Granville House, Inc. v. Department of Health and Human Services, Arthur E. Noot as Commissioner of Public Welfare for the State of Minnesota

HEANEY, Circuit Judge. This case involves a challenge to the interpretation of the phrase “institutions for mental disease” contained in the Medicaid *1294 Act. The district court held that the interpretation adopted by the Department of Health and Human Services (HHS), to include the diseases of chemical dependency and alcoholism within the term “mental diseases,” was unreasonable. We reverse in part and remand to the district court with directions to it to allow the Grant Appeals Board the opportunity to consider and decide this matter, after which any party may seek review in the district court. I. BACKGROUND Granville House, Inc., is a nonprofit corporation that operates three residential chemical dependency treatment programs in Minnesota. Jane Dickman House, established in 1963, serves women aged sixteen and over; Team House, established in 1972, serves men aged seventeen and over; and Warren Eustis House, established in 1980, serves adolescents aged thirteen to eighteen. The three facilities have a total capacity of one hundred six residents. Gran-ville’s primary mission since its founding in 1963 has been to treat poor people who are victims of chemical dependency. Prior to 1980, approximately ninety percent of its residents were indigent clients, whose expenses, since 1972, were paid through public funds. 1 In the mid to late 1970s, Granville made inquiries concerning the availability of federal funding under the Medicaid program, Title XIX of the Social Security Act, 42 U.S.C. §§ 1396 et seq. (1976 & Supp. V 1981). Title XIX provides for the federal government to share with the states the costs of various types of medical assistance, including services provided in intermediate care facilities (ICFs). See 42 U.S.C. § 1396d(a)(15) (1976). 2 Commencing in 1978, Granville invested approximately $300,000 to bring the facilities at Jane Dick-man House and Team House into conformity with the standards for intermediate care facilities. Once its facilities were certified as ICFs, Granville believed its residents would become eligible for Medicaid reimbursement. Approximately $200,000 was spent at Warren Eustis House for a similar purpose. During this time period, however, officials in the Minnesota Department of Public Welfare, the State’s Title XIX agency, advised Granville that its clients would not be eligible for reimbursement under Medicaid because its facilities were institutions for mental disease (IMDs). They indicated that these facilities were IMDs because over fifty percent of the residents were diagnosed as chemically dependent (in fact, all residents have such a diagnosis), a disease which is currently viewed by HHS as a mental disease or disorder. Title XIX provides that no federal funds shall be paid for services provided to an individual who is under sixty-five years of age and who is a patient in an institution for mental disease or tuberculosis. Id; 42 U.S.C. § 1396d(a)(18)(B) (Supp. V 1981). See 42 C.F.R. §§ 435.1008(a)(2), 440.-140(a)(1)(ii), (a)(2) & (c) (1982). In 1972, Congress removed the restriction on federal financial participation for inpatient psychiatric hospital services provided to individuals under age twenty-one. 42 U.S.C. § 1396d(a)(16) (1976 & Supp. V 1981); id § 1396d(h) (1976). See 42 C.F.R. § 440.160 (1982). Thus, the statute currently provides that persons between the ages of twenty-one and sixty-five who are patients in IMDs may not receive Medicaid. 3 *1295 Title XIX also provides that the Secretary of HHS “shall make and publish such rules and regulations, not inconsistent with this chapter, as may be necessary to the efficient administration of the [Medicaid program],” 42 U.S.C. § 1302 (Supp. V 1981), and pursuant to this authority the Secretary has promulgated regulations that define the term “institution for mental diseases.”' The regulations define “institution” as “an establishment that furnishes (in single or multiple facilities) food, shelter, and some treatment or services to four or more persons.” 42 C.F.R. § 435.1009 (1982). An IMD is specifically defined as: an institution that is primarily engaged in providing diagnosis, treatment or care of persons with mental diseases, including medical attention, nursing care and related services. Whether an institution is an institution for mental diseases is determined by its overall character as that of a facility established and maintained primarily for the care and treatment of individuals with mental diseases, whether or not it is licensed as such. Id. HHS has further interpreted this definition of an IMD in a series of internal documents in the Field Staff Information and Instruction Series (FSIIS). These FSIIS documents were issued by the Medical Services Administration of the Social and Rehabilitation Service, the predecessor agency of the Health Care Financing Administration (HCFA), as guidelines for its regional officials. The most important one, for purposes of this case, is FSIIS FY-76-44, which directed the use of the World Health Organization’s International Classification of Diseases (ICD) for determining what constitutes a “mental disease.” FSIIS FY-76-44, dated November 7,1975, specifically states: An institution is characterized as “primarily” one for mental diseases if it is licensed as such, if it advertises as such or if more than 50 percent of the patients are in fact patients with mental disease. * * * Mental diseases are those listed under the heading of mental disorders in the eighth revision, International