Ask the Scholar

{
    "localId": "52875350",
    "label": "Safety Net [Folder 5]",
    "core": "doc",
    "dtoType": "document",
    "citationUrl": "https://catalog.archives.gov/id/52875350"
}

{
    "id": "52875350",
    "sourceUrl": "https://catalog.archives.gov/id/52875350",
    "contentType": "document",
    "title": "Safety Net [Folder 5]",
    "citationUrl": "https://catalog.archives.gov/id/52875350",
    "collections": [
        "Records of the Domestic Policy Council (Clinton Administration)",
        "Devorah Adler's Files"
    ],
    "iiifBase": "https://s3.amazonaws.com/NARAprodstorage/lz/presidential-libraries/clinton/wjc-dpc/7367451/42-t-7367451-20120463S-036-004-2016.pdf",
    "thumbnailUrl": "https://s3.amazonaws.com/NARAprodstorage/lz/presidential-libraries/clinton/wjc-dpc/7367451/42-t-7367451-20120463S-036-004-2016.pdf",
    "largeImageUrl": "https://s3.amazonaws.com/NARAprodstorage/lz/presidential-libraries/clinton/wjc-dpc/7367451/42-t-7367451-20120463S-036-004-2016.pdf",
    "imageCount": 1,
    "hasImages": true,
    "source": "import",
    "hasTranscription": false
}

{
    "url": "https://catalog.archives.gov/id/52875350",
    "naId": 52875350,
    "levelOfDescription": "fileUnit",
    "otherTitles": [
        "42-t-7367451-20120463S-036-004-2016"
    ],
    "recordType": "description",
    "ocrSource": "nara-archive"
}

{
    "seq": 1,
    "pageIndex": 0,
    "type": "document",
    "url": "https://s3.amazonaws.com/NARAprodstorage/lz/presidential-libraries/clinton/wjc-dpc/7367451/42-t-7367451-20120463S-036-004-2016.pdf",
    "mediaId": "8e5f53220d811fe9",
    "ocrText": "Health Security\nAct.\n(b) Such programs are ones for advanced nurse education, nurse practitioners, nurse\nmidwives, nurse anesthetists, and other training in clinical nurse specialties determined by the\nSecretary to require advanced education. Programs must meet the conditions defined in\nsections 821, 822, and 831 of the Public Health Service Act.\nSection 3062. Applicability of Part 1 Provisions. (a) The provisions of part 1,\napply to the graduate nurse training program under section 3061 in the same extent and\nmanner as they apply to the graduate physician training programs, except as they are\nmodified in this part. (b) For the purposes of this program, the council is the National\nCouncil on Graduate Nurse Education. The council will make allocation of nurse training\npositions in the same way as the Council on Graduate Medical Education does for physician\nresidencies.\nSection 3063. Funding. The amount available for graduate nurse training programs\nunder this part is $200,000,000 annually, provided as direct transfers from the Treasury\nunder sections 3034, and 3104. Under section 4051, payments under Medicare for the costs\nof approved educational programs other than direct graduate medical education costs are\nspecifically continued for cost reporting periods beginning on or after October 1, 1995.\nPART 3 -- RELATED PROGRAMS\nSection 3071. Programs of the Secretary of Health and Human Services.\n(a) Authorizes to be appropriated $400,000,000 for fiscal year 1994 and each year thereafter,\nfor carrying out the programs described in this section, in addition to amounts otherwise\nauthorized to be appropriated for such programs. Requires the Secretary of Health and\nHuman Services to carry out the programs in this section.\n(b) Establishes or expands existing programs with respect to training primary care\nphysicians and physician assistants, including programs to train additional numbers of\nphysicians and physician assistants; to retrain mid-career physicians previously certified in a\nnonprimary care specialty; to expand the supply of physicians with special training to serve\nin medically underserved areas; to expand service-linked educational networks for training in\ncommunity settings; to provide training in managed care, practice management, and\ncontinuous quality improvement; and to enhance information on primary care workforce\nissues.\n(c) Establishes or expands programs with respect to training of underrepresented\nminorities and disadvantaged persons, including programs to increase the number of such\npersons in the health professions through financial assistance, recruitment and retention,\nenhancing interest at the preprofessional level, and training of additional minority health\nprofessions faculty.\n(d) Establishes or expands programs with respect to training of nurses, including\ntraining additional numbers of nurse practitioners and nurse midwives; baccalaureate nurses\nTitle III\n113\nfor careers in teaching, community health service and specialized clinical care; nurse\nclinicians and nurse anesthetists; and, school-based community nurses; also programs to\npromote research on nursing workforce issues.\n(e) Establishes a program to develop and encourage adoption of model practice\nstatutes for advanced practice nurses and physician assistants, and other to support efforts to\nremove inappropriate barriers to practice by advanced practice nurses and physician\nassistants.\n(f) Establishes or expands programs with respect to training health professionals and\nadministrators in managed care, cost-effective practice management, continuous quality\nimprovement practices, and provision of culturally sensitive care.\n(g) Authorizes the Secretary to carry out these programs through existing programs in\nTitles VII and VIII of the Public Health Service Act.\nSection 3072. Programs of the Secretary of Labor. (a) Authorizes to be\nappropriated $200,000,000, for fiscal year 1994 and each year thereafter, for carrying out\nthe programs described in this section, in addition to amounts otherwise authorized to be\nappropriated for such programs. Requires the Secretary of Labor to carry out the programs\nin this section.\n(b) Establishes programs to provide for skills upgrading and occupational retraining\n(including retraining health care workers as technicians, nurses, and physician assistants) and\nfor quality and workforce improvement; to assist health care workers in career advancement;\nto develop health worker job banks; to provide for joint labor-management decision-making\non workplace matters; and to facilitate the comprehensive workforce adjustment initiative.\n(c) Requires the Secretary of Labor, in carrying out programs under this section, to\nprovide for specific skill requirements, internal career movement opportunities, employment\nduring retraining, evaluation and dissemination.\n(d) Requires the Secretary, in carrying out programs under this section, to provide for\njoint labor-management implementation and administration and discussion and consultation.\nSection 3073. National Institute for Health Care Workforce Development.\n(a) Requires the establishment of the National Institute for Health Care Workforce\nDevelopment jointly by the Secretary of Health and Human Services and the Secretary of\nLabor.\n(b) Authorizes the Secretary of Labor to carry out section 3073 through the Director\nof the Institute.\n(c) Requires the Director of the Institute to make recommendations regarding health\nTitle III\n114\nSUBTITLE E - HEALTH SERVICES FOR MEDICALLY\nUNDERSERVED POPULATIONS\n(H.R. 3600 and S. 1757 p. 578)\nPART 1 - COMMUNITY AND MIGRANT HEALTH CENTERS\nSection 3401. Authorization of Appropriations. (a) Authorizes funds in addition to\nother authorized funds for community and migrant health centers.\n(b) The additional funds authorized are $100,000,000 per year for each of fiscal years\n1995 through 2000.\n(c) These funds are in addition to other funds available for the centers.\nSection 3402. Use of Funds The additional funds authorized may be used for any\npurposes authorized under sections 329 or 330 of the Public Health Service Act as well as to\nestablish and maintain the financial reserves that are required under Title I for providers of\nhealth services.\nPART 2 - INITIATIVES FOR ACCESS TO HEALTH CARE\nSUBPART A - PURPOSES; FUNDING\nSection 3411. Purposes. The purposes of this program are:\n(1) to improve access of medically underserved populations to health services\nthrough a program of flexible grants, contracts and loans;\n(2) to establish transition to a system in which medically underserved\npopulations have adequate choice of community-oriented providers and health plans;\n(3) to promote the development of community practice networks and health\nplans that integrate public and private health providers in underserved areas;\n(4) to support linkages between provider of care to underserved populations\nand regional and corporate alliance health plans; and\n(5) to expand system capacity with additional practice sites and improvements\nin health-care facilities in need of repair.\n(6) to link providers in underserved areas with each other, regional health care\ninstitutions and academic health centers.\n(7) to support activities enabling underserved populations to gain access to and\neffectively use the health care system.\nSection 3412. Authorizations of Appropriations. For the development of\ncommunity health plans and networks there are authorized to be appropriated $200 million in\nfiscal 1995, $500 million in fiscal 1996, $600 million for fiscal 1997, $700 million for fiscal\nyear 1998, $500 million for fiscal year 1999, and $200 million for fiscal year 2000. These\nTitle III\n123\nfunds are in addition to other funds authorized for this program. Funds under this part are\nalso available for use under section 3692 (school health services).\nSUBPART B - DEVELOPMENT OF QUALIFIED COMMUNITY\nHEALTH PLANS AND PRACTICE NETWORKS\nSection 3421. Grants and Contracts for the Development of Plans and Networks.\n(a) The Secretary may make grants for developing qualified community health plans\nand community practice networks. These plans and networks may be, in this title, referred\nto as community health groups.\n(b) To be a qualified community health plan, a health plan must be public or non-\nprofit private entity whose principal purpose is to provide the comprehensive benefit\npackage, in areas with shortages of medical personnel or to populations with a significant\nnumber of medically underserved persons; the plan must be a member of one or more health\nalliances; and two or more of the categories of providers specified in subsection (d) must be\nrepresented in the plan.\n(c) A qualified community practice network means a consortium of health care\nproviders that is a public or private non-profit entity whose principal purpose is to provide\nservices to underserved populations or in health professions shortage areas, which has an\nagreement with one or more health plans and whose members are governed by a written\nagreement. Two or more categories of providers in subsection (d) must be included in the\nconsortium.