AMENDED PERMANENT INJUNCTION ORDER

JOAN A. LENARD, District Judge.

THIS CAUSE is before the Court following a four-day bench trial. Plaintiffs K.G., I.D., and C.C., by and through their next friends, Iliana Garrido, Nilda Rivera, and Rachelle Crawford, respectively, brought suit against Defendant Elizabeth Dudek, in her official capacity as Secretary, Florida Agency for Health Care Administration ("AHCA" or "the Agency"). Plaintiffs seek, inter alia, a permanent injunction ensuring that Medicaid-eligible minors under the age of 21 in Florida who have been diagnosed with autism or Autism Spectrum Disorder ("ASD") receive Medicaid coverage for medically necessary behavioral health services, including Applied Behavioral Analysis ("ABA").1 (See First Am. Compl., D.E. 63, at 11-12.) Florida's Medicaid program does not cover ABA or any community behavioral health services to treat autism or ASD. See FLA. ADMIN. CODE ANN. r. 59G-4.050 (incorporating by reference The Florida Medicaid Community Behavioral Health Services Coverage and Limitations Handbook); see also The Florida Medicaid Community Behavioral Health Services Coverage and Limitations Handbook Rule 2-1-4, D.E. 63-1, at Exhibit A (stating that "Medicaid does not pay for community behavioral health services for treatment of autism").2 Plaintiffs argue that Defendant's rule regarding behavioral health services violates the federal Medicaid Act that requires certain services be provided to Medicaid recipients under the age of 21 and violates the federal Medicaid Act's comparability requirement. Defendant argues that the Medicaid Act does not require states to cover ABA because ABA is not a "rehabilitative service[] ... for the maximum reduction of physical or mental disability and restoration of an individual to the best possible functional level." See 42 U.S.C. § 1396d(a)(13). Defendant further argues that even if ABA could be covered under § 1396d(a)(13), the State of Florida still is not required to cover ABA under its Medicaid plan because AHCA has reasonably determined that ABA is experimental and not medically necessary for children with autism and ASD. Finally, Defendant argues that the process AHCA used to determine that ABA is experimental was not arbitrary and capricious.

Plaintiff K.G. filed his Complaint on February 28, 2011. (D.E. 1.) On March 10, 2011, Plaintiff K.G. filed his First Amended Motion for Preliminary Injunction. (D.E. 10.) Magistrate Judge John J. O'Sullivan held a hearing on the motion July 18, 2011 (see D.E. 36, 44, 57) and issued a report recommending the Court grant the motion (D.E. 45). On November 1, 2011, the Court adopted the Report and Recommendation of Magistrate Judge O'Sullivan and directed Defendant to "provide Medicaid coverage for K.G.'s ABA therapy as prescribed by his treating physician." (Preliminary Injunction Order, D.E. 74.) The Court denied Defendant's motion for reconsideration (D.E. 82) of the Court's Preliminary Injunction Order on February 10, 2012 (D.E. 97).

On September 12, 2011, Plaintiff K.G. filed his First Amended Complaint, in which I.D. and C.C. were added as Plaintiffs. (D.E. 63.) Because I.D. and C.C. were added as Plaintiffs after Plaintiff K.G. filed his motion for a preliminary injunction and the Magistrate Judge issued his Report and Recommendation on Plaintiff K.G.'s motion, Plaintiffs I.D. and C.C. were not included in the Court's Preliminary Injunction Order.

On October 21, 2011, the Parties cross-moved for summary judgment. (D.E. 70, 72.) The Court denied the Parties' motions on March 7, 2012, concluding that "because the question of whether the Agency was reasonable in its determination that ABA is experimental is a fact-intensive inquiry, and because material issues of fact exist regarding the efficacy of ABA and whether ABA is medically necessary, summary judgment must be denied." (D.E. 105.)