Classification of Diseases, Adapted for Use in the United States * * * except that mental retardation is not included for this purpose[ 4 ] Alcoholism and other chemical dependencies are listed as mental disorders in the ICD. Thus, as a March 29, 1978, letter from Acting Regional Medicaid Director George R. Holland to a Minnesota official in the Department of Public Welfare states: Under Medicaid chemical dependency is viewed as a form of mental disorder. Therefore, patients with such diagnosis who require skilled or intermediate [ICF] care would be eligible for Federal funding under the Medicaid program unless they are in a facility for mental diseases and their age is under 65 years. This latter restriction does not apply to young patients under 21 years of age in psychiatric facilities accredited by the Joint Commission on Accreditation of Hospitals. [Emphasis added.] See generally Schweiker v. Wilson, 450 U.S. 221, 225 n. 5, 101 S.Ct. 1074, 1078 n. 5, 67 L.Ed.2d 186 (1981). Granville filed this action in federal district court on May 12, 1980, against the Department of Health and Human Services and the Commissioner of the Department of Public Welfare in his official capacity, seeking a declaration that the classification of chemical dependency as a mental disease is arbitrary and capricious and that the treatment of chemical dependency in a certified ICF qualifies for federal financial participation under Medicaid. Granville invoked the district court’s jurisdiction pursuant to 28 U.S.C. § 1331, claiming that the matter arose under the Constitution and Title XIX of the Social Security Act, 42 U.S.C. §§ 1396 et seq. (1976 & Supp. V 1981). The Minnesota Department of Public Welfare (hereinafter State) initially defended the suit, maintaining that its policies *1296 concerning Medicaid reimbursement for chemical dependency treatment were proper, that the district court lacked jurisdiction, and that Granville lacked standing and had failed to exhaust its available administrative remedies. The State later changed its position, however, and in February, 1982, moved to file a cross-claim against HHS and to consolidate the instant ease with a companion case, Minnesota v. Schweiker, Civ. No. 4-82-155 (D.Minn., filed Feb. 26, 1982). 5 In its cross-claim, the State alleged jurisdiction under 28 U.S.C. §§ 1331 & 2201 and sought a declaratory judgment that HHS’s interpretation of the Medicaid statute and regulations as precluding coverage for services provided to individuals in residential facilities serving the chemically dependent is contrary to law and regulations, arbitrary and capricious, and an abuse of discretion. HHS opposed the consolidation and filing of the cross-claim, but after the district court ruled against it, HHS filed an answer denying the allegations in the State’s cross-claim. HHS also denied the allegations in Granville’s complaint, and further asserted that the district court lacked jurisdiction, and that Granville lacked standing. Both Granville and HHS moved for summary judgment, and the district court heard evidence on the question of whether chemical dependency should be classified as a mental disorder. Because the agency’s decision to use the ICD’s classification scheme to further define the term “mental diseases” is not a regulation, but rather is simply an interpretation by the agency of a term contained in a regulation, the decision does not have the force and effect of law. See General Electric Co. v. Gilbert, 429 U.S. 125, 141, 97 S.Ct. 401, 410, 50 L.Ed.2d 343 (1976); 2 K. Davis, Administrative Law Treatise § 7:10 at 52 (1979). See also Schweiker v. Hansen, 450 U.S. 785, 789, 101 S.Ct. 1468, 1471, 67 L.Ed.2d 685 (1981) (Claims Manual of Social Security Administration for internal use by its employees “has no legal force”). Of course, an agency’s interpretation of its own regulation is generally accorded deference by a reviewing court. See United States v. Larionoff, 431 U.S. 864, 872-873, 97 S.Ct. 2150, 2155-2156, 53 L.Ed.2d 48 (1977); Udall v. Tallman, 380 U.S. 1, 16-17, 85 S.Ct. 792, 801, 13 L.Ed.2d 616 (1965); Munsinger v. Schweiker, 709 F.2d 1212 at 1214 (8th Cir.1983), citing Oglala Sioux Tribe of Indians v. Andrus, 603 F.2d 707, 718 (8th Cir.1979). This may be especially true in areas involving complicated statutory schemes such as Medicaid. See generally Schweiker v. Gray Panthers, 453 U.S. 34, 43, 101 S.Ct. 2633, 2639, 69 L.Ed.2d 460 (1981); Abbott-Northwestern Hospital, Inc. v. Schweiker, 698 F.2d 336, 340 (8th Cir.1983). Nevertheless, courts are the final authority on matters of statutory construction, and are not required to rubber-stamp their affirmance of administrative decisions that are inconsistent with the statutory mandate or that frustrate congressional policy underlying the statute. Volkswagenwerk Aktiengesellschaft v. Federal Maritime Commission, 390 U.S. 261, 272, 88 S.Ct. 929, 935, 19 L.Ed.2d 1090 (1968). The weight of an interpretive rule or decision in a particular case “will depend upon the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, if lacking power to control.” Skidmore v. Swift & Co., 323 U.S. 134, 140, 65 S.Ct. 161, 164, 89 L.Ed. 124 *1297 (1944); 2 K. Davis, Administrative Law Treatise § 7:13 at 60 (1979). In this case, the district court recognized during the proceedings before it that HHS had not given adequate attention to the classification of chemical dependency for purposes, of the Medicaid statute. The court noted that HHS had never conducted an evidentiary hearing on the matter, its decision to adopt the ICD was made at the lower levels of the agency, and it had never issued a reasoned analysis in support