Full opinion text
FINDINGS OF FACT AND CONCLUSIONS OF LAW ZILLY, District Judge. THIS MATTER came before the Court for trial commencing April 10, 1991. Trial to the Court, sitting without a jury, concluded on May 13, 1991. At the conclusion of trial, the Court took the case under advisement. The Court, having reviewed the testimony presented at trial and the exhibits, having considered the trial briefs and oral argument, and being fully advised, now makes and enters its Findings of Fact and Conclusions of Law. FINDINGS OF FACT I.INTRODUCTION 1. Plaintiffs Multicare Medical Center d/b/a Tacoma General Hospital; Central Washington Health Services Association d/b/a Central Washington Hospital; Public Hospital District No. 2, Snohomish County, Washington d/b/a Stevens Memorial Hospital; St. Joseph Hospital and Health Care Center; Good Samaritan Hospital; Sisters of Providence in Washington d/b/a Providence Hospital of Everett; Lakewood General Hospital; Sisters of Providence in Washington d/b/a St. Peter Hospital; and Health Resources Northwest d/b/a Northwest Hospital, are not-for-profit corporations or Public Hospital Districts organized and operated under the laws of the State of Washington, and licensed under Chapter 70.41 of the Revised Code of Washington (“RCW”), as acute care hospitals providing inpatient services. Plaintiff Legacy Health System does business as Emanuel Hospital and Health Center (“Emanuel Hospital”) and is a not-for-profit corporation organized and operating under the laws of the State of Oregon. Emanuel Hospital is located in Portland, Oregon, where it operates as an acute care hospital providing inpatient services to residents of Oregon and Washington. 2. State defendants are the State of Washington, Department of Social and Health Services (“DSHS”), and Richard Thompson, the Secretary of DSHS (collectively, the “State”). 3. Federal defendants are the United States of America, Department of Health and Human Services (“HHS”), and Louis Sullivan, M.D., Secretary of HHS (collectively, the “federal defendants”). 4. The State of Washington participates in the federal Medicaid program established by the Medicaid Act, Title XIX of the Social Security Act, 42 U.S.C. §§ 1396-1396q. Defendant DSHS, under the direction of Richard Thompson, is the state agency responsible for administering the Medicaid program in the State of Washington. Defendant HHS is the federal agency charged with administering the federal Medicaid program and has delegated much of this responsibility to the federal Health Care Financing Administration (“HCFA”). 5. Plaintiffs bring this action pursuant to 42 U.S.C. § 1983, alleging that State defendants deprived them of their rights secured under the Medicaid Act. Plaintiffs challenge the State of Washington’s new second generation Diagnosis Related Groups (DRG) based reimbursement system for inpatient hospital services for Medicaid recipients which became effective April 1, 1988. Plaintiffs specifically challenge the State’s compliance with the reimbursement standards set forth in 42 U.S.C. § 1396a(a)(13)(A). Plaintiffs’ action against the federal defendants challenges HCFA’s conduct under 5 U.S.C. § 702 as being arbitrary and capricious and otherwise not in accordance with law. 6. The Medicaid Act is a cooperative federal-state program designed to provide medical services to certain low-income persons. Participation by the State of Washington is voluntary. However, once the State makes the decision to participate in the program, it must comply with the federal Medicaid laws and regulations. Amisub, (PSL), Inc. v. Colorado Dept. of Social Services, 879 F.2d 789, 794 (10th Cir.1989), cert. denied, — U.S.-, 110 S.Ct. 3212, 110 L.Ed.2d 660 (1990); Mississippi Hospital Ass’n. v. Heckler, 701 F.2d 511, 515 (5th Cir.1983). 7. The Medicaid Act requires each State to submit its Medicaid plan to the federal government, specifically HCFA, for approval. 42 U.S.C. § 1396. In 1981, Congress amended section 1396a(a)(13)(A) (the “Boren Amendment”) to require each State to “find” and make “assurances” satisfactory to HCFA that its inpatient hospital rates: (1) are “reasonable and adequate to meet the costs which must be incurred by efficiently and economically operated facilities in order to provide care and services in conformity with applicable State and Federal laws, regulations, and quality and safety standards” (the “economy and efficiency requirement”); (2) are reasonable and adequate “to assure that individuals eligible for medical assistance have reasonable access (taking into account geographic location and reasonable travel time) to inpatient hospital services of adequate quality” (the “access requirement”); and (3) take into account “the situation of hospitals which serve a disproportionate number of low income patients with special needs” (the “disproportionate share requirement”). 8. In implementing the Boren Amendment, HCFA promulgated regulations requiring state Medicaid agencies to obtain HCFA approval of any State plan change in payment methods and standards. 42 C.F.R. § 447.253. In order to receive HCFA approval, the State must submit “assurances” satisfactory to HCFA that the requirements of federal law are being met. Whenever the State makes a change in its methods and standards, but not less often than annually, the State is required to make findings and assurances, satisfactory to HCFA, that the rates satisfy all requirements of federal law. 9. To make changes in its State Plan’s methods and standards, DSHS submits plan amendments using Transmittal Number (“TN”) designations. Under HCFA regulations, DSHS can, with proper public notice, implement a change in methods and standards without transmitting the state plan amendment and the required assurances, provided the Transmittal is dated before the end of the quarter in which implementation occurred. 42 C.F.R. § 447.256(c). 10. The Division of Medical Assistance is the division within DSHS which is directly responsible for administering the state’s Medical Assistance programs, including the Medicaid program and the state-only funded Medically Indigent and General Assistance-Unemployable programs (collectively, “MI/GAU”). At all times material hereto, the director of the Division of Medical Assistance was Ron Kero. The Rates Analysis Unit of the Division of Medical Assistance is responsible for developing and implementing medical assistance programs, including Medicaid. Prior to early 1989, Roger Gantz was the Manager of the Rates Analysis Unit. Since early 1989, Thomas Johnson has been the Manager of the Rates Analysis Unit. 11. The plaintiffs claim that DSHS violated the procedural and substantive requirements of the Boren Amendment with respect to the economy and efficiency requirement, the access requirement, and the disproportionate share requirement. II. ROLE OF THE WASHINGTON STATE HOSPITAL COMMISSION 12. From 1973 until 1989 the hospital industry in Washington State was regulated by the Washington State Hospital Commission (“Hospital Commission”), a state agency which established rates that hospitals could charge for health care services. From 1984 until its sunset in 1989, the Washington State Hospital Commission was comprised of nine members, one of whom was the Secretary of DSHS, or his or her designee, for the purpose of “representing the interests of the state as a major purchaser of health care services.” RCW 70.39.030(1)(g). 