The Court held a bench trial from March 20 through March 23, 2012. At trial, Plaintiffs called the following fact witnesses: Iliana Garrido, Plaintiff K.G.'s mother; Alexander Lorenzo, ABA provider to Plaintiff K.G.; and Dr. Roberto Lopez Alberola, Division Chief of Pediatric Neurology at Jackson Memorial Hospital in Miami, Florida, Associate Professor at the University of Miami Miller School of Medicine, and Plaintiff C.C.'s treating neurologist. Plaintiffs presented the transcripts of the deposition testimony for three additional fact witnesses: Michael Bolin, AHCA administrator in the Bureau of Medicaid Services; Darcy Abbot, AHCA administrator in the Bureau of Medicaid Services; and Karen Chang, Senior Management Analyst Supervisor for AHCA. In addition, Plaintiffs called three expert witnesses: Dr. Elza Vasconcellos, Director of the Autism Clinic, the Headache Center and Co-Director of the Neurogenetics Clinic at the Miami Children's Hospital in Miami, Florida, and the treating neurologist for Plaintiffs K.G. and I.D.; Dr. James Mulick, Professor in both the Department of Psychology and in the Department of Pediatrics at Ohio State University and a psychologist at The Children's Hospital in Columbus, Ohio;3 and Dr. Jon Bailey, Professor Emeritus in the Department of Psychology at Florida State University and former editor of the Journal of Applied Behavior Analysis.

Defendant called Elizabeth Kidder, the AHCA Deputy Secretary for Medicaid, as the agency representative for AHCA. In addition, Defendant presented videotaped deposition testimony of its two expert witnesses: Dr. Robert Moon, the Chief Medical Officer and Deputy Commissioner Health Systems of the Alabama Medicaid Agency, and Dr. Alison Little, Evidence and Policy Advisor for the Center of Evidence-based Policy at Oregon Health and Science University.4

To obtain a permanent injunction, Plaintiffs must show by a preponderance of the evidence the following four factors: "(1) that [they have] prevailed in establishing the violation of the right asserted in his complaint; (2) there is no adequate remedy at law for the violation of this right; (3) irreparable harm will result if the court does not order injunctive relief; and (4) if issued, the injunction would not be adverse to the public interest." Thomas v. Bryant, 614 F.3d 1288, 1317 (11th Cir.2010); eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 391, 126 S.Ct. 1837, 164 L.Edid 641 (2006); Sheely v. MRI Radiology Network, P.A., 505 F.3d 1173, 1182 n. 10 (11th Cir. 2007) (stating that "[w]hether a permanent injunction is appropriate ... turns on whether the plaintiff can establish by a preponderance of the evidence that this form of equitable relief is necessary").

Based on the following facts that were established at trial through testimony and exhibits and facts that were stipulated by the Parties in their Amended Joint Pretrial Stipulation (D.E. 98-1), and based on the following conclusions of law, the Court finds that a permanent injunction shall be issued for the following reasons:5

The State of Florida has chosen to participate in the federal Medicaid program and therefore "must comply with federal statutory and regulatory requirements." Moore ex rel. Moore v. Reese, 637 F.3d 1220, 1232 (11th Cir.2011). Under the federal Medicaid act, states must provide "early and periodic screening, diagnostic, and treatment services" ("EPSDT") for Medicaid-eligible children under the age of 21. 42 U.S.C. § 1396a(a)(10)(A); 42 U.S.C. § 1396d(a)(4)(B).

K.G., I.D., and C.C. are minors under the age of 21 who, due to their disabling conditions and their parents' income levels, are Florida Medicaid recipients. During EPSDT screens,6 K.G., I.D., and C.C. were all diagnosed with autism or Autism Spectrum Disorder ("ASD") and were all prescribed ABA by their treating neurologists. Autistic disorder, the full-blown condition, is one of five disorders falling under the umbrella of Autism Spectrum Disorder.7 Autism is a complex neurodevelopmental disability that generally appears during the first three years of life and impacts the normal development of the brain, resulting in impairments of social interaction, verbal and non-verbal communication, leisure or play activities, and learning.89 ABA is a type of early intensive behavioral interaction ("EIBI") health service that uses a structured one-on-one program to treat the behavioral problems associated with autism and ASD.10