of its classification decision. The court then found that HHS’s “classification of alcoholism and other forms of chemical dependency as mental disorders is unreasonable,” based on the “great bulk” of the evidence before it that supported the conclusion that chemical dependency “is predominately physical, as opposed to mental, in nature.” Granville House, Inc. v. Department of Health and Human Services, 550 F.Supp. 628, 636, 638 (D.Minn.1982). Specifically, the court held that (1) HHS’s classification of alcoholism and other forms of chemical dependency as mental diseases or disorders is unreasonable; (2) that HHS’s characterization of Granville’s facilities as IMDs is arbitrary and capricious; and (3) that all of Granville’s residents shall be eligible for Medicaid funding at such time as the State certifies its facilities. Id. at 638. The court also determined that it had jurisdiction to entertain the suit, since Granville had standing to assert its claim and the issues presented were ripe for review. Id. at 630-631. HHS appeals, asserting that the district court lacked jurisdiction .and that its classification of chemical dependency as a mental disorder is reasonable. II. JURISDICTION Article III of the United States Constitution limits the judicial power of this and other courts to the resolution of “eases” and “controversies.” The concept of standing is derived from Article III, which demands at a minimum that the party seeking judicial relief have “such a personal stake in the outcome of the controversy as to assure that concrete adverseness which sharpens the presentation of issues upon which the court so largely depends.” Baker v. Carr, 369 U.S. 186, 204, 82 S.Ct. 691, 703, 7 L.Ed.2d 663 (1962). See Havens Realty Corp. v. Coleman, 455 U.S. 363, 378-379, 102 S.Ct. 1114, 1124, 71 L.Ed.2d 214 (1982). A “personal stake” means that a party must have suffered “injury in fact,” which “fairly can be traced to the challenged action” and “is likely to be redressed by a favorable decision.” Valley Forge Christian College v. Americans United for Separation of Church and State, Inc., 454 U.S. 464, 472, 102 S.Ct. 752, 758, 70 L.Ed.2d 700 (1982) (and cases cited therein). As the Supreme Court stated in Sierra Club v. Morton, 405 U.S. 727, 740, 92 S.Ct. 1361, 1368, 31 L.Ed.2d 636 (1972): The requirement that a party seeking review must allege facts showing that he is himself adversely affected * * * does serve as at least a rough attempt to put the decision as to whether review will be sought in the hands of those who have a direct stake in the outcome. The district court found that Gran-ville had suffered injury in fact, in that the denial of Medicaid funding had caused it financial harm and had caused it “to forego treating the poor in favor of individuals of the middle and upper class” whose insurance plans cover chemical dependency treatment. 550 F.Supp. at 631. In addition to economic injury, the court found that Granville was injured by having “to withdraw from its primary mission of treating the poor.” Id. We agree with the district court that this injury is sufficient to provide Granville with standing to challenge HHS’s classification of chemical dependency as a mental disease. Since its inception, Granville’s “primary mission” has been to serve the poor; to help those most in need. It is a nonprofit organization, whose incorporation charter states that it shall treat victims of chemical dependency regardless of ability to pay. Since 1972, Granville has relied upon public funds to support its poor patients. These funds were “drastically cut” from 1978 to the present time, which has decreased the number of poor patients Granville may ac *1298 cept. Prior to the decrease, its facilities operated at full capacity, with long waiting lists. The overwhelming majority — perhaps ninety percent — of the patients were publicly funded from 1972 to 1978. Many of these publicly-funded patients were members of minority groups. Granville’s Executive Director, Dagny Christiansen, testified that as a result of the lack of federal funds, the majority of Granville’s patients were no longer indigent, and very few were minority. Christiansen indicated that some of these indigent individuals may be admitted to general hospitals for treatment, where the per diem charges which are paid through the Medicaid program are substantially larger than those at Granville. See 550 F.Supp. at 638. In fact, the per diem charge for each resident in Granville’s program ranges from $35.00 per day to $52.00 per day, id, while hospital charges may be as high as $300.00 per day. Granville’s facilities offer services that are equal to or superior to those offered in the hospitals, yet the current administration of the Title XIX program encourages placement in the higher-cost general hospitals. Id Granville has sought Title XIX funds, and has challenged the characterization of its facilities as “institutions for mental diseases,” not only because of the financial impact of this decision on its operations, but also to allow it to fulfill its primary purpose of serving the poor and minorities who suffer from chemical dependency. The district court specifically found that “[u]pon the State’s action granting Team House certification as an Intermediate Care Facility and its repeated assertions that in its view all the Granville House facilities were certifiable as such, the Court must conclude that the only obstacle in [Granville’s] path is the federal government’s characterization of chemical dependency as a mental disorder.” Id at 631. Under these circumstances," we believe that Granville has suffered injury in