13. The Hospital Commission was charged with ensuring that hospital costs did not exceed those “necessary for prudently and reasonably managed hospitals.” RCW 70.39.140. The Hospital Commission required submission of an annual budget with budgeted volumes and expenses. To assist it in analyzing and ruling upon hospital budget requests, the Hospital Commission engaged a professional staff of employees. The staff had knowledge of and expertise in hospital cost analysis. 14. The Hospital Commission regulated Washington hospitals’ budgetary processes. It established by regulation an accounting and reporting manual (“the Manual”) containing a detailed uniform system of hospital cost accounting and reporting. WAC 261-20-020. Annually, each hospital filed a detailed report with the Hospital Commission pursuant to the Manual, which required expenses to be broken down into nine natural classifications. The Hospital Commission required hospitals to submit detailed cost reports each year. Pursuant to statutory requirements, the Hospital Commission held public hearings on, among other things, each hospital’s operating expenses, capital allowance, and deductions from revenue. The sum of these three components became the hospital’s total rate setting revenue. Revenue was not allowed for cost items disapproved by the Hospital Commission. The staff made recommendations to the Hospital Commission as to whether or not budgeted expenses, capital allowance and deductions from revenue met the statutory criteria applicable to the Hospital Commission set forth in Revised Code of Washington, Chapter 70.39, and the regulatory criteria set forth in the Washington Administrative Code at Chapter 261-40. 15.The Hospital Commission annually published and distributed to hospitals extensive cost and productivity data that compared hospitals to other hospitals in groupings called peer groups. The Hospital Commission classified Washington hospitals into four peer groups. Peer group A hospitals were essential rural hospitals, peer group B hospitals were non-essential rural and non-teaching urban hospitals, peer group C hospitals were teaching hospitals, and peer group D hospitals were hard-to-classify, specialty hospitals. The Hospital Commission used the peer groups to establish a percentile to trigger further budget review. The Hospital Commission presumed that hospitals with operating expenses below the 70th percentile in each peer group had reasonable operating expenses. Hospitals with operating expenses at or above the 70th percentile in each peer group were subject to further budget review. 16. In addition to approving costs and revenues for the budget year, the Hospital Commission also utilized a year-end conformance review to insure that hospitals’ actual performance conformed with the revenue and cost projections approved in the prior year. Each hospital was required to file a “year-end conformance report” containing the following: its Medicare cost report filed with Medicare, a financial statement including a statement of income and expenses audited by a public accounting firm, and a year-end report of expenses incurred and revenue received during the year. RCW 70.39.110; WAC 262-40-150(6). The Hospital Commission staff conducted a desk-audit, comparing the year-end conformance report with the Medicare cost report and the audited financial statements. A hospital was deemed out of conformance if its actual revenue exceeded the approved revenue adjusted for actual patient volumes. Hospitals could justify excess revenue, but all justifications had to be related to unanticipated and justifiable increases in operating expenses. WAC 261-40-150(6)(f). If the Hospital Commission refused to accept the justifications, a Decision and Order was issued (“year-end Conformance Adjustment”), reducing the hospital’s current year budget and rates by the amount that actual revenue exceeded approved revenue. 17. Because the Legislature took no action to reauthorize the Hospital Commission’s authority by June 30, 1989, the Commission’s rate-regulation authority sunset as of June 30, 1989 pursuant to RCW 43.-131.253. Although the Hospital Commission’s authority ended, its data collection function survived and was transferred to the newly created Washington State Department of Health (“DOH”) as of July 1, 1989 and continues to this day. DOH requires hospitals to continue preparing and submitting certain financial information such as budgets and budget amendments, year-end reports, quarterly reports, and changes in hospital prices. III. THE WASHINGTON MEDICAID REIMBURSEMENT SYSTEM 18. Prior to January 1, 1985, hospitals were paid for inpatient services to Medicaid patients on a retrospective basis. In 1983 and 1984, for example, payment rates were established by multiplying a hospital’s allowable billed charges by the ratio of Hospital Commission-approved operating expenses to Hospital Commission-approved total rate setting revenue (the “OE/TRSR” ratio). 19. Effective January 1, 1985, DSHS adopted its first truly prospective methodology for reimbursing inpatient hospital services. The new system, described in TN 85-15, used Diagnosis Related Groupings (“DRGs”) to pay hospitals prospectively-determined rates based on a Medicaid patient’s diagnosis. This system became known as the “first generation DRG system.” The first generation DRG system classified illnesses according to 486 different diagnoses and procedures, and assigned each a relative weight, designed to reflect relative resource consumption. Payments for Medicaid patients under the first generation DRG system were calculated by multiplying a DRG’s relative weight by a hospital-specific base rate, called a conversion factor. 20. Under the first generation DRG system, each hospital had a different payment rate, representing the hospital’s average cost per Medicaid patient. To calculate each hospital’s payment rate, DSHS first determined its 1982 costs, using the hospital’s 1982 Medicare cost reports. DSHS then divided the hospital’s 1982 base year costs into three components: operating costs, capital costs, and teaching costs. The hospitals’ 1982 base year component costs were then inflated to payment year levels. Teaching and capital costs were inflated annually based on increases for such costs approved during the Hospital Commission’s annual rate setting procedure. Operating costs were inflated annually using the DRI Hospital Market Basket Index. 21. Effective February 1, 1986, DSHS filed plan amendment TN 86-7, reducing payments to hospitals under the first generation DRG system by 23.7 percent across the board. On May 18, 1987, Judge Carolyn Dimmick declared State plan amendment TN 86-7 void, and found that the hospitals were entitled to payment at the full rates provided for in TN 85-15, the approved State plan in effect prior to the reduction. Washington State Hospital Association v. Washington, No. C86-463D, Slip op., 1987 WL 88997 (W.D.Wash., May 18, 1987), reprinted in Medicare & Medicaid Guide (CCH), ¶ 36,352 (1987). 22. Soon after Judge Dimmick invalidated TN 86-7, DSHS adopted TN 87-15. TN 87-15 made one significant change to the first generation DRG system: it substituted the Medicare Prospective Payment Update Factor (“MPPUF”) for the DRI Hospital Market Basket Index as the inflation factor used to inflate 1982 base year operating expenses. Except for this one change, the first generation DRG system remained in place until April 1, 1988, when the State replaced it with TN 88-8. 