Because Plaintiffs' conditions were discovered during EPSDT screens, the State of Florida is required to provide the children with any service described in 42 U.S.C. § 1396d(a) that is necessary to correct or ameliorate the children's conditions, whether or not the service is covered for adults under the State's Medicaid plan. 42 U.S.C. § 1396d(r)(5)11; Moore, 637 F.3d at 1233-34; Smith v. Benson, 703 F.Supp.2d 1262, 1269 (S.D.Fla.2010); Rosie D. v. Romney, 410 F.Supp.2d 18, 26 (D.Mass.2006). Thus, in determining if AHCA's decision not to cover ABA for children with autism and ASD violates the EPSDT provisions in the Medicaid Act, this statutory framework presents two questions: (1) is ABA among those services which can be covered under 42 U.S.C. § 1396d(a), and (2) is ABA necessary to correct or ameliorate an illness or condition for a Medicaid recipient under age 21?

ABA falls within the scope of 42 U.S.C. § 1396d(a)(13).12 The services under this provision are as follows:

42 U.S.C. § 1396d(a)(13) (emphasis added). Federal regulations define "rehabilitative services" as:

42 C.F.R. § 440.130(d) (emphasis added). Plaintiffs have established beyond all doubt through testimony and exhibits at trial that ABA is necessary to restore children with autism or ASD to the normal trajectory of development that all children have through the first twelve years of life. Dr. Mulick testified as follows:

Dr. Mulick further testified that the purpose of ABA is to restore autistic children to their normal rate of development. Dr. Lopez Alberola testified that ABA restores autistic children's functional capacity to "develop normally and continue in the normal pattern of human development." Dr. Vasconcellos testified that ABA restores children to their normal potential by "restor[ing] them back to the curve of the developmental milestones that they had before the autism started[,] ... similar as what you expect for normal development in childhood." As to the Plaintiffs in this case, Dr. Lopez Alberola, C.C.'s treating neurologist, and Dr. Vasconcellos, K.G. and I.D.'s treating neurologist, clearly established that ABA would restore K.G., I.D., and C.C. to their best possible functional levels. Other courts have interpreted 42 U.S.C. § 1396d(a)(13) to mean that "if a licensed clinician finds a particular service to be medically necessary to help a child improve his or her functional level, this service must be paid for by a state's Medicaid plan pursuant to the EPSDT mandate." Parents League for Effective Autism Servs. v. Jones-Kelley, 565 F.Supp.2d 905, 915 (S.D.Ohio 2008), aff'd, 339 Fed.Appx. 542, 544 (6th Cir.2009); Rosie D., 410 F.Supp.2d at 26; see also Collins v. Hamilton, 349 F.3d 371, 375 (7th Cir.2003); John B. ex rel. L.A. v. Menke, 176 F.Supp.2d 786, 800 (M.D.Tenn.2001); Pediatric Specialty Care, Inc. v. Ark. Dep't of Human Servs., 293 F.3d 472, 480-81 (8th Cir.2002).

Although the EPSDT statutory provision uses the term "necessary," the Medicaid Act does not define the term. See generally 42 U.S.C. § 1396d. Instead, the Medicaid Act and its implementing regulations grant states the authority to set reasonable standards for the terms "necessary" and "medical necessity." See 42 U.S.C. § 1396a(a)(17); 42 C.F.R. § 440.230(d); see also Rush v. Parham, 625 F.2d 1150, 1154-55 (5th Cir.1980).13 States are not required to pay for services that are not medically necessary. See Rush, 625 F.2d at 1155. "`Experimental forms of treatment, i.e., treatment not generally recognized as effective in the medical community'" is "an example of a service the state is not required to pay for because it is not medically necessary." Moore, 637 F.3d at 1248 n. 48 (quoting Rush, 625 F.2d at 1154-55). However, "a `state may not arbitrarily deny or reduce the amount, duration, or scope of a required service ... solely because of the diagnosis, type of illness, or condition.'" Rush, 625 F.2d at 1157 n. 12 (quoting 42 C.F.R. § 440.230(c)(1)).14