\n(d) The relevant categories of providers described in subsections (b) and (c) are the\nfollowing:\n(1) physicians or other health professionals or health care institutions providing\ncare in a shortage area or to an underserved population;\n(2) migrant and community health centers;\n(3) entities funded under sections 340 and 340A of the Public Health Service\nAct (homeless health care providers and health care providers in public housing);\n(4) entities furnishing health services under Section 1001 or Title XXIII of the\nPublic Health Service Act (family planning clinics and Ryan White program\nproviders);\n(5) entities furnishing services under Title V of the Social Security Act\n(maternal and child health);\n(6) entities that are rural health clinics or federally qualified health centers;\n(7) entities providing health services to Indians in urban areas under Title V of\nthe Indian Health Care Improvement Act or outpatient services to Indians through the\nIndian Self Determination Act; and\n(8) state or local public health agencies.\nThe Secretary also may make grants to a health plan that is not a community health\nTitle III\n124\nplan but that seeks to develop a community practice network with the entities described in\nthis subsection.\nSection 3422. Preferences in Making Grants. In making grants, the Secretary shall\ngive the preference to those applications in which a maximum number of entities described in\nsection 3421(d) are represented and added weight if a large group also includes private\nphysicians, other health professionals or institutions that provide health services in a health\nprofessions shortage area, or provide health services to significant numbers of ilndividuals in\nmedically underserved populations.\nSection 3423. Certain Uses of Awards. (a) Awards under section 3421 may be\nspent for the following purposes:\n(1) planning the network or health plan, including entering into contracts;\n(2) recruitment and compensation of health and administrative staff;\n(3) acquisition and expansion of facilities;\n(4) acquisition and development of information systems;\n(5) other expenditures recognized by the Secretary.\n(b) An award that includes funding for capital costs must obligate the recipient to the\nUnited States for the amount of the award, plus interest during the 20 year period beginning\non the date of completion, if the applicant ceases to be a qualified health plan or network or\nis sold or transferred to an entity that is not a community health plan or network.\nSection 3424. Accessibility of Services. (a) Community health plans and networks\nmust assure that their services are available to persons seeking care, whether or not they are\neligible individuals under title I.\n(b) A community health group's providers must be approved as Medicare and\nMedicaid providers. A group must seek reimbursement for the cost of caring for persons\nentitled to benefits under Title I, Medicare, Medicaid, any other public assistance programs,\nor private health insurance plans.\n(c) The network or plan must prepare a fee schedule that is consistent with local rates\nand a corresponding schedule of discounts to be determined by a patient's ability to pay.\n(d) The plan or network must maximize the accessibility of its services to residents in\nits area by eliminating barriers resulting from geographical or demographic characteristics,\nincluding limited ability of patients to speak English. A plan or network also must\nperiodically determine the accessibility of its services.\nSection 3425. Additional Agreements. (a) Networks and plans must provide enabling\nservices (as defined in section 3461(g)) as part of their funding agreement.\n(b) Networks and plans must maintain ongoing systems for patient-oriented,\nTitle III\n125\ncommunity-responsive quality control, and for collecting and making public information on\ncosts, health-care and financial performance, and other matters.\n(c) The plan or network must agree to maximize its use of existing resources.\nSection 3426. Submission of Certain Information. (a) The Secretary may make\ngrants only if applicants submit information on the needs (including the need for enabling\nservices) of the medically underserved population to be served by the applicant.\n(b) The applicant must also include a description of how the applicant will design the\nplan or network, a description of efforts to secure financial and technical assistance, and\nevidence of significant community involvement in the initiation, development and ongoing\noperation of the project.\nSection 3427. Reports; Audits. Funded plans and networks must provide reports and\ninformation as required by the Secretary and must submit to annual audits.\nSection 3428. Application for Assistance. Applications must include the agreements\nand information required in other sections of the subtitle and additional agreements and\ninformation that the Secretary deems necessary.\nSection 3429. General Provisions. (a) The Secretary may not make more than two\ngrant awards under section 3421 for the same project.\n(b) The Secretary may determine the amount to be granted for any project.\nSUBPART C - CAPITAL COST OF DEVELOPMENT OF\nQUALIFIED COMMUNITY HEALTH PLANS AND PRACTICE NETWORKS\nSection 3441. Loans and Loan Guarantees Regarding Plans and Practice\nNetworks. (a) The Secretary may make loans to public and private entities for capital costs\nof developing qualified community health groups, and may also guarantee such loans by\nFederal and non-Federal lenders.\n(b) The Secretary shall use the same preferences in making loans that apply to grants\nunder section 3421.\n(c) Funds under this section may be used to finance facilities, major equipment,\nincluding information systems, to establish financial reserves and other capital costs that are\nnecessary to the purpose of the section (as determined by the Secretary). Priorities shall be\nplaced on loans to modernize facilities, prevent or eliminate safety hazards, and to repair or\nreplace obsolete facilities.\n(d) The principal of the loan or loan guarantee, when added to other assistance under\nTitle III\n126\nthis section, may cover up to 100 percent of the costs involved.\nSection 3442. Certain Requirements. (a) The Secretary may approve loans only if\nreasonably satisfied that the grantee can repay the loan, and only if the grantee provides\nassurances that additional funds are available to complete the project for which the loan is\nmade. Also, a loan under section 3441 must be on the terms and conditions that are\nnecessary to protect the financial interests of the United States (as determined by the\nSecretary).\n(b) The Secretary may guarantee loans only if the loan conditions, terms and\narrangements for repayment are sufficient to protect the financial interests of the United\nStates. Such guarantees are also subject to further terms as determined by the Secretary.\n(c) An applicant for a loan or loan guarantee must agree to use existing resources to\nthe maximum extent feasible.\nSection 3443. Defaults; Right of Recovery. (a) The Secretary may take necessary\naction, including waiver of regulatory conditions, deferral of loan repayments or other\nactions as needed to prevent a default on a loan or loan guarantee. The Secretary may also\nforeclose a loan in default, or waive, for good cause, any right of recovery from a borrower\nwho fails to make payments on a loan. A waiver of the right of recovery does not modify\nthe Secretary's obligation to make payments for a loan that has been sold and guaranteed.\n(b) A loan becomes due and payable immediately if a facility for which loan funds\nhave been used is sold within 20 years after the federally-financed work on it is completed.\nThe loan becomes due if sale is to an entity not eligible for assistance under the section, or\nnot approved by the Secretary, or if the facility ceases to be a public or nonprofit private\nentity eligible for assistance. The Secretary may also subordinate or waive the right of\nrecovery and any other Federal interest based on a loan or loan guarantee for capital\nprojects, if such waiver(s) would further the purpose of serving medically underserved\npopulations.\nSection 3444. Provisions Regarding Construction or Expansion of Facilities. (a)\nThe Secretary may provide loans or loan guarantees for the construction, conversion\nexpansion or modernization of a facility, only if the applicant describes the facility site,\nprovides plans and specifications which meet the Secretary's requirements, and demonstrates\nthat title is vested in one or more of the applicants.\n(b) An applicant for a loan must make the following agreements:\n(1) Title to the site will be vested in one or more of the applicants;\n(2) Adequate financial support is available for completing and maintaining and\noperating the facility;\n(3) The construction contract complies with the Davis Bacon Act (relating to\npayment of laborers); and\nTitle III\n127\n(4) The facility will be available to persons seeking service there, regardless of\ntheir ability to pay.\nSection 3445. Application for Assistance. The Secretary may provide assistance only\nif the applicant files the application in the form and manner prescribed by the Secretary.\nSection 3446. Administration of Programs. The loans and loan guarantees for\ncapital projects must be administered from a centralized unit in the Department of Health and\nHuman Services.\nSUBPART D - ENABLING SERVICES\nSection 3461. Grants for Enabling Services. (a)(1) The Secretary may make grants\nto qualified community health groups (plans and networks) and to other public and private\nnon-profit groups that provide services in one or more health professional shortage areas or\nto medically underserved populations and are experienced in providing services to increase\nthe capacity of individuals to use health services. The grants are to be used to provide\nenabling services.\n(b) Enabling services are transportation, community and patient outreach, patient\neducation, translation services, and other services that would increase the capacity of\nindividuals to use the comprehensive benefits to which the Act entitles them.\n(c) Grants may be made only if the applicant submits information demonstrating the\nneed for the services, a proposed grant budget and evidence of significant community\ninvolvement in the project.