fact sufficient to give it standing to challenge the classification of chemical dependency as a mental disease. HHS’s interpretation of the term mental diseases has perceptibly impaired Granville’s ability to provide its services to indigent patients, just as the realty company’s practices in Havens Realty Corp. v. Coleman, supra, 455 U.S. at 379, 102 S.Ct. at 1124, impaired the ability of the nonprofit corporation in that case to provide its services. In Havens Realty Corp., several individuals and a nonprofit corporation, Housing Opportunities Made Equal (HOME), brought suit against a realty company, seeking declaratory, injunctive, and monetary relief from the company’s racial steering practices. The complaint identified HOME as a nonprofit corporation “whose purpose was ‘to make equal opportunity in housing a reality in the Richmond Metropolitan Area.’ ” Id at 368, 102 S.Ct. at 1118. Because of a settlement that had been entered into by the parties, the Supreme Court did not decide whether HOME had standing to assert its members’ interests, but the Court held that HOME had standing in its own right because it was frustrated in its efforts to assist in equal access through its counseling and referral services. Id at 378-379, 102 S.Ct. at 1124. The Court stated: “If, as broadly alleged, petitioners’ steering practices have perceptibly impaired HOME’S ability to provide counseling and referral services for low- and moderate-income homeseekers, there can be no question that the organization has suffered injury in fact.” Id at 379, 102 S.Ct. at 1124. In this case, as in Havens Realty Corp., the harm to the organization involved “constitutes far more than simply a setback to the organization’s abstract social interests,” id, as was the case in Sierra Club v. Morton, supra, 405 U.S. at 739, 92 S.Ct. at 1368. Indeed, in Sierra Club, the organization specifically declined “to rely on its individualized interest as a basis for standing.” Id at 735 n. 7, 92 S.Ct. at 1366 n. 7. See id at 735-736, 92 S.Ct. at 1366-1367. Cf. Valley Forge Christian College v. Americans for the Separation of Church and State, Inc., supra, 454 U.S. at 476 n. 14, 102 S.Ct. at 761 n. 14 (organization alleged no injury to itself as an organization). *1299 HHS claims that as a provider, Granville fails to satisfy the “zone of interests” test enunciated in Association of Data Processing Service Organizations v. Camp, 397 U.S. 150, 153, 90 S.Ct. 827, 829, 25 L.Ed.2d 184 (1970). Although there is apparently some controversy surrounding the precise status and applicability of this test, see K. Davis, Administrative Law Treatise § 22.02-11 at 347-352 (1982 Supp.), the Supreme Court has recently characterized the test as “prudential” in nature. Valley Forge Christian College v. Americans for the Separation of Church and State, Inc., supra, 454 U.S. at 474-475, 102 S.Ct. at 759-760. HHS correctly notes that other courts have stated that the very similar Medicare statute was enacted to benefit patients, not to benefit or subsidize providers. See, e.g., Dialysis Centers, Ltd. v. Schweiker, 657 F.2d 135, 138 (7th Cir.1981). See also Pennsylvania Pharmaceutical Ass’n v. Department of Public Welfare, 542 F.Supp. 1349, 1355-1356 (W.D.Pa.1982) (Medicaid Act not intended to benefit pharmacists who sought to challenge rate of reimbursement for prescriptions). The present case is distinguishable from those cited by HHS, however, because Granville’s interest is not simply financial. Cf. Town Court Nursing Center, Inc. v. Beal, 586 F.2d 266, 277 (3d Cir.1978) (en banc) (provider claiming financial hardship “can look to private patients”). Granville has sought private paying patients to continue to exist as a chemical dependency treatment facility, but it remains frustrated in its nonprofit corporate purpose, which is to serve indigent and minority clients, by the lack of Medicaid eligibility of those clients because its facilities are viewed as “institutions for mental diseases.” This nexus between Granville’s purpose and the purpose of the Medicaid statute, which is to provide medical assistance to the needy, combined with Granville’s economic injury, is sufficient, we believe, to confer standing in this case. III. PROPRIETY OF JUDICIAL REVIEW HHS also argues that the matter is not ripe for judicial review. In Abbott Laboratories v. Gardner, 387 U.S. 136, 87 S.Ct. 1507, 18 L.Ed.2d 681 (1967), the Supreme Court enunciated the two factors relevant to a ripeness decision: first, the fitness of the issues for judicial resolution, and second, the hardship to the parties of withholding court consideration. Id. at 149, 87 S.Ct. at 1515. The district court found that the issue presented in this case was ripe, rejecting HHS’s argument that the court should await “the more fully developed factual record emanating from the administrative agency charged with responsibility for interpreting and enforcing the statutory framework under review.” 