the remaining claim, and (2) submit a proposed schedule. 7 This terminates Docket Nos. 70 and 71. 8 IT Is SO ORDERED. ? || Date: July 27, 2020 10 YVONNE GONZALEZ ROGERS

UNITED STATES DISTRICT COURT JUDGE

11 12 8

5 14

2 5 16 a 17 o Ss

E18

Zz 19 20 21 22 23 24 25 26 27 28

29