\n(d) Grant applicants must agree not to charge fees for grant-funded enabling services.\n(e) Grant applicants must make maximum use of existing resources.\n(f) Applications must be filed in a form and manner prescribed by the Secretary, and\ninclude agreements and assurances deemed necessary by the Secretary.\n(g) Enabling services are services described in subsection (b), when furnished by an\nentity described in subsection (a).\nPART 3 - NATIONAL HEALTH SERVICE CORPS\nSection 3471. Authorization of Appropriations. (a) Funds for carrying out subpart\nII of part D of title III of the Public Health Service Act, and section 3472 of the Health\nSecurity Act, are authorized in the following amounts: $50,000,000 for fiscal year 1995,\n$100,000,000 for fiscal year 1996, and $200,000,000 for each of fiscal years 1997 through\n2000.\nTitle III\n128\n(b) The authorizations are in addition to funds otherwise authorized.\n(c) Funds may be appropriated under this section at any time before the fiscal year\nfor which they are appropriated.\nSection 3472. Allocation for Participation of Nurses in Scholarship and Loan\nRepayment Programs. Of amounts appropriated under section 3471, the Secretary shall\nreserve such amounts as may be needed to ensure that of the aggregate number of persons\nparticipating in the scholarship program or loan repayment program of the National Health\nService Corps (section 338A of the Public Health Service Act), the total proportion of\nindividuals being educated as or serving as nurses increases to 20 percent.\nPART 4 - PAYMENTS TO HOSPITALS\nSERVING VULNERABLE POPULATIONS\nSection 3481. Payments to Hospitals. (a) The Secretary shall make payments to\neligible hospitals from funds made available under this part. The amounts specified in\nsubsection (b) must be made available to the Secretary on behalf of eligible hospitals, but\npayment is not guaranteed to the state in which an eligible hospital is located or to any\nindividual receiving services from the hospital.\n(b)(1) The total amount of the payments is $800,000,000 for the fiscal year in\nwhich the general effective date occurs and for each subsequent year.\n(2) For any year prior to the general effective date, the amount specified shall\nequal the aggregate disproportionate share hospitals (DSH) percentage of the amount\notherwise available under this section. The aggregate DSH percentage is equal to the\npercent of total payments to DSH hospitals in all states represented by the payments\nto DSH hospitals in participating states.\n(c) Hospitals qualifying for payments shall receive them for five years, without regard\nto the first year in which a hospital receives payment.\n(d) Payments shall be made on a quarterly basis.\nSection 3482. Identification of Eligible Hospitals. (a) In order to qualify for\npayments, a hospital must be located in a participating state. However, a qualifying hospital\nmay continue to receive payments even if the state in which it is located is no longer a\nparticipating state.\n(b) States shall identify for the Secretary those hospitals that meet the qualification\ncriteria.\n(c) In order to qualify, a hospital must have a low income percentage caseload (as\ndefined in section 1923(b)(3) of the Social Security Act) during the base year of not less\nTitle III\n129\nthan 25 percent.\nSection 3483. Amount of Payments. (a) Of total amounts available for payment, 75\npercent shall be allocated based on the hospital's low income percentage of the allocation for\nthe year.\n(b) Twenty-five percent of the total shall be allocated to hospitals for services that\nare not covered services under the Act. The Secretary shall develop an allocation\nmethodology.\n(c) An eligible hospital's low income percentage shall equal the amount of all low\nincome days attributable to the hospital. Low income days equal the total amount of\ninpatient days multiplied by the hospital's low income utilization rate under section\n1923(b)(3).\nSection 3484. Base Year. The base year is the year prior to the year of the general\neffective date of this Act.\nTitle III\n130\nSUBTITLE F - MENTAL HEALTH; SUBSTANCE ABUSE\n(H.R. 3600 and S.1757 p. 615)\nPART 1 - FINANCIAL ASSISTANCE\nSection 3501. Authorization of Appropriations. (a) Funds are authorized for the\npurposes of this subtitle in the following amounts: $100,000,000 in fiscal year 1995;\n$150,000,000 in fiscal year 1996, and $250,000,000 for each of fiscal years 1997 through\n2000.\n(b) Of amounts made available, the Secretary shall reserve as much as is deemed\nappropriate for activities under section 3503. Of the remaining amounts, the Secretary shall\nreserve 50 percent for activities described in subsection (a) of section 3502 and 50 percent\nfor activities under subsection (b) of section 3502.\n(c) The amounts authorized above are in addition to any other funds authorized for the\nsame purposes.\nSection 3502. Supplemental Formula Grants for States Regarding Activities\nUnder Part B of Title XIX of the Public Health Service Act. (a) The Secretary shall make\nmental health grants to states that have submitted applications meeting the requirements of\nsubsection (e). The allotment formula shall be the same one used under section 1918(a)(2) of\nthe Public Health Service Act.\n(b) The Secretary shall make grants for substance abuse services to states that have\nsubmitted applications. The amounts of the grants are determined by using the formula\nunder section 1933(a)(1)(B)(i) and section 1918(a)(2)(A).\n(c)(1) The Secretary shall approve grant uses that are consistent with the\nmental health and substance abuse activities that are described in this section. The\nstate must agree to spend grant funds in accordance with the Secretary's approved\nuses.\n(2) Approved uses are as follows:\n(A) transportation and translation, patient and community\noutreach, patient education, and such other services as the Secretary\ndeems appropriate for the purpose of increasing the access of\nindividuals to services relating to mental health and substance abuse;\n(B) improving the capacity of state and local service systems to\ncoordinate and monitor mental health and substance abuse services,\nimproving information systems, and establishing linkages between\nmental health and substance abuse services and primary care providers\nand health plans;\n(C) providing incentives to integrate public and private systems\nfor treatment of mental health and substance abuse treatment systems;\nand\nTitle III\n131\n(D) any activity for which a grant may be made under sections\n1911 and 1921 of the Public Health Service Act.\n(d) As a condition of grant receipt, a state, may not reduce its funding for mental\nhealth and substance abuse activities below the level spent in the fiscal year preceding the\nfirst year for which the state receives the grant. The Secretary may waive this requirement\nif the state agrees to spend the funds that would otherwise be subject to the requirement on\nthe developing community based care systems for the eventual integration of the public and\nprivate systems for treating mental health or substance abuse (as applicable to the grant).\n(e) The state's application must be submitted in a form, time and manner and include\nagreements and assurances as required by the Secretary.\nSection 3503. Capital Costs of Development of Certain Clinics and Centers. (a)\nThe Secretary may make loans and loan guarantees to public and non-profit private entities,\nfor the capital costs of developing non-acute, residential treatment centers and ambulatory\nclinics.\n(b) Priority must be given to loan and loan guarantees for centers and clinics in areas\nwith a health professions shortage or with significant numbers of medically underserved\nindividuals.\n(c) Loans and loan guarantees shall be made only in accordance with procedures in\nsubpart C of Part 2 of subtitle E.\nPART 2 - AUTHORITIES REGARDING PARTICIPATING STATES\nSUBPART A - REPORT\nSection 3511. Report on Integration of Mental Health and Substance Abuse\nSystems.\n(a) As a condition of being a participating state, each state must submit, not later than\nOctober 1, 1998, a plan to achieve the integration of state and local mental health and\nsubstance abuse services with services that are included in the comprehensive benefit plan.\n(b) The state's report shall contain the following information:\n(1) the number of persons served by state and local mental health and\nsubstance abuse systems and the proportion who are eligible persons under title I of\nthe Act;\n(2) services furnished to eligible persons, including each type of benefit\nfurnished, the diagnoses for which the benefits are furnished, the amount, duration\nand scope of coverage of each benefit furnished, applicable limits on benefits, and\ncost-sharing rules that apply;\n(3) the extent to which mental health and substance abuse providers providing\nservices under a state plan participate in alliance health plans and reasons for any lack\nTitle III\n132\nof participation by these providers;\n(4) the amount of revenues from health plans received by mental health and\nsubstance abuse providers that do participate in health plans and that are funded under\none or more state program(s);\n(5) the amount spent by the state and its political subdivisions in each of the\ntwo years before it became a participating state, for items and services covered in the\ncomprehensive benefit package; also, the amount spent on medically necessary care\nnot included in the benefit package, including medical and other health care and\nrelated supportive services;\n(6) an estimate of the amount the state will need to spend on uncovered\nbenefits and services after mental health and substance abuse services are expanded in\nthe year 2001;\n(7) a description of how the state will assure that all eligible individuals served\nby state-funded mental health and substance abuse programs will be enrolled in a\nhealth benefit plan and how mental health and substance abuse services not covered in\nthe benefit package will continue to be furnished;\n(8) a description of the conditions under which integration of mental health and\nsubstance abuse providers into health plans can be achieved and an identification of\nchanges in provider participation and health plan certification requirements that are\nneeded to achieve integration; and\n(9) if integration is not medically appropriate or feasible for one or more\ngroups of individuals treated in state programs, a description of the reasons therefor,\nand a plan for assuring coordination of care and services covered in the benefit\npackage and the state program for these people.