550 F.Supp. at 630. The court based its decision on a determination that it would be futile to pursue the matter through the administrative process, and on a finding that the case presented “issues of sufficient immediacy and reality to justify the issuance of a declaratory judgment.” Id. at 631. HHS’s contention regarding “ripeness” raises the issue of whether the district court was the appropriate forum in which to resolve the validity of HHS’s classification of chemical dependency. See Craigg v. Russo, 667 F.2d 153, 160 (D.C.Cir.1981). The inquiry centers on whether, once the court’s jurisdiction is established, it is proper for the court, under all of the circumstances, to reach the merits of the case before it. Although we agree with the district court that HHS’s consideration of the matter to this point has been inadequate, we agree with HHS that the district court should not have acted on the basis of this record without the benefit of a final decision of the Grant Appeals Board (GAB). There is no clear indication in the Medicaid statute, legislative history, or regulations as to what Congress meant by the term “mental disease.” The phrase “institutions for mental disease” was first inserted in the Social Security Act in 1950, as an exemption to “old age assistance” under Title I, apparently to avoid providing assistance to patients in state mental hospitals, which Congress considered a state responsibility. See H.R.Rep. 1300, 81st Cong., 1st *1300 Sess. 42 (1949), U.S.Code Cong. & Admin. News 1950, p. 3287; Schweiker v. Wilson, supra, 450 U.S. at 225 n. 5, 237 n. 19, 101 S.Ct. at 1078 n. 5, 1084 n. 19. See also 5. Rep. No. 404, 89th Cong., 1st Sess. 146 (1965), reprinted in 1965 U.S.Code Cong. & Ad.News 1943, 2086. Although Congress used the term again in 1965 in Title XIX, and subsequently amended this statute in 1972 to lift the exemption as it applied to persons under age twenty-one who received inpatient psychiatric services, the record is simply silent as to whether Congress intended to include chemical dependency within the term “mental disease.” There is also very little information in the record concerning the decision by the Health Care Financing Administration (HCFA) to use the ICD to define this term. Apparently the Bureau of Program Policy made this decision in 1975, without the benefit of a hearing or of comments from interested parties. At oral argument before this Court, counsel for the government stated that the ICD was widely used throughout HHS as a reference manual for disease classification purposes, but there is no indication that HHS has ever carefully considered whether chemical dependency should properly be viewed as a mental disease under the statutory framework of the Medicaid program. Moreover, there has been no “final” decision by the Grant Appeals Board that the agency’s use of the ICD’s classification of chemical dependency is proper. We agree with the district court that resort to the administrative procedures culminating in a GAB decision may not persuade the agency to change its views, since the GAB has decided in a related matter that the use of the ICD is reasonable. See generally Grant Appeals Board Decision No. 231, JOINT CONSIDERATION: “Institutions for Mental Diseases,” Docket Nos. 79-52-MN-HC, 79-89-MN-HC, 80-44-IL-HC, 80-150-CT-HC, 80-184-CA-HC, slip op. at 21-22 (November 30, 1981). Nevertheless, given the lack of any hearings or formal consideration prior to the adoption of HHS’s classification of chemical dependency, we believe that a final GAB decision would offer the agency an opportunity to articulate its rationale for its use of the ICD and for its rejection of the position argued by Granville in this case. HHS correctly argues that there is a factual dimension to the issue presented — a serious question concerning whether chemical dependency is properly viewed as a mental disease was raised by the evidence offered by Granville in the district court proceedings. Granville presented testimony from eight highly qualified physicians, psychiatrists, and chemical dependency practitioners that chemical dependency is a primary disease that should not be viewed as a mental disease for purposes of the Medicaid program. 6 All of Granville’s witnesses had been actively involved in the chemical dependency field in Minnesota, in programs that are regarded as outstanding: two nationally and internationally renowned facilities, the Hazel-den Foundation and the Johnson Institute, are located in Minnesota. Granville’s witnesses testified that the basic assumption of the only successful chemical dependency treatment programs, which all utilize the Alcoholics Anonymous (A.A.) approach, is that chemical dependency is a disease, a primary diagnosis in its own right, and not a symptom of some underlying mental disorder. Dr. Daniel Anderson, a consulting clinical psychologist and President and Director of the Hazelden Foundation, described the history preceding the adoption of the disease concept in this way: First of all, developmentally, alcoholism was simply considered to be a condition of willpower. Most alcoholics were treated in the criminal justice system. *1301 Then an attempt was made to explain alcoholism or chemical dependency as an illness. And these early explanations were mainly psychiatric explanations where it was presumed that the excessive drinking or excessive drug use was symptomatic of some underlying psychiatric problem, and that the goal was to treat the underlying psychiatric problem and then the symptomatic behavior of alcoholism or drug addiction would go away. Unfortunately that treatment never did prove to be successful, to my knowledge. Dr. Robert Premer, Chief of Orthopedics at the Veterans Administration Hospital in Minneapolis, Minnesota, associate professor at the University of Minnesota Hospital, and Chairman of the Board of the Johnson Institute, testified that the disease concept of alcoholism is commonly accepted in the medical profession today. Premer criticized the classification of chemical dependency as a mental disorder, stating: I think one of the things that has kept more chemically-dependent people ill than anything else is the idea that it is a mental illness; because they tend to attach to this, and “If you could just treat my mental illness, then I could continue to take my drugs.” So I think this idea that it is a mental illness is not only fallacious, it is pernicious. It is not supported by the facts. The facts are that people who are chemically dependent, and particularly alcoholics, have lost control of the use of alcohol. It is a primary condition caused by the misuse of alcohol. Dr. Richard Heilman, a physician and psychiatrist who is the Director of the chemical dependency treatment program at the Veterans Hospital, as well as senior consulting psychiatrist at Hazelden, indicated that among his colleagues, chemical dependency is recognized and understood to be a disease that does not necessarily imply a mental disorder or mental disease. 