23. TN 88-8 implemented a new second generation DRG system which was the subject matter of this trial. While structurally similar to the first generation system, TN 88-8 made several significant changes to the rate setting methodology which resulted in a substantial reduction in overall Medicaid payments to hospitals. The various components of the second generation DRG system have been the focus of this litigation and are described in detail below. 24. The State’s new second generation Diagnosis Related Groups (DRG) based reimbursement system employs three methods to determine hospital payment rates. These methods are (1) DRG cost-based rates, (2) DRG selective contract rates, and (3) rates based on a hospital’s ratio of operating expenses to total rate setting revenue (OE/TRSR). The State’s selective contracting program is not part of the State Medicaid plan and is exempt from Boren Amendment requirements pursuant to a waiver from HCFA. Contracting hospitals participating in the State’s selective contracting program are paid in accordance with their contract price. Hospitals not located in selective contracting areas or otherwise exempt are reimbursed on a cost-based DRG rate. Non-contracting hospitals in selective contracting areas are paid on a cost-based DRG rate, but are restricted to treating emergency Medicaid cases only. IV. THE PLAINTIFFS’ CHALLENGE TO THE STATE’S CURRENT SYSTEM A. Scope of Trial 25. Although plaintiffs’ complaint challenges various State plans and amendments instituted since January 1, 1985, the Court limited the instant trial to the State’s second generation reimbursement system, described in TN 88-8 (the “second generation DRG payment system”), and subsequent plan amendments (collectively, the “State Plan”). TN 88-8, implemented on April 1, 1988, was submitted with the State’s assurances and related information to HCFA for approval on June 29, 1988, and eventually approved by HCFA on July 16, 1990. Subsequent to the implementation of TN 88-8, the DSHS submitted to HCFA three plan amendments: (1) TN 88-11, effective July 1, 1988, which increased reimbursement to hospitals serving a disproportionate share of low income patients; (2) TN 89-14, effective July 1, 1989, which increased reimbursement to hospitals for certain psychiatric and AIDS cases; and (3) TN 89-17, effective October 1, 1989, which provided reimbursement for services to chemically dependent pregnant women. These amendments did not alter the basic rate setting methodology established in TN 88-8. 26. The Court limited its review at trial to TN 88-8 and its amendments in order to comply with the Eleventh Amendment of the United States Constitution. While a federal court may award injunctive relief requiring state officials to conform their conduct to federal law, the Eleventh Amendment stands as a bar to claims against a State in federal court for money damages or other retrospective relief. Edelman v. Jordan, 415 U.S. 651, 666-67, 94 S.Ct. 1347, 1357-58, 39 L.Ed.2d 662 (1974). Moreover, a federal court may not issue a declaratory judgment that federal officials violated federal law in the past where there is no ongoing violation of federal law. Green v. Mansour, 474 U.S. 64, 106 S.Ct. 423, 88 L.Ed.2d 371 (1985). In the present case, there would be no continuing violation of federal law to enjoin if the current state plan was held valid. 27. On February 28, 1991, this Court issued a temporary injunction, enjoining the State from implementing a new state plan which was scheduled to go into effect on March 1, 1991. The Court enjoined the implementation of the new plan in part to preserve the plaintiffs right to challenge the current DRG reimbursement plan. The injunction shall remain in effect until final judgment is entered or until further order of this Court. B. DRG Payment Rates 28. A DRG (diagnosis related groups) system provides for payment to hospitals of prospectively-determined rates based on a Medicaid patient’s diagnosis related group. A DRG system is a patient classification system which classifies patients into groups based on diagnoses, the presence of surgical procedures, patient age, presence or absence of significant comorbidities or complications, and other relevant criteria. The DRGs categorize patients into clinically coherent and homogeneous groups with respect to resource use. The Washington Medicaid program currently uses 467 of 475 DRGs as classified by Grouper Version 5. Eight DRGs are not used because they represent exempt neonatal services, services no longer valid, or invalid or ungroupa-ble cases. (Ex. 439, pp. 41-42). Under a DRG reimbursement system, a hospital is reimbursed a fixed amount based upon the diagnosis grouping into which a patient's condition falls. The fixed amount is derived from hospitals’ historical costs incurred in treating patients in that DRG classification and generally does not vary, regardless of the patient’s length of stay in the hospital or the actual charges that are billed on account of that stay. The fixed amount or DRG rate for each hospital is multiplied by a relative weight assigned to each DRG, in order to determine the actual payment amount for an inpatient admission. For example, to determine how much payment plaintiff Central Washington Hospital should receive for performing a skin graft (DRG 264), DSHS multiplies Central Washington’s base payment rate of $1,569 by the relative weight of DRG 264, which is 2.2677, for a total payment of $3,558.02. (1) Calculation of DRG Payment Rates under the Second Generation DRG System 29.Under the second generation DRG payment system, the DRG payment rate is a hospital specific dollar amount which is multiplied by the applicable DRG weight to produce the DRG payment. The rate is established on the basis of a hospital’s average cost for treating a Medicaid patient during a base year period. For TN 88-8 and its amendments, the State selected 1985 as the “base year.” The base year rate is adjusted for the hospital’s case mix and, with the exception of capital-related costs, updated for inflation to the current payment year. The first payment year under the new plan commenced April 1, 1988. The specific calculations were made as follows: (a) DSHS determined fiscal year 1985 base year costs per admission for each hospital using audited Medicare cost reports, also known as HCFA Form 2552 reports. DSHS excluded from base year costs the costs of patients exempt from DRG reimbursement (“non-DRG patients”), such as rehabilitation, burn and neonate patients. Base year costs per admission were then divided into components consisting of operating expenses, capital costs (depreciation, interest and leases), and medical education costs. The three component costs per admission were then standardized by application of a Medicaid case-mix index. (b) The hospitals were then segregated into four peer groups that were developed and used by the Hospital Commission for budget review purposes. Peer Group A consists of essential rural hospitals, Peer Group C contains teaching hospitals, Peer Group D has specialty hospitals and Peer Group B contains all remaining hospitals. All Washington hospitals are divided into these four peer groups as shown by Exhibit 439. The standardized component costs for the hospitals in each peer group were then arrayed from lowest to highest so that peer group median component costs could be identified. DSHS used the median (50th percentile) Medicaid cost per