\n(c) Reports shall be submitted in a form and manner prescribed by the Secretary.\nSUBPART B - PILOT PROGRAM\nSection 3521. Pilot Program. (a) The Secretary shall establish a pilot program to\ndemonstrate model methods of integrating mental health and substance abuse services with\nthe mental health and substance abuse services covered in the comprehensive benefit package\nunder title I.\n(b) In establishing the pilot program, the Secretary must consider the following\nfactors:\n(1) the types of items and services needed by patients in addition to those\ncovered under Title I;\n(2) optimal methods of treating persons with long term mental illness and\nsubstance abuse conditions;\n(3) the capacity of alliance health plans to furnish such treatment;\n(4) necessary modifications in coverage and services furnished by health plans;\nand\n(5) the role of publicly funded providers in integrating acute and long-term\ntreatment.\nTitle III\n133\nDOLE BILL\n330A\nState block grant program to create or enhance primary care services\nprovided to low income or medically underserved populations.\nFunding Allocation\ntotal funds = State need adjusted population (State FMAP) / national need\nadjusted population (national FMAP)\nneed adjusted population = State population (need index)\nneed index = weighted sum of geographic percentage + poverty percentage\n+ multiple grant percentage : general population percentage of State\ngeographic percentage = estimated population of State in non-urban areas\n/ total national non urban population\nnon-urbanized population = 1 - urbanized population of State, expressed as\na percentage of total State population\npoverty percentage = the estimated number of people in the State below\n200% FPL / total number of people below 200% nationally\nmultiple grant percentage = amount of State funding recieved under 329\n330, 340 / total funding recieved under these grants for all States.\n(ceiling of twice the general population percentage and floor of less than\n50% of general population percentage)\nFMAP = 1- State matching percentage\nState matching percentage = .25 (taxable resource % : need adjusted pop)\ntaxable resource percentage = total taxable resources of a State / total\ntaxable resources nationally\nPurpose\nGrants are to be used to deliver primary care services to people without\nother access to those services in a cost effective, geographically\nequitable manner.\n321\nGrants for technical assistance when establishing health networks and\nplans in underserved areas.\nLOOK AT 152 - 155\n331\nLoans to health networks, health plans covering individuals in rural,\nfrontier, or underserved areas, or health care providers serving rural,\nfrontier, or underserved areas\nAssistance can be used for facility improvement (SEE 155 - 8)\n341\nProvision of a tax credit for primary health services providers.\nThe tax credit amount is equal to:\nthe number of months during the year that the taxpayer is a qualified\nprimary health services provider (1,000) or (500) if the person is not a\nphysician.\nQualified primary health services provider is anyone deemed by BPHC.\nSAFETY NET\nThe Evolution\nOf Support For\nSafety-Net Hospitals\nChanges in the medical care marketplace are placing funding\nfor uncompensated care and clinical education provided by\nsafety-net hospitals at risk.\nby Linda E. Fishman and James D. Bentley\nPROLOGUE: Teaching hospitals hold a special place in America's\npluralistic system of health care delivery and financing. They\nare citadels of medical learning and institutions that care for\nsome of our most vulnerable citizens. But they also are the\nplaces where much of the innovation that leads to medical\nprogress is conducted. Because teaching hospitals are complex\n30\nEVOLUTION OF\ninstitutions that perform multiple missions, federal and state\nSUPPORT\ngovernments have woven a complex set of policies that seek to\nbalance public and private interests.\nIn this paper two veterans of federal health policy making\nexamine the rich history and current status of hospital support\nfor financing uncompensated care and graduate medical\neducation. Linda Fishman, who has a reputation in policy\ncircles as a level-headed realist, is associate vice-president in\nthe Office of Governmental Relations of the Association of\nAmerican Medical Colleges (AAMC). Fishman holds two\ndegrees from the University of Washington, where she was\nelected to Phi Beta Kappa. Fishman recently produced a\nmonograph for the AAMC that is essential reading for anyone\nwho wants to understand the complexities of Medicare's\nfinancing of graduate medical education. James Bentley, senior\nvice-president of the American Hospital Association (AHA),\nhas been instrumental in shaping federal policy as it applies to\ngraduate medical education and hospital financing. Bentley\ndirects the AHA's public policy analyses as well as related\nactivities. He holds a doctorate in medical care organization\nfrom the University of Michigan. He spent ten years at the\nAAMC before joining the AHA.\nHEALTH AFFAIRS\nVolume 16, Number 4\nc 1997 The People-to-People Health Foundation, Inc.\nEVOLUTION OF SUPPORT\nABSTRACT: The federal government, mostly through the Medicare and Medicaid\nprograms, has created and maintained a set of structural mechanisms to sup-\nport uncompensated care and clinical education: disproportionate-share hospi-\ntal payments and direct and indirect graduate medical education payments. This\npaper provides a history of how these traditional supports have evolved. We note\nthat the need to reduce federal and state spending threatens the level of these\npayments, while changes in the health care delivery system highlight a range of\ndesign and technical inadequacies in the current support mechanisms.\nHROUGHOUT THE TWENTIETH CENTURY, the U.S. health\nT\ncare system has supported uncompensated care and graduate\nmedical education (GME) through a complex patchwork of\nrevenue generated by patient care. Hospitals, the sites where most\nuncompensated care and physician training occur, have traditionally\ncharged privately insured patients more than the cost of their hospi-\ntal care. More recently, the federal government, through the Medi-\ncare and Medicaid programs, has created and maintained a set of\nstructural support mechanisms based on patient care payments to\nhelp finance uncompensated care and health professions clinical edu-\ncation, particularly GME. Now the health care system is transforming\nfrom one based on a delicate web of confusing cross-subsidies to a\nsystem based on price competition in which both private and public\nSAFETY NET\n31\npurchasers want to pay only for the cost of the services their en-\nrollees receive. Pressure to curb the rate of growth in state and\nfederal health care spending threatens to erode the existing public\nsupport mechanisms for uncompensated care and GME.\nThe question in the current competitive environment is whether,\nhow, and to what extent society will continue to support the addi-\ntional roles of hospitals that now are funded with patient care reve-\nnue. Although other types of health care providers render uncom-\npensated care and participate in health professions education, most\nof the public structural support mechanisms have been designed\nspecifically as institutional- or hospital-level payments. This paper\nexamines the history and current structure of hospital support for\nfinancing uncompensated care, GME, and its related activities. Al-\nsthough the competitive environment does not yet jeopardize the\nhospital community overall, certain types of hospitals, which form a\nsafety net of care for the poor and/or provide GME and cutting-edge\nresearch, are at risk.¹\nA Historical Perspective\nCare for the poor. The evolution of support for uncompensated\ncare and GME closely parallels the development of the hospital.\nHospitals in the United States, first built primarily for the poor,\nwere organized as charities under the sponsorship of religious or-\nMEALTH AFFAIRS - July/August 1997\nSAFETY NET\nganizations and wealthy patrons. By the late nineteenth century,\nhospitals' orientation had changed from charitable institutions to\nbusinesses as they began attracting patients from all socioeconomic\nclasses. Today, public, private, and proprietary hospitals all, to vary-\ning degrees, serve their communities and provide care to persons\nwho cannot pay.\nClinical education and medical technology centers. Some\nhospitals also invest resources in a variety of other medical prod-\nucts. As scientific knowledge exploded in the twentieth century,\nsome hospitals developed close affiliations with medical schools,\nbecoming centers of advanced medical technology and providing the\nsettings for clinical education with organized teams of attending\nphysicians, residents, and students. In these institutions, commonly\nknown as \"teaching hospitals,\" education and research are con-\nducted simultaneously with clinical hospital practice.\nFinancing: private and public. As hospitals evolved from\ncharitable institutions to complex business enterprises, their reli-\nance on patient care revenue increased. In 1922 patient care revenue\naccounted for 65.2 percent on average of the total revenue of general\nhospitals.² In 1994, after the growth of private insurance and intro-\n32\nEVOLUTION OF\nduction of Medicare and Medicaid, 94 percent of hospital revenue\nSUPPORT\non average was derived from services to patients.³\nPayments from private payers indirectly assist hospitals in meet-\ning the costs of uncompensated care and GME-related activity. To\nlarge and varying degrees, hospitals \"cost shift,\" or obtain revenues\nin excess of costs from one payer of service to offset shortfalls in\nother categories.\nHospitals also rely on a set of publicly funded structural support\nmechanisms (Exhibit 1). Some federally appropriated funds, accessi-\nble under Titles VII and VIII of the Public Health Service Act, sup-\nport a variety of clinical education programs. Other appropriated\nfunds are available to hospitals to support certain at-risk popula-\ntions, such as patients with acquired immunodeficiency syndrome\n(AIDS), migrant workers, and the homeless, but the bulk of public\nsupport is embedded in the Medicare and Medicaid programs.