7 Heilman rejected the terminology “mental illness” to describe chemical dependence, as did Dr. Randall Lakosky, another physician and psychiatrist who treats chemically dependent patients in his private practice. Lakosky stated that psychiatrists do not generally recognize alcoholism as a mental disorder. Leonard Boche, a chemical dependency practitioner since 1959, past Director of Minnesota’s alcohol and drug program and current Director of a private chemical dependency program, and Dr. Thomas Briggs, a physician who has specialized in alcoholism treatment for twenty-two years and who is currently the medical director of the Alcohol Treatment Center at St. John’s Hospital, in St. Paul, Minnesota, shared the view that chemical dependency was a disease in its own right, and criticized the use of the ICD to define it as a mental disease. Both acknowledged that the ICD was used for reimbursement purposes by the HCFA for Medicaid and by insurance companies, but they did not consider such use reasonable because as Boche stated: I believe it is somewhat of an artificial system that was designed for another purpose; and that purpose * * * was * * * for. professional communication. And to use a tool — essentially designed for professional communication — as a method of determining payment, in my judgment is inappropriate. Boche stated that the concept of mental disorder is not applied to the treatment of chemical dependency in Minnesota, because “the treatment is so very different in this State — between mental illness and chemical dependency — that they are literally two different systems.” He noted that while there was a clearly established history of recovery in chemical dependency treatment, “there is not the same optimism about the successful treatment of mental illness.” *1302 Dr. Premer rejected the classification schemes of both the ICD and the Diagnostic and Statistical Manual of Mental Disorders (DSM-III); 8 Premer would classify alcoholism and other drug dependencies “in a separate classification because [the diseases are] being caused by the drugs.” Premer made the following comments concerning the ICD: It’s just a classification, putting a name on something. And, you know, the classification could be accurate or inaccurate as far as the medical facts are concerned. And it is my view that that classification is inaccurate as far as alcoholism is concerned. And in the future I would suspect— when they come to their senses — they will reclassify it. As the thing is set up right now, it’s inaccurately classified. And, you know, that classification doesn’t really have any relationship to the facts — as we know them today — about alcoholism and chemical dependency. It was apparently made by a group of psychiatrists who consider it a mental illness [and who have] little experience with alcoholism.[ 9 ] Dr. Heilman agreed that chemical dependency should be reclassified outside of the mental disorder category, even though he agreed that in general a primary disease could also be a mental disorder. The following dialogue between Heilman and counsel for HHS reveals his views more fully: Q: Does the fact that a disorder could be considered a primary disease, does that necessarily exclude the idea that it is also a mental disorder? A: Well, I think that is important, to make a distinction, I think this is where we may be having differences. Years back with Freud, the field of psychiatry was dealing primarily with consequences of physical problems— Freud was a neurologist — and there was a lot of mental illness caused by syphilis; and until they found out that that was its cause, they thought it was a mental illness. And when they found out that syphilis was underlying it, they reclassified it. And the same way with thyroid diseases that manifested themselves as psychiatric problems — when they found out it was thyroid disease, they reclassified it. And the same way with brain tumors. And I think that is kind of the issue here, that [chemical dependency] is addiction. In Minnesota we’ve kind of honed in on what [chemical dependency] is, underneath, as primary. It is an addictive problem — which is not made reference to in the [DSM-III], They talk about only the consequences in [DSM-III] — not about what the problem is. They don’t define what “addiction” is. HHS defends its use of the ICD and DSM-III classifications despite this substantial criticism by chemical dependency physicians, psychiatrists, and practitioners in effect suggesting that while the ICD and DSM-III are not perfect, they are the best classification systems available, and their use, as many of Granville’s witnesses acknowledged, is thus a “necessary evil.” See Appellant’s Br. at 17. The agency also *1303 presented the testimony of Dr. Robert Spitzer, a physician and psychiatrist who is Chief of Psychiatric Research with the Biometrics Research Department, New York State Psychiatric Institute, and Professor of Psychiatry at Columbia University, in