admission for each component in peer groups A, B and C to limit or cap component costs. Peer group D, which consists of certain hard-to-classify specialty hospitals, was not subject to any peer group capping. If a peer group A, B or C hospital’s component cost was below the median, that component cost was not capped. The result was that each peer group A, B or C hospital’s allowable component cost was the lower of its own cost or the median cap. (c) The “capped” or, if below the median, actual operating and medical education component costs were then added together and inflated by the Medicare Prospective Payment Update Factor (“MPPUF”) from the 1985 base year to the current payment year. The 1985 “capped” capital cost component (depreciation, interest and leases), without any inflation, was then added to this amount to arrive at the base payment rate for the current year. For purposes of determining the payment amount for an inpatient admission, a hospital’s base payment rate is multiplied by the relative weight associated with the DRG into which an admission falls. C. OE/TRSR Payment Rates 30. Under TN 88-8 and its plan amendments, certain hospitals and services are exempt from the DRG payment method. Exempt hospitals include psychiatric hospitals, rehabilitation units, alcoholism treatment centers, pain treatment centers, managed health care, out-of-state hospitals, and military hospitals. Exempt services include neonatal services, rehabilitation cases, and AIDS-related services. Such hospitals and services are reimbursed under the “OE/TRSR method.” Under the OE/TRSR method, the basic payment is calculated by multiplying the hospital’s assigned OE/TRSR ratio by the allowed charges for medically necessary services. The OE/TRSR ratio, an acronym for the ratio of operating expenses to total rate setting revenue, is a ratio of cost to charges derived from using the Hospital Commission’s categories of operating expenses and total rate setting revenue. Recipient responsibility or third party liability payments are deducted from the basic payment to determine the actual payment for an admission. There is no inflation factor applied to payments made under the OE/ TRSR ratio. 31. The State also uses the OE/TRSR ratio in reimbursing extraordinarily expensive cases known as “outliers.” Under the State Plan, if the charges for a DRG case exceed a threshold of three times the DRG payment amount or $14,000, whichever is higher, the amount above that threshold generally is reimbursed at the level of 60 percent of reasonable cost, as estimated using the OE/TRSR ratio. Total payment in such cases equals the DRG rate plus a percentage of charges exceeding the outlier threshold amount. Cases that fall below the outlier threshold amount are reimbursed solely by the DRG rate. D. Selective Contract Rates 32. At the same time that it implemented the second generation DRG system in TN 88-8, the State also implemented a hospital “selective contracting” program. Under the selective contracting program, Medicaid services in certain selective contracting areas are provided primarily by hospitals that have entered into Medicaid contracts with the State. As of April 1, 1988, there were six selective contracting areas including Seattle, South King County, Spokane, Tri-Cities, Vancouver/Portland, and Yakima. Hospitals that are not in selective contracting areas (“noncon-tracting hospitals”), are paid pursuant to the formula rates determined under the second generation DRG system. None of the plaintiff hospitals have entered into selective contracts with the State. Where hospitals are located in a selective contracting area but do not enter into contracts with the State, they are limited, for the most part, to providing emergency services to Medicaid recipients. As of March 1, 1991, there were 24 Washington hospitals that had entered into selective contracts with the State. These hospitals perform approximately 43 percent of all Medicaid services in the State. The selective contracting hospitals and their contract rates for the period April 1, 1988 through March 1, 1991 are shown on Exhibit A-137 A. 33. To implement the selective contracting program, DSHS had to obtain from HCFA a waiver of the Boren Amendment and had to demonstrate to HCFA that the program was cost effective by showing that the selective contract prices in any given area were, in the aggregate, equal to or less than the formula prices that would be paid under the second generation DRG payment system in TN 88-8. The plaintiffs do not challenge the validity of the State’s selective contracting program. Y. FACTORS LEADING TO ADOPTION OF SECOND GENERATION DRG SYSTEM A. The Role Played By Budgetary Constraints in DSHS’s Development of the Second Generation DRG System 34. DSHS’s decision to design a new Medicaid reimbursement system was based in substantial part on the need to address an unexpected State budget shortfall for the 1987-89 biennium, which became apparent in December of 1986. DSHS knew in the last week of December that the Governor had cut DSHS’s Medicaid appropriation request to the State Legislature by $25.8 million ($14 million of State funds). DSHS informed the State Comptroller on January 15, 1987, that the Governor’s reduction in appropriations presented a “formidable task” to effect “a reduction of that magnitude from an already low level,” and that “the earliest possible implementation date for any new reimbursement methodology would be January 1, 1988.” (Ex. 44). DSHS knew that, to counter the effect of the Governor’s reduction in DSHS’s appropriation level, DRG rates for Medicaid cases would have to decrease 7.4 percent, effective July 1, 1987. DSHS concluded that the budget decreases would “necessitate a new hospital reimbursement system.” (Ex. 69). Ron Kero admitted in his testimony that the new system was required because the then available funding was known and fixed. 35. In early 1987, Kero made the decision to develop a new reimbursement system. DSHS contacted consultants from the Compass Consulting Group of Peat Marwick Main & Co. (“Peat Marwick”), who had experience in designing Medicaid prospective payment systems. 36. The second-generation DRG payment system was developed by the State with the assistance of Peat Marwick, over a period of approximately one year. Peat Marwick reported to the State in a series of documents called “Deliverables.” 37. Ron Kero devised a point-ranking system which was used “by and large” to make the decisions that led to adoption of the new reimbursement system and its elements. This ranking system was based on 100 points, of which 55 points were assigned to categories dealing with the impending budget shortfall: 50 points for “Option Achieves Necessary Savings” and 5 points for “Expenditures Are Predictable.” (Ex. 62). Fifteen points were assigned to providing access and quality of care. Ten points each were attributed to (1) simple administration, (2) management flexibility, and (3) equitable treatment of hospitals. Ron Kero’s handwritten notes on Exhibit 71 demonstrate that the system was designed based on DSHS’ knowledge that the appropriation did not provide as much money as it would like to pay hospitals and the need to decrease aggregate reimbursements based on a funding level that was known and fixed. In May 1987, DSHS’s budgetary constraints became more acute when Judge Dimmick invalidated the 23.7 percent across-the-board reduction in first generation DRG payment rates that DSHS had imposed in TN 86-7. 