\nMedicare makes special payments to hospitals, called dispropor\ntionate-share hospital (DSH) payments, direct graduate medical\neducation (DGME) payments, and indirect medical education\n(IME) payments, to maintain access to care for its beneficiaries and\nto support GME and its related costs. The Medicaid program,\nfi\nnanced through a federal/state partnership, also makes DSH pay?\nments to hospitals to ensure access to care and, in many states,\nsupports GME and its related costs.\nThese public support mechanisms are in jeopardy for two real\nHEALTH AFFAIRS Volume 16, Number 4\nEVOLUTION OF SUPPORT\nEXHIBIT 1\nPublic Structural Supports For The Health Care Safety Net\nFY 1996\namount\n(billions of\nType of support\ndollars)*\nPurpose\nMechanism\nMedicare\nDirect graduate medical\n$2.4\nDirect costs of graduate\nNumber of full-time-\neducation paymentsᵇ\nmedical education\nequivalent residents\nIndirect medical education\n4.3\nDifferences in patient\nIntern- and resident-\npayments\ncare costs/access\nto-bed ratio\nDisproportionate-share hospital\n4.3\nDifferences in patient\nSupplemental\npayments\ncare costs/access\nSecurity Income and\nMedicaid inpatient\ndays\nMedicald\nDisproportionate-share hospital\n19.0\nAssistance to hospitals\nVaries\npayments\nfor services to low-\nincome and Medicaid\npatients\nClinical education payments\n1.0+\nDirect costs of graduate\nVaries\nmedical education\nand/or indirect patient\ncare costs related to\nteaching\nFederally appropriated fundsc\n2.0\nVaries\nAppropriations/grants\nSOURCE: Department of Health and Human Services.\nNOTE: Medicare and Medicaid funding estimates are for fee-for-service spending and exclude amounts embedded in managed\ncare rates where applicable.\na\nMedicare funding estimates from the Congressional Budget Office, January 1997 baseline. Medicaid disproportionate-share\nhospital estimate is for 1995; see J. Holahan and D. Liska, \"Where Is Medicaid Spending Headed?\" (Washington: The Urban\nInstitute for The Kaiser Commission on the Future of Medicaid, December 1996). Medicaid estimate for clinical education\napending is for 1995; see D. Plumb and T. Henderson, Medicaid Funding of Graduate Medical Education: A Survey of the States\n(Washington: The Intergovernmental Health Policy Project at The George Washington University, October 1995). Federally\nappropriated funding levels from the Health Resources and Services Administration overview in the proposed U.S. federal budget\nfor FY 1998.\nIncludes payments for nursing and allied health training programs.\nIncludes those programs in Titles VII and VIII of the Public Health Service Act, Ryan White funds, consolidated health centers,\nNational Health Service Corps, rural health programs, and a few other programs.\nsons. First, federal and state governments are seeking large targets\nto help them meet increasing budgetary constraints, and these spe-\ncial payments are substantial. Second, the structural support pay-\nments are based on fee-for-service payment systems and inpatient\nuse of hospital services. Managed care uses different payment meth-\nods, reduces hospital use, and moves care to lower-cost and ambula-\ntory settings. This dynamic distorts and weakens the methodologies\ncalculating these hospital payments.\nPALT AFFAIRS July/August 1997\nSAFETY NET\nUncompensated Care\nAmount. Hospitals and physicians traditionally have cared for\npatients regardless of their ability to pay. In this paper, uncompensated\ncare is charity care and bad-debt expense. To a limited extent, all types\nof providers provide some uncompensated care, but data on total un-\ncompensated care costs generally are available only from hospitals.\nHospitals provided about $16.8 billion in uncompensated care in 1994,\n19 percent of which was offset by government subsidies.⁴\nCurrent financing. Hospitals finance uncompensated care\nthrough a variety of revenue sources. They may receive funds from\nfederal, state, or local grant programs for specific services or special\npopulations; philanthropies; gifts or their own charitable activities;\nand earnings from the fiscal year. Public municipal or state-owned\nhospitals receive state or local government appropriations, as do\nsome private institutions. The federal government, through the\nMedicare and Medicaid programs, has implicitly supported uncom-\npensated care by targeting additional funds toward certain types of\nhospitals that serve large numbers of poor persons. These payments\ninclude two separate and distinct DSH adjustments: one under the\nMedicare prospective payment system (PPS), and another under the\n34\nEVOLUTION OF\nSUPPORT\nMedicaid program. Some policymakers also believe that the current\nlevel of the Medicare IME adjustment is justified by the uncompen-\nsated care losses incurred by teaching hospitals.\nMedicare DSH adjustment. In April 1986, with passage of the\nConsolidated Omnibus Budget Reconciliation Act of 1985 (COBRA)\n(P.L. 99-272), Congress mandated an explicit payment adjustment\nin the PPS for hospitals that serve large numbers of low-income\npatients. It required no new money. Funding was obtained by low-\nering the basic rate paid to hospitals and decreasing the level of the\nIME adjustment in recognition that teaching hospitals would re-\nceive a large proportion of DSH payments.\nThe original rationale for the DSH adjustment was to compensate\nhospitals for the higher operating costs they incurred in treating\ndisproportionately large numbers of low-income patients. Several\nstudies, conducted by the Health Care Financing Administration\n(HCFA) and the American Hospital Association (AHA), had dem\nonstrated a relationship between higher Medicare costs and the\npercentage of a hospital's patients covered by Medicare or Medic-\naid.5 A 1984 study by the Congressional Budget Office (CBO), based\non 1981 data, showed that certain groups of hospitals, particularly\nthose with relatively large shares of Medicaid patients, more than\n100 beds, and large-city locations, would do worse, on average, un-\nder the new Medicare PPS than would other hospitals, because\nHEALTH AFFAIRS Volume 16, Number 4\nEVOLUTION OF SUPPORT\ntheir higher costs.⁶ By 1990 another CBO analysis of more recent\n(1987) data showed that except for urban hospitals with more than\n100 beds and the highest levels of service to the low-income poor,\nthe higher cost differences associated with serving the poor had\ndisappeared, and there was little justification for a DSH adjustment\nbased on differences in costs.⁷\nHowever, the CBO noted another rationale for the DSH adjust-\nment. Congress had become increasingly concerned that certain\nhospitals were at risk of closing as a result of treating large numbers\nof poor patients and began to view the DSH payment as a mecha-\nnism for mitigating hospitals' financial distress. Under this ration-\nale, DSH payments were justified because they aided hospitals in\nmaintaining access to care for low-income Medicare beneficiaries\nand other patients. Today, most policymakers acknowledge that the\nDSH adjustment is more a mechanism for channeling payments to\nhospitals that serve a high proportion of poor patients than it is a\nmeans of compensating hospitals for differences in operating costs.\nMechanics of the DSH adjustment. A qualifying hospital\nreceives a DSH payment for each Medicare patient it treats under\nthe PPS. The DSH payment is calculated as a percentage add-on to\nthe basic diagnosis-related group (DRG) payment. Different DSH\nSAFETY NET\n35\nformulas are used, depending on where the hospital is located, how\nmany beds it has, and its status as a rural referral center or sole\ncommunity provider. The value of the hospital's DSH \"index\" deter-\nmines the hospital's eligibility for a DSH payment and the size of the\npayment. The index, whose definition has not changed since the\noriginal legislation, is the sum of two ratios: the proportion of all\nMedicare days that are attributable to beneficiaries of Supplemental\nSecurity Income (SSI), a means-tested cash benefit program for aged\nand disabled people, and the proportion of all patient days for which\nMedicaid is the primary payer.8\nMedicare DSH expenditures. Since its enactment in 1986, the\nDSH payment has been modified in every budget reconciliation act\nexcept OBRA 1993 (P.L. 103-66). Originally intended to sunset, the\nadjustment was made a permanent part of the PPS in 1990. Each bill\nadded more money to the adjustment but not always to every cate-\ngory of hospital. Legislation passed in 1990 (P.L. 101-508) added the\nmost money to the adjustment, almost $1 billion, through formula\nmodifications. Medicare DSH spending has increased almost four-\nfold, from about $1 billion in 1989 to more than $4 billion in 1996,\nand has grown faster than overall inpatient hospital payments.9 The\nCBO estimates that Medicare DSH payments will total $4.5 billion\nin fiscal year (FY) 1997, increasing to $4.8 billion in FY 1998.\nAs DSH spending has increased, so has the number of hospitals\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\nreceiving DSH payments. In FY 1991, the Prospective Payment As-\nsessment Commission (ProPAC), which monitors the PPS for Con-\ngress, estimated that 1,558 hospitals, or 28 percent of all PPS hospi-\ntals, received DSH payments.\" By FY 1996, ProPAC found that 1,957\nhospitals, or 38 percent of all PPS hospitals, were receiving DSH\npayments.\"\nAlthough almost 2,000 hospitals receive this adjustment, Medi-\ncare DSH payments are highly concentrated. Ninety-three percent\nof total DSH payments go to large hospitals in urban areas, and\nteaching hospitals receive about 65 percent of all DSH payments.\nFinally, because Medicaid eligibility and coverage vary widely\nacross states, Medicare DSH payments are distributed unevenly\nacross geographic areas: The Middle Atlantic, South Atlantic, and\nPacific regions account for 60 percent of all DSH payments but only\n46 percent of Medicare discharges.\"\nMedicaid DSH payments. The Medicaid DSH payment is\nbased on the assumption that certain hospitals, in addition to pro-\nviding care to Medicaid enrollees, also serve indigent persons who\nare not eligible for Medicaid and maintain many public health and\nsocial services for all area residents. Legislators recognized that\n36\nEVOLUTION OF\nthese hospitals could not shift the cost of uncompensated care to the\nSUPPORT\nrelatively few privately insured patients they serve. Congress took\naction so that access to care could be maintained. OBRA 1981 (P.L.\n97-35) required states to \"take into account the situation of hospi-\ntals which serve a disproportionate number of low income patients\nwith special needs\" when setting inpatient hospital payment rates.