ABA is "medically necessary" and is not "experimental" as defined under Florida administrative law and federal law. See FLA. ADMIN. CODE ANN. r. 59G-1.010(166)(a)3 (defining "medical necessity")15 FLA. ADMIN. CODE ANN. r. 59G-1.010(84)(a)3 (defining "experimental" and "reliable evidence")16; Rush, 625 F.2d at 1154-58 (setting forth guidelines for determining whether a treatment is "experimental") 17; Moore, 637 F.3d at 1248 n. 48. The role of the Court is to decide whether AHCA's determination that ABA is experimental was reasonable. See Rush, 625 F.2d at 1157 (remanding to the district court to decide "whether [the state's] determination that transsexual surgery is experimental is reasonable"); see also Miller v. Whitburn, 10 F.3d 1315, 1321 (7th Cir.1993) (stating that "the district court may decide only whether [the state's] determination that the liver-bowel transplant procedure is experimental is reasonable" (citing Rush, 625 F.2d at 1157)).

Based on the testimony and exhibits at trial, the Court finds that the determination by AHCA that ABA is experimental was arbitrary, capricious, and unreasonable both in its process and in its conclusion. Elizabeth Kidder, the AHCA Deputy Secretary for Medicaid, testified that AHCA has a standard process for determining whether a certain treatment or service should be covered under the State's Medicaid plan, and that AHCA did not follow this standard process when determining not to cover ABA. Kidder testified that AHCA's standard practice involves assigning a treatment to an analyst and a registered nurse to research the treatment, and the analyst and nurse review published, peer-reviewed literature on the treatment, as well as whether other entities are covering the service (e.g., other state Medicaid programs,18 Medicare, and commercial insurance plans19). In addition, Michael Bolin, an AHCA employee, testified that as part of its standard practice AHCA "routinely" consults with physicians for a "clinical evaluation of the data." After the analyst and nurse have finished this research, the analyst puts their findings in a memorandum that is reviewed by four to five levels of administrators up the chain of command before being adopted by the Agency. Kidder testified that this standard practice was in place when AHCA decided that ABA was experimental, but ABA did not go through this process. Instead, after Plaintiff K.G. filed the Complaint in this Court on February 28, 2011, Roberta Bradford, the then AHCA Deputy Secretary for Medicaid, determined that ABA was experimental by reviewing the following material: a metaanalysis20 by Spreckley and Boyd published in the Journal of Pediatrics, a metaanalysis by Ospina published in PloS ONE, and the TEC Report. The TEC Report was funded by Blue Cross and Blue Shield Association, a federation of 38 health insurance organizations and companies in the United States, and Kaiser Permanente, a nonprofit health plan service.21 This material was provided to Bradford by AHCA's legal counsel. No analyst and/or nurse in AHCA was assigned to research ABA, no analyst ever reviewed any "published reports and articles in the authoritative medical and scientific literature" about ABA, and no memorandum regarding ABA was ever prepared by an analyst and reviewed by the four to five levels of management in AHCA. See FLA. ADMIN. CODE ANN. r. 59G-1.010(84)(b) (defining "reliable evidence"). No one in AHCA reviewed whether ABA was covered by other states' Medicaid programs, Medicare, or commercial insurance plans. No one in AHCA consulted with a physician about ABA. After reviewing only these three reports, Bradford alone determined that ABA is experimental.