38. As of July, 1987, DSHS’s best estimate was that it would have to cut hospital Medicaid inpatient rates by 9.4 percent as of January 1, 1988. DSHS concluded that if the rate reduction was implemented after January 1, 1988, the reduction would have to be greater than 9.4 percent. 39. Handwritten notes of Roger Gantz, the DSHS official responsible for the development of TN 88-8, refer to the need for “inpatient rate reductions to meet both inpatient and outpatient target levels,” with target levels referring to the amount of the 1987-89 biennium appropriation. (Ex. 56, p. 40). 40. The ability to increase budget savings by picking and choosing the formula price elements was a theme that continued throughout development of the second generation DRG system. In October of 1987, DSHS provided Peat Marwick with appropriation levels for use in assessing the impact on state budget targets of the formula elements chosen to that date. DSHS informed Peat Marwick that to meet the appropriation level, hospital Medicaid inpatient rates had to be reduced by 11.6 percent effective January 1, 1988. (Ex. 953). This became the final “budget target.” 41. On November 12, 1987, Peat Mar-wick performed a budget impact analysis, Exhibit 57, showing that the formula elements chosen to that date produced reimbursement rates that were $11,948,276 above the budget target for the biennium. 42. As a result, DSHS made a number of policy decisions after November 12, 1987, to decrease the amount of state funds that would be required by the new system. These decisions included changes in outlier policies and changes in the OE/TRSR ratio between the first-generation and the second-generation system. DSHS’s decision to exclude outlier cases, the most expensive cases in the Medicaid system, from the calculation of the base year costs reduced rates by over 4 percent. (Ex. 35, p. 1). 43. A “Policy Impact Analysis” dated January 25, 1988, Exhibit 60, shows that the reimbursement system had, by that time, met the exact budget target of achieving an 11.6 percent reduction in payments. 44. During the period early 1987 through the adoption of the second generation DRG system in April 1988, the State expended thousands of hours in the design and implementation of the new program. This was the largest study and investigation ever undertaken by DSHS. The State also used the services of Peat Marwick who had substantial experience in the Medicaid reimbursement area. Peat Marwick prepared dozens of reports for the State setting forth available policy options and documenting the design for the system. DSHS also solicited and received input from the hospital industry. However, the Court finds that the driving force behind the decision to develop a new system and the policy decisions relating to the new system was the State’s desire to meet budgetary targets. Ultimately, the State selected various rate components and made certain policy decisions in order to achieve the desired reduction in payments, from an “already low level” to an even lower level reflecting an additional reduction in Medicaid benefits of over eleven million dollars ($11,000,000) per year. (Ex. 60). VI. COMPONENTS OF THE SECOND GENERATION DRG SYSTEM 45. The plaintiffs claim that the key methodological features of the second generation DRG payment rate, namely, the Hospital Commission peer groups, median capping of component costs, the selected inflation rate (MPPUF), and the capital freeze, were selected solely to meet the budget target, without a bona fide and objective investigation and due consideration of all relevant facts. Much of the evidence at trial focused on aspects of these specific components and, therefore, it is appropriate to describe each one in more detail. A. The Peer Groups 46. The second-generation DRG system uses the peer groups formerly used by the Hospital Commission. The Hospital Commission, as part of its budget review process, grouped hospitals into four peer groups in order to compare hospitals’ costs and charges. The first group, peer group A, is comprised of 32 essential rural hospitals, that is, those hospitals in rural areas whose continued operations are essential to the public interest. Peer group C contains 12 teaching hospitals, which are expected to have higher costs than non-teaching hospitals. Peer group D consists of hospitals thought to be difficult to classify or incomparable to others in their cost and budget structures. Peer group D includes a total of five (5) children’s and other highly specialized facilities. The last group, peer group B, consists of all hospitals that do not meet the criteria established for the other three peer groups. There are 50 hospitals in this peer group. Most of these hospitals are urban non-teaching and nonessential rural hospitals. During the Hospital Commission’s tenure, hospitals were free to request placement in a different peer group. As of September 1987, three hospitals had successfully challenged their classification and were moved to another peer group by the Hospital Commission. (Ex. 68). No appeal process was provided under the new State Plan to allow hospitals to challenge peer group classification and no hospital has attempted to challenge their peer group designation under the new Plan. 47. One of the tasks Peat Marwick was hired to perform in connection with the formulation of a new reimbursement system was the development of hospital peer group classifications. DSHS intended that the peer groups would be used to establish “efficiency caps” for determining conversion factor components. (Ex. 27). On August 17, 1987, Roger Gantz informed Ron Kero that there appeared to be a general consensus among Peat Marwick representatives that the Hospital Commission peer groupings were inadequate and “a ‘weak’ system which may have to be defended in court.” (Ex. 27). Peat Marwick presented DSHS with a series of options, including a proposal that DSHS commission a study to develop a Medicaid specific peer group system. Peat Marwick estimated that a study of this type would take six months to one year to develop and implement. Ron Kero rejected this option, explaining that there was “no time” to undertake such an endeavor. 48. In a subsequent report dated September 28, 1987, Peat Marwick identified again the methodological deficiencies of the Hospital Commission peer groupings, but ultimately recommended that DSHS adopt the Hospital Commission peer groups for use in setting caps for the formula pricing system. Peat Marwick concluded that “while the peer groups may have a number of methodological deficiencies, they are recognized by the State and the hospitals as the official peer groups currently in use.” (Ex. 68). 49. DSHS adopted the Hospital Commission’s peer groups as the basis for the second generation DRG payment system’s relative definition of efficiency. DSHS did not conduct any independent investigation or study of the Hospital Commission peer groups to determine whether they were a proper basis for capping costs. 