¹⁴\nAt first, states were slow to establish-DSH payment adjustments,\nbut in the late 1980s the federal government stimulated the creation\nof state DSH programs through legislation and regulation. In 1987\nbudget reconciliation legislation (P.L. 100-203), Congress estab-\nlished minimum criteria for designating and paying DSH hospitals\nand required states to designate every hospital that met those crite-\nria as a DSH hospital. States could be more generous in their desig-\nnation criteria or in their payment levels. This led to great variation\nacross states and even among hospital types within states, as many\nstates went beyond the minimum criteria.\nIn the late 1980s and early 1990s, Medicaid DSH payments ex-\nploded. States became very creative in increasing their Medicaid\nfunding via provider-specific taxes, intergovernmental transfers,\nand donations from hospitals as part of the state share of Medicaid\nspending. These strategies increased federal payments to states\nwith little impact on state general revenue funds. For example,\nproviders, usually hospitals, would pay a tax or donate funds to the\nstate. The state would then use these funds to make Medicaid pay-\nHEALTH AFFAIRS Volume 16, Number 4\nEVOLUTION OF SUPPORT\nments and receive a matching payment from the federal government.\nThe federal government objected to this because it believed that the\nmechanisms increased the federal contribution inappropriately.\nIn December 1991 Congress passed legislation, the Medicaid Vol-\nuntary Contribution and Provider-Specific Tax Amendments of\n1991 (P.L. 102-234), as a compromise between the federal govern-\nment and the states. It restricted the types of provider taxes that\nstates could use and banned the use of provider donations. The law\nalso placed caps on DSH payments, limiting them to 12 percent of\nprogram expenditures at both the national and state levels. States\nwhose DSH payments were more than 12 percent of state spending,\nso-called high-DSH states, were frozen at 1993 levels until the rest of\ntheir Medicaid expenditures grew so that DSH payments were 12\npercent or less of Medicaid spending. States where DSH spending\nwas below 12 percent were allowed to grow at the same rate as the\nrest of the Medicaid program. In 1993 Congress, responding to infor-\nmation that some hospitals were receiving excess DSH funds and\nthat some states were diverting federal matching funds for purposes\nother than health care, placed further restrictions on particularly\nlarge Medicaid DSH payments to certain hospitals.\nCurrent rules for Medicaid DSH payments. The 1991 and\nSAFETY NET\n37\n1993 legislative provisions for Medicaid DSH payments remain in\neffect. Federal matching payments for state Medicaid spending that\nis financed with revenues from provider taxes continue to be\ncapped, and the 12 percent limit is still applied. States use a variety\nof methods to make DSH payments, but all are tied in some way to\nservice for Medicaid enrollees and low-income persons.\nMedicaid DSH expenditures. Between 1988 and 1992 Medic-\naid expenditures grew on average 22.4 percent annually, while DSH\nspending grew a whopping 149.9 percent annually, from $400 mil-\nlion in 1988 to $17.5 billion in 1992. 15 Legislative changes to the DSH\nprogram in 1991 and 1993 moderated the growth of DSH payments\nand overall program spending. Between 1992 and 1995 Medicaid\nDSH spending grew only 2.7 percent per year, increasing to $19\nbillion by 1995, considerably slower than the overall growth of the\nMedicaid program, which had dropped by then to 9.5 percent per\nyear. 16 As a result, whereas in 1992 DSH payments were 14.6 percent\nof all Medicaid spending, in 1995 DSH payments had dropped to\nabout 12 percent of total Medicaid spending.\nOther federal government support. The federal government\nprovides support for special populations through certain programs\nin the Health Resources and Services Administration (HRSA).\nWithout the availability of federal funds for programs such as the\nRyan White AIDS program, migrant health centers, and health care\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\nfor the homeless, hospitals would have to fund such activities with\nother revenues obtained through cross-subsidization. These feder-\nally appropriated funds, about $2 billion in FY 1996, are available to\nthe health care system through grants and appropriations.\"\nNonfederal government support. Another important source\nof support for uncompensated care is state and local government\nappropriations. State, county, or city tax dollars may be the primary\nsource of funds for certain types of hospitals, such as public general\nhospitals. Data on appropriations are difficult to collect at the na-\ntional level, but among members of the Council of Teaching Hospi-\ntals and Health Systems (COTH), twenty-three public municipal\nteaching hospitals reported receiving $1.7 billion in state, county,\nand city support for indigent care and general operations in FY\n1995. Another thirty-eight state-owned public teaching hospitals\nreceived $700 million from state and county appropriations for indi-\ngent care, general operations, and medical education purposes.\nClinical Education\nMedical education in the twentieth century has been intertwined\nwith care for the poor, particularly in public hospitals where medical\n38\nEVOLUTION OF\neducation programs have provided a workforce for care of indigent\nSUPPORT\npatients. Until the 1960s residents gained their primary clinical expe-\nrience in public hospitals and on the charity wards of voluntary\nhospitals. In the mid-1960s and 1970s enactment of the Medicare and\nMedicaid programs sharply reduced the number of indigent patients,\nand GME changed, expanding training to other clinical settings.\nHowever, since the clinical component of medical education involves\nstudents and residents in the direct diagnosis and treatment of pa-\ntients, and the poor have represented a substantial portion of pa-\ntients cared for in educational settings, many view support for GME\nas support for uncompensated care in underserved communities.\nThe cost of clinical education. Teaching hospitals, which\nsponsor GME programs, incur significant direct and indirect costs\nin operating physician training programs. The direct costs consist of\nstipends and fringe benefits for residents, salaries and fringe bene-\nfits for supervising faculty, costs directly associated with support-\ning the GME program (such as the clerical personnel working exclu-\nsively in the GME administrative office), and allocated institutional\noverhead costs (such as maintenance, cafeteria, and depreciation).\nIndirect costs are those incurred in providing an appropriate en-\nvironment for clinical education. They include the higher patient\ncare costs that accompany an academic infrastructure because\nteaching institutions tend to treat a much higher proportion of se-\nverely ill patients who require intensive resources. Teaching hospi-\nHEALTH AFFAIRS Volume 16, Number 4\nEVOLUTION OF SUPPORT\ntals also maintain a broader scope of highly specialized services and\nstand-by capacity, often on an around-the-clock, regional basis.\nHigher patient care costs also may result when a teaching program\nis located, for example, in the central core of a large urban\narea-where labor, land, and operating costs may be higher than in\nthe suburbs. Finally, indirect costs include the reduced productivity\nof the hospital staff because they are educating residents and the\nprocessing of additional diagnostic tests or ancillary services that\nresidents may order during their clinical learning experience.\nCurrent support for clinical education. Support for clinical\neducation comes from many sources, but most of it comes from\nhospital patient service revenue, such as payments from individuals;\nthird-party payers, such as commercial insurance companies;\ngovernment-financed programs, such as Medicare and Medicaid;\nand state and local appropriations. Fees from faculty physician prac-\ntices, foundation grants, grants from the National Institutes of\nHealth, and other diverse nonhospital sources, in addition to the\nDepartments of Veterans Affairs (VA) and Defense, also support\nGME. The VA is the largest single provider of physician training\nsites in the United States today and funds about 9 percent of all\nresidency positions each year at about 130 VA medical centers.\nSAFETY NET\n39\nBecause the Medicare program makes two explicit payments to\nteaching hospitals for their direct and indirect costs, many persons\nmistakenly believe that Medicare is the only payer of GME-related\ncosts. Other purchasers of health services also participate in GME\nfinancing. Medicaid makes payments through a variety of explicit\nand implicit mechanisms. Private insurance companies contribute\nimplicitly to direct and indirect GME costs by paying higher prices\nthan these companies would pay to nonteaching hospitals.\nMedicare DGME payments. Medicare DGME payments com-\npensate teaching hospitals for the costs that are directly related to\nthe graduate training of physicians, dentists, and podiatrists.¹⁹\nMedicare does not pay the costs of clinical undergraduate medical\neducation, although it occurs in teaching hospitals, usually along-\nside and intertwined with residency training. In establishing Medi-\ncare in 1965, Congress recognized the need to support residency\ntraining programs to meet the nation's need for fully trained health\ncare professionals and acknowledged that educational activities\nheighten the quality of care in hospitals.²⁰\nMechanics of the DGME payment. From 1965 until 1985\nMedicare paid its share of each hospital's historical DGME costs.\nReimbursement was open-ended: If a hospital increased its costs,\nMedicare paid its share of the costs incurred. However, COBRA\n1985 dramatically changed the DGME payment methodology in\nHEALTH AFFAIRS - July/August 1997\nSAFETY NET\nApril 1986 by uncoupling the link between costs and payments.\nToday, Medicare pays a portion of a hospital-specific per resident\namount, which is updated annually by an inflation factor. These\ntotal per resident amounts vary widely, and in the FY 1984 or FY\n1985 period on which the per resident amounts are based, they\nranged from $20,000 to more than $100,000. Medicare pays a per-\ncentage of the per resident amount based on its share of total inpa-\ntient days in each hospital. In recent years the per resident amount\nhas been adjusted to pay a higher rate for primary care residents.