support of its classification of chemical dependency as a mental disease. Spitzer is a generally recognized expert in disease classification, and he was a consultant to the last revision of the ICD, as well as chairperson of the task force that developed the American Psychiatric Association’s DSM-III. Spitzer testified that the primary reason he believed it was reasonable to classify alcoholism and chemical dependency as mental disorders in these references was because that classification “identifies correctly the nature of the symptoms,” which are primarily psychological and behavioral. He also notéd that this classification “correctly identifies that we don’t yet know the entire etiology [or] cause of alcoholism, and * * * we generally classify as mental disorders those behavioral or psychological conditions for which the ultimate etiology is still unknown.” Finally, he indicated that classifying chemical dependency as a mental disorder “suggests that treatment at the present time is probably going to be nonphysical * * * that is, drugs are not very effective, and the treatment generally consists of programs and approaches which are essentially psychological.” Spitzer’s view is that the A.A. model of treatment is “behavioral or psychological rather than physical.” While Spitzer suggested that one reason to classify chemical dependency as a mental disorder would be to indicate that psychiatrists ought to be able to study — in a helpful way — this disease, he agreed that “psychiatrists generally recognize that they don’t have an awful lot to offer — as psychiatrists — in the treatment of alcoholism.” Spitzer also conceded that the A. A. model is an effective, non-psychiatric treatment for chemical dependency, and he acknowledged that although there was not much controversy about the classification of alcoholism as a mental disorder among members of the substance abuse classification committee, neither this classification nor the DSM-III were designed with a reimbursement purpose in mind. In fact, he stated that “there were no considerations given to the use of this classification for reimbursement.” The Grant Appeals Board has not had the opportunity to consider this very important factual information. Moreover, as the agency charged with administering the Medicaid Act, HHS legitimately claims that the district .court has prematurely interfered with its processes in ruling on the merits of Granville’s case. The Grant Appeals Board does have at least two cases pending before it which present the precise issue litigated here. See Appeal of Colorado, Department of Social Services, Docket Nos. 82-125-CO-HC, 83-22-CO-HC (filed July 22, 1982 & January 28,1983). A hearing has been held before the Board on these cases, but counsel for the government informed this Court that the Board has stayed its consideration of the issue, pending our decision in this case. Clearly, an adequate factual record has been developed to permit a decision to be made concerning the validity of the use of the ICD to classify alcoholism and chemical dependency as mental diseases under Medicaid, but we are concerned that the agency’s Grant Appeals Board has not had the opportunity to evaluate this evidence nor to offer its considered analysis of the issue before us. HHS argues that if we agree that it should be given the opportunity to make a final decision, we should reverse the district court with directions to it to dismiss the ease. We are reluctant to dismiss the case outright, however, because of the hardship to the parties that such action would create. This is not a case such as was presented in Abbott Laboratories v. Gardner, supra, 387 U.S. at 152-154, 87 S.Ct. at 1517-1518, where the party challenging the agency action was faced with compliance at a very high cost or the risk of serious civil and criminal penalties, id. at 153, 87 S.Ct. at 1517, but the impact on Granville is certainly “sufficiently direct and immediate.” Id. at 152, 87 S.Ct. at 1517. One of Granville’s three facilities, Team House, has received certification by the State Department of *1304 Health, the first step in the Medicaid eligibility process. Granville voluntarily delayed pursuing certification of all its facilities pending some indication that its clients would be eligible for Medicaid funds because of the costs of the requirements for certification; costs which Granville did not believe would improve the quality of its programs. Granville has made other substantial expenditures in bringing its facilities into compliance with Title XIX requirements, and has likely incurred costs in advancing the present litigation, all in its attempt to advance its legitimate corporate purpose and to have declared invalid an interpretation of “mental diseases” that it views as wholly unsupported by the current state of knowledge concerning the nature of the diseases of alcoholism and other chemical dependencies. HHS would offer Granville no recourse but to state administrative procedures, which culminate in a state court review of whether the Department of Public Welfare acted properly. We deem this remedy inadequate under the present circumstances. Alternatively, since the State has now aligned itself against HHS in this matter, presumably the agency would require the State to exhaust its administrative remedies. Under this scenario, Granville would have to admit a Medicaid-eligible patient to Team House, treat the patient with its own funds, submit a request to the State for reimbursement, convince the State to agree to reimburse, even though the State has little hope of receiving any federal financial participation for some time, and