50. DSHS and the Hospital Commission used the peer groups for different purposes and achieved vastly different results. The Hospital Commission used the peer groups to make comparative analyses and to establish a percentile, the 70th percentile, to trigger further budget review. The Hospital Commission did not use the peer groups to set or limit payment rates. In contrast, DSHS used the peer groups to set the payment rates for hospitals. Specifically, DSHS used the peer groups to cap the hospitals’ component costs at 1985 peer group medians. The impact of any methodological deficiencies in the peer groupings was not as significant to hospitals in the context of budget review as they were in DSHS’s rate setting process. The consequence of a hospital’s falling above the 70th percentile in the Hospital Commission’s budget review process was merely exposure to further budget review. Hospitals with high operating expenses relative to their “peers” were given the opportunity to justify their anticipated expenses and have such expenses approved. This was not true with respect to DSHS’s rate setting process. Component costs above the median were presumed to be excessive and inefficient and were disallowed. The State did not consider, and hospitals were not given an opportunity to identify, special factors or circumstances which set them apart from others in the peer group or which made their placement in a particular peer group inappropriate. 51. In October of 1988, after TN 88-8 was adopted, the Hospital Commission conducted a study of its peer grouping system and considered whether the Commission should revise peer groups A, B, and C. (Ex. A-489). The Hospital Commission concluded that the peer groups should be retained, explaining that none of the other alternative peer groups considered could be shown to be the “best,” and that hospitals were relatively comfortable with their peer group placements. (Ex. A-489). DSHS subsequently relied upon this Hospital Commission report as support for its later findings that use of the Hospital Commission peer groups to establish “efficiency caps” was reasonable. B. Use of Median (50th percentile) Caps 52. Peat Marwick explained the cap to be used in each peer group as setting “a level which defines the cost to an economic and efficient hospital of providing care to Medicaid recipients.” (Ex. 28, p. 4). 53. DSHS chose to cap hospitals’ 1985 base year costs at the median of each peer group, except Group D, for each of the three cost components. 54. DSHS witnesses testified that the median was selected because HCFA had approved this standard in other states and courts had accepted it as a standard of relative efficiency. DSHS made no attempt to ascertain the validity of DSHS’s peer grouping methodology relative to the peer grouping methodology used in these other states. DSHS did not investigate whether costs disallowed by adoption of the median cap were due to inefficiency or lack of economy. 55. DSHS witnesses testified that they knew of no other state that had ever applied peer group caps on a component basis and could not identify a judicial decision that approved peer group capping by component parts rather than in the aggregate. 56. Capping by components results in 64 percent of the hospitals in Washington having their base year costs capped. For peer group A hospitals, the capping resulted in an average reduction in the base payment rate of $93.76. For peer group B and C hospitals, the reductions were $126.37 and $99.85, respectively. (Ex. 20). Several of the plaintiff hospitals experienced a significant decrease in payments as a result of the capping. For St. Peter Hospital, a peer group B hospital, capping resulted in a reduction in its base payment rate of $383.48. Tacoma General, a peer group C hospital, experienced a reduction in its base payment rate of $183.70. 57. DSHS did not examine the appropriateness of using the median to cap component costs for each peer group of Washington hospitals. When it adopted the median caps, DSHS did not know that the Hospital Commission had gone from the median to the 70th percentile for its budget screens. DSHS conducted no investigation and performed no cost analysis to determine whether Commission-approved base year costs above the 50th percentile were costs that an economic and efficient hospital must incur. C. Use of the MPPUF as the Inflation Factor 58. DSHS first adopted the MPPUF as its inflation factor in TN 87-15, effective July 1, 1987, which amended the State’s first generation DRG system. In TN 87-15, DSHS substituted MPPUF for the DRI Market Basket Index as the inflation factor for that reimbursement system. 59. The DRI Market Basket Index (published by HCFA in the Federal Register) is a national measure of the increase in input prices for a standard “market basket” of goods and services typically purchased by hospitals. Thomas Johnson of the DSHS admitted that the DRI Market Basket is the best available and most common estimator of inpatient hospital costs. 60. MPPUF is the update factor developed annually and adopted by Congress to update payment rates in the Medicare prospective payment system. Medicare is the largest payor in the State of Washington, paying 35 to 40 percent of hospital revenues. Medicare has no requirement such as the Boren Amendment that it pay costs that must be incurred by efficiently and economically operated providers. 61. The Prospective Payment Assessment Commission (“ProPAC”) and HCFA annually recommend an MPPUF level to Congress. ProPAC arrives at the recommended MPPUF level by making adjustments to the DRI Market Basket Index to reflect various factors including increases in prices of goods and services, previous Medicare overpayment, medical technology changes, changes in productivity, changes in practice patterns, site selection, and coding improvements, also known as DRG creep. (Ex. A-372) Congress then sets the MPPUF on the basis of budgetary and political reasons. The MPPUF ultimately set by Congress is generally at the low end of the range of update factors recommended by ProPAC. 62. DSHS adopted MPPUF shortly after Judge Dimmick invalidated DSHS’s 23.7 percent across-the-board reduction in DRG rates under the first generation system. This decision, together with a reduction in DSHS’s budget for the 1987-89 biennium, resulted in a severe appropriation shortfall. In June of 1987, DSHS staff developed a series of options on how to address the shortfall. DSHS selected an option known as Option l.B, which called for the immediate substitution of the MPPUF for the DRI Market Basket Index as the inflation index in the first generation DRG payment system. At the time, DSHS expected the MPPUF, which is set by Congress in October of each year, to be .75 percent. DSHS made the substitution of the MPPUF in TN 87-15, which was implemented on July 1, 1987 and submitted to HCFA on September 28, 1987. Without the change to MPPUF as of July 1, 1987, DSHS estimated that there would have been a Medicaid budget shortfall of more than $31.7 million dollars ($14.6 million in State funds). (Ex. 51, pp. 2, 4). By replacing the DRI Market Basket Index with MPPUF, DSHS estimated that the shortfall in its Medicaid budget would be reduced to $28.3 million ($12.8 million in State funds). (Ex. 51, p. 7). See also Depo. of Roger Gantz, Vol. 8, pp. 13-14. 63. During the formulation of the second generation DRG system, Peat Marwick advised DSHS that the MPPUF was a “political football.” Peat Marwick also advised DSHS that it should not adopt an inflation factor lower than DRI without determining that such an inflation factor reflected local Washington State conditions. Peat Marwick further recommended that DSHS should not adopt the ProPAC and HCFA discretionary adjustments used to set the MPPUF unless DSHS first determined that the various policy adjustments addressed specific features in the Washington hospital industry. 