\nThe COBRA legislation also limited the number of years for which\nMedicare would fully support its share of residency training, and in\nAugust 1993 Congress made additional changes. Today, after the\nperiod required for a resident's initial board certification in a spe-\ncialty, Medicare pays only 50 percent of its share of the per resident\namount. The program imposes no limit on the number of residents it\nsupports, either at an individual hospital or in the national aggregate,\nas long as the residents are enrolled in an approved training program.\nHospitals may receive payments for residents who are graduates of\nU.S. medical, osteopathic, dental, and podiatric schools and for train-\nees who have graduated from foreign medical schools.\n40\nEVOLUTION OF\nMedicare DGME expenditures. According to CBO estimates,\nSUPPORT\nthe Medicare DGME payment totaled about $2.4 billion in FY\n1996.2 Included in this estimate are about $300-$350 million in\npayments to hospitals for a portion of the direct costs of hospital-\nbased nursing and allied health professions education.\nThe Medicare IME adjustment. The IME adjustment, part of\nthe Medicare PPS, compensates teaching hospitals for their higher\nMedicare inpatient operating costs relative to nonteaching hospi-\ntals. The roots of the IME adjustment lie in the limits placed on\nroutine hospital costs in the 1970s. Even though Medicare initially\nreimbursed all Medicare-allowable hospital costs, the federal gov-\nernment soon imposed limits on acceptable costs. Section 223 of the\nSocial Security Amendments of 1972 (P.L. 92-603) authorized the\nsecretary of health, education, and welfare to set payment limits on\nroutine inpatient hospital costs. The cost limits, intended to reduce\nthe variation in hospital costs acceptable to Medicare and the\nmethod of setting them, evolved between 1974, when the regulations\nwere first published, and 1979. As the cost limits became more strin-\ngent, federal policymakers and the teaching hospital community\nexpressed concern that teaching hospitals were being dispropor-\ntionately harmed because the limits did not initially recognize the\ncosts associated with operating an educational program. Eventually,\nteaching hospitals were permitted to remove their DGME costs\nbefore determining whether their costs were below the limits.\nHEALTH AFFAIRS - Volume 16, Number 4\nEVOLUTION OF SUPPORT\nThe concept of indirect costs was recognized in 1980.² Govern-\nment researchers, in studying the relationship between hospital\ncosts and teaching status, found that even after removing DGME\ncosts, teaching hospitals more often reached or exceeded their cost\nlimits than nonteaching hospitals did. Researchers noted that a hos-\npital's intern- and resident-to-bed (IRB) ratio was related to an\nincrease in hospital costs. As a result, the Medicare routine-cost\nlimits for teaching hospitals were increased to incorporate a differ-\nential based on the IRB ratio in each hospital. In 1982 the adjustment\nfor teaching hospital costs was included in the extension of the\nroutine-cost limits to cover total hospital operating costs.\nIn December 1982, when the secretary of health and human serv-\nices proposed a new Medicare payment system for hospitals, the\nresident-to-bed adjustment to the routine-cost limits was con-\nverted to a PPS payment, the IME adjustment, to adjust for the\nhigher costs of teaching hospitals.23 The secretary's estimate indi-\ncated that Medicare operating costs per case increased approxi-\nmately 5.79 percent with each 10 percent increase in the number of\nresidents per bed. However, two months after the secretary's report,\nthe CBO presented an impact analysis showing that the proposed\nDRG-based payment system would have adversely affected 71 per-\nSAFETY NET\n41\ncent of teaching hospitals if the IME adjustment were set at the 5.79\npercent level. The administration then suggested that the estimate\nbe doubled to 11.59 percent for each 10 percent increase in the IRB.\nCongress supported this modification, and the IME adjustment was\nincorporated into the prospective payment legislation.²⁴\nAs more information became available, the IME adjustment was\nrecalculated and lowered. The original adjustment of 11.59 percent\nwas reduced to 8.7 percent in 1986 when better data became avail-\nable. However, the 8.7 percent adjustment factor was reduced by 0.6\npercentage point to finance part of the DSH adjustment, resulting in\nan IME adjustment factor of 8.1 percent during the time that the\nDSH adjustment would be in effect. The 8.1 percent IME adjustment\nrecognized that teaching hospitals would receive a large share of\nDSH payments. The current 7.7 percent IME adjustment, enacted in\nOBRA 1987 (P.L. 100-203), took effect 1 October 1988.\nMechanics of the IME adjustment. For every Medicare case\npaid under the PPS, a teaching hospital receives an additional pay-\nment. A teaching hospital's IME payment is determined by inserting\nits IRB into a formula that is written in statute. In FY 1996 a hospital\nwith five residents for every 100 beds (IRB = 0.05) received a 3.77\npercent add-on payment for each PPS case. A hospital with fifty\nresidents for every 100 beds (IRB = 0.50) received a 33.73 percent\nadd-on to its basic DRG payment.\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\nplans, the program pays 95 percent of an \"up-front\" monthly per\ncapita (capitation) amount (the adjusted average per capita cost\n[AAPCC]) directly to the health plan. The plan's capitation pay-\nment includes what Medicare traditionally spends on DSH, DGME,\nand IME payments under the fee-for-service payment system.\nThe managed care plan then contracts with hospitals and physi-\ncians to provide services. However, instead of using Medicare fee-\nfor-service payment methods, the risk plan negotiates with provid-\ners, including hospitals. The rates that the plan negotiates with the\nhospital do not necessarily include the DSH, DGME, or IME pay-\nments that would be made to the hospital if the beneficiary re-\nmained in the fee-for-service system. Alternatively, the risk plan\nmay direct patients away from the teaching or DSH hospital to a\nlower-cost site of care because the plan receives the same capitation\nrate regardless of the provider with whom it has a contract. In either\ncase, under a risk contract, the teaching/DSH hospital would not\nreceive the DSH, DGME, or IME payment. These earmarked funds\nmay be used by the risk plan for purposes other than those intended\nby Congress. Some policymakers and advocacy groups have pro-\nposed separating the payments from the calculation of the managed\n44\nEVOLUTION OF\ncare rates and paying them directly to a teaching or DSH hospital\nSUPPORT\nwhen the facility serves a Medicare risk-plan patient.\nMedicaid managed care programs pose the same problem. In gen-\neral, states set managed care rates using fee-for-service historical\nclaims data. Unless removed before calculating the health mainte-\nnance organization (HMO) rate, hospital payments for DSH and\nGME-related costs are included in the capitated rates the state pays\nto managed care plans, which are not required to distribute the\nfunds to hospitals. A few states (New York and Michigan, for exam-\nple) have created mechanisms for \"carving out\" the GME-related\ndollars from managed care rates.\nProblems with current support mechanisms. The transfor-\nmation of the health care delivery system highlights a range of design\nand technical inadequacies with the current government-funded\nsupport structures for uncompensated care and GME. These mecha-\nnisms base the level of funding on measures of Medicare or Medicaid\ninpatient hospital use. As states move Medicaid recipients into man-\naged care plans, the identification of these persons for purposes of\ncalculating both types of DSH payments under fee-for-service be-\ncomes more difficult. Because these payments use inpatient days or\ndischarges to distribute funds, managed care's emphasis on reduc-\ning inpatient utilization eventually will result in diminished public\nsupport for uncompensated care and GME. Finally, these payments\nare targeted only to hospitals when the delivery system is shifting to\nHEALTH AFFAIRS Volume 16, Number 4\nEVOLUTION OF SUPPORT\n\"Managed care's emphasis on reducing inpatient utilization eventually\nwill result in diminished support for uncompensated care and GME.\"\nambulatory sites of care and nonhospital providers.\nTo some degree, all support mechanisms use proxies rather than\ndirect measures. Medicaid activity represents service to the poor.\nThe number of residents in the IME's resident-to-bed ratio serves as\na surrogate for differences in inpatient operating costs. Because they\nuse proxies, the payments may not be targeted to the appropriate\ninstitutions. In addition, the purpose of the payments may be com-\nmingled, such as IME payments ensuring access to medical care.\nEvolution of the competitive market. The transition to a\ncompetitive health care market has additional implications for mu-\nnicipal hospitals that traditionally have served the poor. As the mar-\nket constrains prices for all hospitals, Medicaid rates become more\nattractive to other hospitals. Public hospitals that have relied on\nMedicaid patients as sources of revenue must now compete with\nprivate hospitals for these same patients. Municipal hospitals may\nbe unable to compete because they offer fewer amenities or have\nSAFETY NET\n45\nmore unattractive facilities or locations than their competitors have.\nUnder Medicaid managed care, former patients may seek or be di-\nrected to other providers, leaving municipal hospitals with an even\nless desirable patient mix. In response to the fragile situation of\nthese hospitals, local governments are making decisions to sell tax-\nsupported hospitals or further reduce their support.\nAnother sign of increased competition is the relatively recent\nproliferation of niche providers, such as ambulatory surgical centers\nor cancer management companies. Niche providers siphon off the\nmore profitable lines of traditional hospital services, leaving to hos-\npitals the responsibility of providing costly, often unprofitable, serv-\nices to very sick populations.\nThe competitive market also exposes the need to identify the\ninstitutional costs related to clinical research and the provision of\ncommunity services, both of which traditionally have been funded\nwith patient care revenue. As with medical education, there are\nadded patient care costs and lost efficiencies because of the kind of\nenvironment needed to conduct clinical research. Hospitals now\nfinance some research through contributions to medical schools,\nwhich may support the conduct of unsponsored clinical research.