then convince the State to press its argument before the Grant Appeals Board and the federal courts. This expenditure of time, effort, and additional money would do nothing to sharpen the issue for decision, however, and would add nothing to the present record save the Grant Appeals Board’s official decision in the matter. For the above reasons, we believe the best course in this case is to reverse and remand to the district court, with directions to it to remand to the Grant Appeals Board to make an initial determination on the issue of whether otherwise-eligible residents of Granville’s facilities (which have been certified as intermediate care facilities) are eligible for Medicaid under 42 U.S.C. §§ 1396 et seq. The district court shall certify the record made in the present case to the Board, which may consolidate this record with the two Colorado cases now pending before it. The Board may conduct such further evidentiary hearings as it deems appropriate, see 45 C.F.R. § 16.13 (1982), and the district court shall retain jurisdiction pending the outcome of the Grant Appeals Board decision. Cf. Craigg v. Russo, supra, 667 F.2d at 160; Hayes v. Secretary of Defense, 515 F.2d 668, 675 (D.C.Cir.1975) (district courts ordered to retain jurisdiction pending remand to administrative agencies). Any party may resubmit the case to the district court after the Grant Appeals Board renders a decision. The district court shall, after reviewing the decision of the Grant Appeals Board, make such further rulings as it deems necessary, and any party dissatisfied with the district court’s decision may again appeal to this Court. Accordingly, the decision of the district court is reversed in part and the case is remanded for further proceedings consistent with this opinion. Each party to this appeal shall bear its own costs. Footnotes [1] Prior to 1972, the costs of treating indigent patients at Jane Dickman House were paid through private donations. In 1972, Granville began receiving federal money through Title IV-A of the Social Security Act. In 1974, Title IV-A monies were replaced by Title XX funds. Commencing in 1978, these funds have been dramatically reduced, and Granville has no longer been able to rely upon them as the major source of its funding. [2] An intermediate care facility is defined by statute as an institution that is licensed under state law to provide, on a regular basis, health-related care and services to individuals who do not require hospitalization, but who, because of their physical or mental condition, require care and services above the level of room and board. 42 U.S.C. § 1396d(c) (Supp. V 1981). [3] Granville’s younger patients may be eligible for Medicaid under this statutory scheme. See 42 C.F.R. § 440.160 (1982). [4] Mental retardation is listed as a mental disorder in the International Classification of Diseases (ICD), but Title XIX specifically allows reimbursement to residents in facilities for the mentally retarded. See 42 U.S.C. § 1396d(d) (1976). [5] After granting the motion to consolidate, the district court subsequently held in the companion case that by “institution for mental diseases” Congress meant those institutions which provided primarily long-term care for the mentally ill by administering psychiatric treatment for its residents on the premises. Minnesota v. Schweiker, Civ. No. 4-82-155, slip op. at 15 (D.Minn. Aug. 25, 1982). An appeal by HHS from that decision is presently pending before a different panel of this Court. Minnesota v. Heckler, Nos. 82-1164, 82-2297 (8th Cir., submitted Nov. 8, 1982). See also Connecticut v. Schweiker, 557 F.Supp. 1077, 1088-1089 (D.Conn.1983) (the term IMD refers only to facilities that provide total care to patients). The narrower question presented in the instant case is whether chemical dependency is properly characterized as a mental disease under the Medicaid statute. [6] See Testimony of Dr. Thomas Briggs, Tr. I-11; testimony of Dr. Randall Lakosky, Tr. 1-64, 65; testimony of Dr. Sharon Woods, Tr. 1-106; testimony of Leonard Boche, Tr. 1-251; testimony of Dr. Robert Premer, Tr. 11 — 12, 17; testimony of Dr. Daniel Anderson, Tr. II — 48; testimony of Dr. Richard Heilman, Tr. 11-78, 80; testimony of Dagny Christiansen, Tr. 11-135. See also Deposition testimony of Dagny Christiansen at 25. [7] In fact, most people suffering from chemical dependency do not also suffer from a mental disease. Granville does not treat patients who have underlying psychiatric difficulties — those clients are referred elsewhere. It only admits patients who have a primary diagnosis of chemical dependency. [8] The American Psychiatric Association is responsible both for publishing the DSM-III and for providing a revision of the mental disorders section of the ICD. Both publications classify chemical dependencies as mental disorders. [9] Dr. Robert Spitzer testified that the committee for the ninth revision of the substance use disorders classified in the ICD was composed of five to seven psychiatrists and one sociologist. Members of the committee included: Dr. Lee Robins, a sociologist, who had received a grant from the federal government to study the problems of addiction; Dr. Jerome Jaffee, a psychiatrist, who was a special consultant on drug abuse problems during the Nixon Administration; Dr. Donald Goodwin, a professor in psychiatry, who has specialized in alcoholism and has done studies on genetic twins and alcoholism; Dr. Sheldon Zimberg, a psychiatrist, who was running an alcoholism treatment program in New York City; and Dr. Phillip Zeidenberg, a psychiatrist who had experience in the treatment of alcohol problems.