64. DSHS elected to use the MPPUF, the Medicare update factor, for changes in inflation and other variables because the State’s Medicaid system “is also a DRG-based system, which impacts on factors affecting provider costs (e.g., provider efforts at cost containment, practice patterns, site selection) in a manner similar to the Medicare DRG System.” (Ex. 439, p. 10). The State represented to HCFA that MPPUF “represents a reasonable rate of growth in the cost of hospital services,” thereby ensuring that State rates comply with federal law and “are well above hospitals’ average variable costs.” (Ex. 439, p. 10). 65. Ron Kero testified that he selected MPPUF from a range of options in order to meet budget targets and to reign in excessive hospital inflation, which was increasing at rates three to five times greater then the Consumer Price Index. This is in contrast to the State’s written representations to HCFA on September 28,1987 that one of the reasons for the change from the DRI Market Basket Index to MPPUF was that Washington hospitals’ costs “reflect historically low lengths of stay and low costs per discharge in the State of Washington.” (Ex. 53). Mr. Kero testified further that he rejected the DRI Market Basket Index as an inflator because in his view DRI does not represent the actual rate of hospital inflation; rather, DRI measures only increases in hospital inputs. Kero also concluded that the DRI Market Basket Index does not take into account other factors affecting hospitals’ costs such as changes in productivity, practice patterns and technology. 66. MPPUF does not represent the actual rate of growth in the cost of hospital services in Washington State. (Exs. 803 and 897). The State has admitted that it knew that since 1985 Washington hospitals have been experiencing increases in operating costs exceeding DRI Market Basket rates. Pretrial Order, Admitted Fact 22. Notwithstanding this fact, prior to the adoption of TN 88-8, DSHS did not conduct any investigation to determine whether the inflation experienced by Washington hospitals was due to increases in costs that must be incurred by economic and efficient providers. In addition, DSHS did not investigate whether it was possible for economic and efficient providers to keep increases in their operating and medical education expenses to the level of the MPPUF. DSHS also failed to determine whether the discretionary adjustments ProPAC made to the DRI Market Basket Index were applicable to Washington hospitals. 67. MPPUF is only used prospectively by Congress to increase Medicare rates. DSHS applied MPPUF retroactively to limit payment amounts to a level substantially below actual costs when it used MPPUF to trend forward two of the three 1985 base year components. 68. On September 22, 1988, after DSHS submitted to HCFA a proposal to implement the second generation DRG system, HCFA informed DSHS that it did not have a reasonable basis upon which to accept the State’s assurance that the proposed rates were reasonable and adequate to meet the costs that must be incurred by economic and efficient providers. HCFA noted in Exhibit 34 that: [I]t was the intention of Congress that States, when establishing Medicaid payment rates for hospitals and long-term care facilities, use rates which take into account economic conditions during the period for which the rate has been established. In its methodology, the State proposes to calculate current and future rates based upon updates of the base year (1985) cost using the [MPPUF]. This factor is used to establish the annual increases in Medicare’s DRG payments. It is intended to account for a number of variables other than inflation, such as coding improvements and productivity improvements, and is designed to reflect hospitals’ Medicare payment experience under PPS. It is not clear how the State’s use of this factor will support its assurance that the proposed rates are reasonable and adequate to meet the costs which must be incurred by an economically and efficiently operated facility. 69. In its first formal submission of TN 88-8 to HCFA on May 12, 1989, DSHS told HCFA that it selected MPPUF to adjust 1985 base year costs because MPPUF “represents a reasonable rate of growth in the cost of hospital services, as determined by Congress and HCFA,” and because selection of MPPUF ensured that Washington’s Medicaid rates would be in compliance with the upper limits requirements in 42 C.F.R. § 447.253(b)(2). (Ex. 23). 70. A comparison of inflation factors for the period 1985 to 1988 is shown on Exhibit 803. The percentage change in hospital costs using MPPUF was 2.9 percent as compared to 12.6 percent using the DRI Market Basket Index. During the same period of time, the actual increase in costs of Washington hospitals was approximately 17.7 percent. As a result of using MPPUF, in conjunction with other rate components in the new second generation system, DRG payments to the hospitals in 1988 were less than their 1985 actual Medicaid costs. Dr. Dobson, one of plaintiffs’ expert witnesses, estimated that the amount of loss in payments to Washington hospitals during the first six months of the new plan (April 1, 1988 through September 30, 1988) due to the use of MPPUF and the capital freeze was approximately $4.5 million. The total shortfall in reimbursements for each full year of the Plan would be more than twice this amount. Thomas Johnson admitted at trial that as late as July 1, 1989, when the most recent amendment, TN 89-17, was implemented, the State still had made no investigation and had no evidence that an efficiently and economically operated hospital could limit its costs to the MPPUF rate of inflation. 71. The effect of capping costs at the median and using MPPUF is illustrated by Exhibit 114. The following four hospitals, which had uncapped component costs in the base year, received the following payments relative to their actual costs as of June 30, 1988, as shown at page 19 of Exhibit 114: Central Washington was paid 59% of cost (payment = $1484; cost = $2511) Good Samaritan was paid 79% of cost (payment = $2013; cost = $2542) Olympic Memorial was paid 56% of cost (payment = $1272; cost = $2245) St. Joseph — Tacoma was paid 49.6% of cost (payment = $2208; cost = $4444) 1. Upper Limits Requirements 72. Federal financial participation in Washington’s Medicaid program is not available to the extent that payment levels exceed certain defined “upper limits.” 42 C.F.R. § 447.257. The upper limits requirement provides that the aggregate Medicaid payments “may not exceed the amount that can reasonably be estimated would have been paid for those services under Medicare payment principles.” 42 C.F.R. § 447.272. DSHS must, whenever it makes a change in its methods and standards, but not less often than annually, make a finding that proposed payment rates will not exceed the upper payment limits. 73. In adopting MPPUF in TN 87-15, DSHS informed HCFA that the change was necessary to ensure compliance with the upper limits requirement. HCFA instructed DSHS on how to do the upper limits calculation, informing DSHS that the Medicare upper limit is to be determined by taking 1981 Medicare base year costs and inflating them to the payment year using the Tax Equity and Fiscal Responsibility Act of 1982 (“TEFRA”) target rates published each year in the Federal Register (Ex. 130). 