Kidder replaced Bradford in August 2011. The week before her deposition in this case, Kidder asked for the file of materials that Bradford reviewed in making her determination that ABA is experimental. Kidder testified that she wanted to see if she could "replicate [Bradford's] conclusion" by using the materials Bradford used. In addition to the Spreckley and Boyd meta-analysis, Ospina meta-analysis, and TEC Report, Kidder reviewed the Hayes Report22 and the AHRQ Report, also referred to as the Warren study.23 Kidder read only the summary of the Hayes Report, which included the analyses of the three reports previously reviewed by Bradford and read by Kidder, and four additional meta-analyses, which Kidder chose not to read. Kidder admitted that out of the four meta-analyses she chose not to read, the Hayes summary stated that three meta-analyses concluded that ABA results in beneficial improvement in outcomes and suggested that the treatment should be considered a therapy of choice and the other concluded that ABA is an effective therapy for a subpopulation of children with autism. In addition, Kidder read only the executive summary of the AHRQ Report, and she did not read the 100 pages in the AHRQ Report that addressed ABA. Again, AHCA completely failed to follow its own standard procedures. No analyst and/or nurse in AHCA ever reviewed any "reliable evidence"24 about ABA, no one assessed whether ABA was covered by other states' Medicaid programs, Medicare, or commercial insurance, no one consulted with any physician about ABA, and no memorandum regarding ABA was ever prepared by an analyst and reviewed by AHCA's management. Instead, Kidder took on the role of analyst for the purpose of this litigation, and upon a cursory review of these materials, decided that ABA was experimental.

Testimony from Dr. Bailey established that almost all of the materials relied upon by Bradford and Kidder are not "reliable evidence" as defined under Florida law. Under Florida law,

FLA. ADMIN. CODE ANN. r. 59G-1.010(84)(b) (emphasis added). The TEC Report, Hayes Report, and the AHRQ Report are not peer-reviewed and have not been published "in the authoritative medical and scientific literature." See id. The Spreckley and Boyd metaanalysis (which was also used in the Hayes Report) was published in the Journal for Pediatrics, which would ordinarily be considered "reliable evidence." See id. However, Dr. Bailey testified that the Spreckley and Boyd meta-analysis made a clear error in its evaluation of the Sallows and Graupner study. Dr. Bailey cited to a letter to the editors of the Journal of Pediatrics from Smith, Eikeseth, Sallows, and Graupner, and Dr. Bailey testified that the authors stated that Spreckley and Boyd misrepresented the findings in the Sallows and Graupner study. Dr. Bailey further testified that the authors claimed that removing the mischaracterized Sallows and Graupner study from the Spreckley and Boyd meta-analysis would show that ABA yields significant findings on three of the four outcome measures in the meta-analysis. The Hayes Report did not mention this letter to the editor. It was also not included in the materials Bradford reviewed and turned over to Kidder, and Kidder did not review the subsequent discourse published in the Journal of Pediatrics. Dr. Bailey testified that this letter to the editor of the Journal of Pediatrics exemplified the peer-review process, whereby studies get published in the public domain and everyone in the field may examine them and respond.25 Kidder testified that she did not review this letter to the editor. It is reasonable to infer that if AHCA had followed its standard practice for making coverage decisions in this case, an analyst may have found this letter in his or her survey of the "published reports and articles in the authoritative medical and scientific literature." See FLA. ADMIN. CODE.ANN. r. 59G-1.010(84)(b). Dr. Bailey further testified that the Ospina meta-analysis made the same mistake as the Spreckley and Boyd meta-analysis.

In addition, AHCA did not comply with the definition of "experimental" under Florida law when determining that ABA is experimental. Florida's definition of "experimental" states, in relevant part:

FLA. ADMIN. CODE ANN. r. 59G-1.010(84)(a)3 (emphasis added).26 Kidder testified that she did not compare ABA to any other treatment as required by Florida law, that ABA is "the only treatment out there," and that she felt that further studies were necessary to determine ABA's maximum efficacy. Kidder further testified that ABA may even be the standard of care for autism and ASD. In essence, Kidder testified she only applied the first prong of the definition, and never made a comparative analysis because ABA either is the standard of care or there is no standard of care.