74. In February of 1988, DSHS submitted to HCFA an upper limits analysis for TN 87-15 which compared Medicaid payments under TN 87-15 to payments under Medicare’s prospective payment system. HCFA rejected this analysis, calling it “questionable,” and informed DSHS that the “correct comparison” in the upper limits analysis “is the State’s aggregate payments with the amount that would have been paid had the applicable Medicare reimbursement principles been applied.” (Ex. 130). DSHS then submitted a correct upper limits analysis that showed that rates under TN 87-15 were 25 percent below the upper limit. 75. In adopting TN 88-8, DSHS again represented to HCFA that use of the MPPUF inflator was necessary to comply with the upper limits requirements. When DSHS performed an upper limits calculation for TN 88-8, the results showed that the payment rates under TN 88-8 were approximately 6 percent under the upper limits. This presents a mathematical impossibility since TN 88-8 payment rates were 11-12 percent lower than those under TN 87-15, which were 25 percent below the upper limits. All other factors being equal, the payment rates under TN 88-8 should have been more than 25 percent below the upper limits. In April 1990, in its final submission to HCFA, the State once again computed the upper limits calculation and concluded that the rates would be 5 percent below upper limits. This calculation was incorrect because it was based on a 1985 Medicare base year rather than the required 1981 TEFRA base year. DSHS staff offered various explanations for the discrepancies in the various calculations of the upper limits throughout the course of the trial, including claims of lost data and a desire to shift to better methodologies. Thomas Johnson testified at trial that the upper limits calculations for TN 87-15 and TN 88-8 were done differently and that this accounts for the inconsistent results. Mr. Johnson stated that the analysis for TN 87-15, showing that the State was 25 percent below the upper limits, used correct 1981 base year Medicare data, but excluded from its calculation the 20 percent of all claims paid on an OE/TRSR basis. The calculation for TN 88-8, showing the State to be 5 percent below the upper limits, used 1985 base year data and included all claims. According to Mr. Johnson, this accounts for the significant increase in payment levels in TN 88-8’s upper limits calculation. DSHS admitted, however, that DSHS also included in its upper limits calculation for TN 88-8, payments that are not subject to the upper limits test, including disproportionate share payments and certain capital reimbursements. Thus, the payment levels in TN 88-8’s upper limits calculation were overstated to the extent such payments were included. 76. The Court finds that DSHS’s claim that it had to adopt MPPUF in order to meet the upper limits requirements is not credible. The evidence establishes that DSHS's initial adoption of the MPPUF in TN 87-15 and its use of the MPPUF in TN 88-8 were driven by budgetary constraints and were not necessary to satisfy the upper limits requirements of federal law. This finding is further supported by the fact that DSHS has informed the Court that it intends to substitute the DRI Market Basket index for MPPUF in its 1991 amendment to the Medicaid reimbursement plan which was previously enjoined by this Court pending trial. 2. Presence of Adjustment Factors in Washington 77. DSHS adopted and maintained use of the MPPUF knowing that the MPPUF for 1986, 1987, and 1988 contained a downward adjustment for DRG creep. DRG creep is an increase in average case mix due to the assignment of cases to higher DRGs than the DRGs to which they were typically assigned in the past. Generally, DRG creep is caused by the assignment of certain other secondary diagnoses or additional procedures which cause the case to be assigned to a DRG with a higher relative weight. This results in a higher payment than would otherwise be made had the secondary diagnosis not been recorded or the procedure not performed. Mr. Gantz testified that DSHS conducted no investigation to determine if DRG creep had occurred in Washington. Mr. Gantz’s predecessor, Mr. Thomas Bedell, had determined in 1986 that the DRG creep phenomenon had not occurred in Washington State, but DSHS discounted Mr. Bedell’s study as too quick to be reliable. 78. Thomas Johnson testified that he believes that DRG creep had occurred in Washington. His belief is based on marked changes in case mix at Children’s Hospital from base year 1985 to 1988, which he attributes to coding improvements, and information contained in documents produced by the plaintiff hospitals during discovery in this case. Mr. Johnson admits, however, that when the State selected MPPUF as its inflation factor it did so without having conducted any study as to whether DRG creep had occurred. He admits that as of the time TN 88-8 was implemented the State had' no evidence of DRG creep in Washington. 79. DSHS adopted and maintained use of the MPPUF without investigating whether and to what extent the other discretionary adjustment factors used by Pro-PAC to set MPPUF occurred in Washington. Mr. Johnson testified that DSHS’s Washington-specific investigation into the applicability of the MPPUF discretionary adjustment factors consisted only of a thorough reading of ProPAC reports and DSHS’s reliance on its professional knowledge of the hospital industry. 80. Since 1986, the State, through the Hospital Commission and now the Department of Health, has collected and annually published data for each hospital in the state that reflect actual cost increases or decreases for operating expenses and capital costs that have been adjusted for changes in case mix of patients. The data are called “Trend Reviews.” The Trend Review data are detailed and displayed in the same manner as the inflationary inputs of the components of the HCFA-DRI Market Basket including wage and salary costs, changes in drug and supply costs, and the like. The Trend Reviews also separately show changes in hospital productivity, changes in inpatient/outpatient treatment, and changes in volume and case mix of patients. 81. The Trend Review analysis displays a net increase or decrease in operating expenses and capital costs for each hospital, which takes into account all changes in price inputs and any reductions, called “Operating Expense/Adjusted Case Mix Value Unit,” and displays the changes or trends over a period of five years. DSHS did not consult the Hospital Commission Trend Reviews, available for 1987 and 1988, and subsequent years, to determine whether use of the MPPUF was appropriate in Washington. 82. These Trend Reviews reveal that Washington hospitals experienced increases in operating and capital costs greater than the DRI even after taking into account the effect of various discretionary adjustments. 83. DSHS knew at the time it adopted the MPPUF that the Hospital Commission had approved net cost increases that exceeded the DRI, but adopted the MPPUF without investigating whether the Hospital Commission-approved increases were increases in costs that had to be incurred by economic and efficient providers in Washington. D. Freeze of Capital at the 1985 Level 84. In calculating a hospital’s DRG formula price rate under TN 88-8 and all subsequent plan amendments, the capital cost component was limited to the lesser of its 1985 capital costs or the 1985 peer group median capital costs. Capital costs include depreciat