Dr. Vasconcellos, Dr. Bailey, and Dr. Mulick all testified that ABA is the standard means of treatment for autism and ASD. Dr. Vasconcellos testified that she prescribes ABA to all of her autistic patients and believes it would be medical malpractice not to prescribe ABA for a child with autism. Dr. Bailey testified that "we know ABA works. It's been well established. It's accepted in the medical community." Furthermore, Dr. Mulick testified that the consensus in the medical community is that ABA has been the standard means of treatment for children with autism and ASD since the 1990's, evidenced by consensus statements from the following sources: Centers for Medicare and Medicaid Services, United States Surgeon General Schachter, the Center for Disease Control and Prevention, the National Institute for Child Health and Human Development, the National Institute for Neurological Disorders and Stroke, the National Institute for Mental Health, the American Society of Child and Adolescent Psychiatry, the American Academy of Pediatrics, and the American Psychological Association. Kidder testified that she did not consider any of these consensus statements when determining the standard means of treatment for autism. In sum, the Court finds that AHCA's failure to follow its own unwritten but formal standard practice for making treatment coverage decisions, failure to apply Florida's definition for "experimental," and failure to use "reliable evidence" as defined by Florida law, was unreasonable, arbitrary, and capricious.27

The "reliable evidence," as defined by Florida law, conclusively shows that ABA is not "experimental." Plaintiffs have established through their expert witnesses that there exists in the medical and scientific literature a plethora of peer-reviewed meta-analyses, studies, and articles that clearly establish ABA is an effective and significant treatment to prevent disability and restore developmental skills to children with autism and ASD. Dr. Bailey testified that the four peer-reviewed meta-analyses listed at the end of the Hayes Report (two by Eldevik, one by Reichow, and one by Virues-Ortega) show that ABA is effective though they received scarce attention in the Hayes summary report. These meta-analyses included findings of large to moderate changes in IQ, intensive ABA intervention leads to positive medium to large effects in terms of intellectual functioning, language development, acquisition of daily living skills, and social functioning in children with autism and ABA is the treatment of choice. Dr. Bailey further testified about three additional studies (by Dawson, Zachor, Smith) that show that ABA is effective. For example, Dr. Bailey testified that the Zachor study is "well-conducted," "well-respected," "highly cited," and "published in a good journal... with peer review," and that the study concluded that the children receiving ABA "showed significantly greater improvements" than those children in the eclectic group. Further, Dr. Bailey stated that AHRQ did not acknowledge the significance of the Zachor study or its findings. Dr. Bailey testified that he is aware of 25 to 30 reviews of the literature and an additional 30 to 50 meta-analyses showing that ABA has been proven effective for children with autism. Notably, all the experts that testified for Plaintiffs and Defendant stated that they have never seen a study in the peer-reviewed literature where the authors concluded that ABA was ineffective as a treatment for children with autism or a study that characterized ABA as experimental.

Based on the foregoing, the Court finds that Plaintiffs have clearly established that Florida's exclusion of ABA for Medicaid-eligible minors diagnosed with autism or ASD violates the broad EPSDT mandate in the federal Medicaid Act. Regarding the remaining factors for a permanent injunction, the Court first finds that there is no adequate remedy at law for the violation of this right. Second, there would be irreparable injury to Plaintiffs and all Medicaid-eligible minors in Florida who have been diagnosed with autism or ASD if these children do not receive ABA pursuant to this injunction. If these children do not receive ABA in the primary years of development up to age 6 and then to 12 years of age, the children may be left with irreversible language and behavioral impairments. As Dr. Lopez Alberola testified:

Dr. Mulick expressed similar concerns, testifying:

K.G. is already 6 years old, C.C. is 5 years old, and I.D. is 3 years old. Dr. Vasconcellos testified that starting ABA at 4 years old is "late." Therefore, it is imperative that autistic children in Florida receive ABA immediately to prevent irreversible harm to these children's health and development.

Finally, regarding the public interest element, "[i]ssuance of an injunction to enforce the federal Medicaid Act is without question in the public interest." Edmonds v. Levine, 417 F.Supp.2d 1323, 1342 (S.D.Fla.2006). Furthermore, the Court finds that paying for the cost of ABA for autistic children will ultimately save public funds. Dr. Mulick testified that autistic children who do not receive ABA become "totally dependent on the community and state and the Federal Government for support. They receive Social Security support, domiciliary care. They often become injured and require ... expensive medical interventions." Therefore, the Court finds Plaintiffs have satisfied by a preponderance of the evidence the four factors necessary to be granted a permanent injunction.

Accordingly, it is hereby ORDERED AND ADJUDGED that:28