\nFinally, hospitals provide community services, ranging from poi-\nson control to Meals-on-Wheels programs. These services are now\nfinanced from patient care revenues, gifts, grants, and retained earn-\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\nings and may be the first types of programs that hospitals reduce or\neliminate when confronted with financial difficulty.\nConclusion\nIn the changing medical care marketplace, if government is unwill-\ning to support the uncompensated care, clinical education, research,\nand community service missions of hospitals, private payers will feel\ncomfortable avoiding these costs. No one opposes hospitals' provid-\ning these services, but few see a responsibility to pay a price differ-\nential to support them. In this environment government policies set\na benchmark for appropriate behavior and provide critical revenues.\nFederal and state initiatives to control health care costs, balance\nbudgets, or extend trust funds place payments for uncompensated\ncare and clinical education at risk.\nThe missions of uncompensated care and GME will not be pre-\nserved simply by increasing hospital efficiency. New and modified\nstructures to finance uncompensated care and clinical education\nmust be developed to preserve these functions. Action today can\npreserve the critical strengths of the U.S. health care system for\ntomorrow. Support structures for these missions are easy to destroy\n46\nEVOLUTION OF\nbut costly and devilishly difficult to rebuild.\nSUPPORT\nThe views expressed in this paper are those of the authors and do not necessarily represent\nthose of their respective associations. The authors acknowledge Richard M. Knapp and\nseveral reviewers for their constructive comments. A portion of this paper was presented\nat The Robert Wood Johnson/Health Affairs/Alpha Center conference, \"What Is\nHappening to the Safety Net?,\"9 January 1997, in Washington, D.C.\nNOTES\n1. See L.E. Fishman, \"What Types of Hospitals Form the Safety Net?\" in this\nvolume of Health Affairs for a general description of safety-net hospitals.\n2. Public appropriations, endowment income, donations, and other sources con-\nstituted the remainder. See P. Starr, The Social Transformation of American Medicine\n(New York: Basic Books, 1982), 161.\n3. Personal communication from Lloyd Wackerling of the Trend Analysis Group\nat the American Hospital Association (AHA), using data from the AHA's\nAnnual Survey of Hospitals, 1994.\n4. Prospective Payment Assessment Commission, Medicare and the American Health\nCare System, Report to the Congress (Washington: ProPAC, June 1996), 48.\n5. Congressional Budget Office, Medicare's Disproportionate Share Adjustment for Hos-\npitals (Washington: CBO, May 1990).\n6. Nancy M. Gordon, assistant director for human resources and community\ndevelopment, Congressional Budget Office, testimony before the Senate Fi-\nnance Subcommittee, 29 July 1985.\n7. CBO, Medicare's Disproportionate Share Adjustment for Hospitals, xiv.\n8. The inpatient days attributable to Medicaid managed care enrollees may be\ncounted for Medicare DSH payment purposes. For a more complete discus-\nsion, see \"Interpretation of Medicaid Days in Medicare DSH Adjustment Cal-\nHEALTH AFFAIRS Volume 16, Number 4\nEVOLUTION OF SUPPORT\nculation,\" HCFA Ruling no. 97-2 (27 February 1997).\n9. ProPAC, Report and Recommendations to the Congress (Washington: ProPAC, 1\nMarch 1997), 30.\n10. Congressional Budget Office, January 1997 baseline estimates.\n11. ProPAC, Medicare and the American Health Care System, Report to the Congress (Wash-\nington: ProPAC, June 1991), 43.\n12. ProPAC, Medicare and the American Health Care System (June 1996), 62.\n13. Ibid., 63-64.\n14. See ProPAC, Analysis of Medicaid Disproportionate Share Payment Adjustments, Con-\ngressional Report C-94-01 (Washington: ProPAC, 1 January 1994); and L. Ku\nand T. Coughlin, \"Medicaid Disproportionate Share and Other Special Fi-\nnancing Programs,\" Health Care Financing Review (Spring 1995): 27-54.\n15. J. Holahan and D. Liska, Where Is Medicaid Spending Headed? (Washington: The\nUrban Institute for The Kaiser Commission on the Future of Medicaid, De-\ncember 1996).\n16. Ibid., 2, Table 1.\n17. Derived from the FY 1998 federal budget overview, which reports actual 1996\nspending.\n18. Association of American Medical Colleges, COTH Survey of Hospitals' Financial\nand General Operating Data, 1995 (Washington: AAMC, Autumn 1996).\n19. See L.E. Fishman, Medicare Payments with an Education Label: Fundamentals and the\nFuture (Washington: Association of American Medical Colleges, 1996). Medi-\ncare also recognizes the direct costs incurred by hospital-based nurse educa-\ntion and allied health professions training programs.\n20. Senate Report 404, Pt. 1, 89th Cong., 1st sess. 36 (1965); and House Report 213, 89th\nCong., 1st sess. 32 (1965).\nSAFETY NET\n47\n21. Federal Register (30 August 1996): 46206-46207.\n22. K. Davis et al., Health Care Cost Containment (Baltimore: The Johns Hopkins\nUniversity Press, 1990), 16-21.\n23. Secretary of the Department of Health and Human Services, Hospital Prospective\nPayment for Medicare: A Report to Congress (Washington: U.S. Government Print-\ning Office, December 1982), 48-49.\n24. J.R. Lave, The Medicare Adjustment for the Indirect Costs of Medical Education: Historical\nDevelopment and Current Status (Washington: Association of American Medical\nColleges, January 1985).\n25. ProPAC, Medicare and the American Health Care System (June 1996), 62-63.\n26. Ibid., 62.\n27. D. Plumb and T. Henderson, Medicaid Funding of Graduate Medical Education: A\nSurvey of the States (Washington: The Intergovernmental Health Policy Project\nat The George Washington University, October 1995).\n28. AAMC, COTH Survey of Hospitals' Financial and General Operating Data, 1995.\n29. 1997 Annual Report of the Board of Trustees of the Federal Hospital Insurance Trust Fund,\nHouse Document 105-73 (April 1997).\n30. ProPAC reported at its January 1997 meeting that in 1995-1996, major teach-\ning hospitals' aggregate PPS inpatient margin was 18.6 percent, compared\nwith 3.7 percent for nonteaching hospitals. Major teaching hospitals continue\nto have the lowest total margins of any hospital group, however, at 3.7 percent,\ncompared with 6.5 percent for nonteaching hospitals in 1995-1996.\nHEALTH AFFAIRS July/August 1997\nHEALTH TRACKING : TRENDS\nthe first four steps are in-\ne mail-only steps used for\na discussion of these tech-\nan, Mail and Telephone Surveys:\nA Changing Picture Of\n(New York: John Wiley and\nher described in a technical\nUncompensated Care\nwork on the Internet at\ne.net/chpvs.\nweighted cross-sectional\nHospitals, physicians, and community health centers all provide\nnd 1995 samples. Inferences\ncharity care. To ignore any one of the groups could distort notions about\nat the cross-sectional sam-\ninal members. There is no\nchanges in access to care for the uninsured.\nble comparisons.\nd 1995 means are calculated\nBY PETER J. CUNNINGHAM AND HA T. Tu\nwithout weighting.\neighted means for the longi-\n993 and 1993-1995 changes\nC\nHANGES IN THE HEALTH CARE sys-\nproblematic, both because they usually serve\ng longitudinal weights.\ntem are causing renewed concern\nother patients and because a substantial\nThe Employee Health Care\nabout access to care for uninsured per-\namount of charity and uncompensated care is\nnd One,\" Health Affairs (Fall\nsons. This paper reviews trends in the amount\ndispensed by providers that serve the general\nof resources that providers are devoting to\nresults is the position that\npopulation. Also, while hospitals have tradi-\nmines satisfaction, not vice\ncare of the medically indigent, as indicated by\ntionally been considered the providers of last\nn is supported by the strik-\nuncompensated care costs. The results of this\nresort for the medically indigent, ignoring\n995 changes on the PCS and\nstudy show that the amount of services pro-\ntrends in physician uncompensated care may\nes of the self-rated outcomes\nvided to the medically indigent has indeed\nresult in false conclusions about changes in\n5 overall sample: For the PCS:\nchanged over the past ten to fifteen years, al-\n-2.24, good -1.34, very good =\naccess. Thus, this study examines trends in\n167\n-0.62. For the MCS: poor\nthough the direction of the change is not con-\nuncompensated care for all hospitals as well\ngood -1.11, very good -0.67,\nsistent across types of providers. Hospital un-\nas physicians and community health centers.¹\nFor further support, see the\ncompensated care costs, which rose rather\nHOSPITALS. Hospitals subsidize care to\non the Internet.\nsubstantially during the early 1980s, have\nthe medically indigent through both direct\nare similar to Exhibit 4's, but\nbeen relatively stagnant during the 1990s,\nIts by level of health. Healthy is\nfunding from public sources-tax revenue,\nes with three or fewer physical\nwhile provision of uncompensated care by\nuncompensated care pools, and Medicare and\nin 1993 and values on the PCS\nphysicians and community health centers has\nMedicaid disproportionate-share adjust-\nSF-12 Health Status survey of\nincreased. In addition, hospital uncompen-\nments-and private philanthropy, as well as\nis defined as patients with four\nsated care costs have become more concen-\nindirectly by shifting the costs of uncompen-\naronic conditions, or PCS ≤ 42,\ntrated among public hospitals and other hos-\nltiple comparisons are not ad-\nsated care onto other payers.\n4 or 5. For documentation on\npitals that provide a disproportionately high\nDespite increases in the number of unin-\nage used for these analyses, see\nlevel of uncompensated care. These findings\nsured persons, there have not been increases\nStata Statistical Software Refer-\nmay have important implications for access to\nin the relative amount of uncompensated care\nn 5.0, Reference P-Z (College\ncare among the uninsured.\nprovided by hospitals during the first part of\nCorp., 1997), 135-137. For docu-\nole selection and scoring for the\nthe 1990s, either in terms of inflation-adjusted\nalth measures, see H. Allen Jr.,\nUNCOMPENSATED CARE TRENDS\ncosts or as a percentage of total costs (Exhibit\nof Health and Health Care: The Cen-\n1).² However, it appears that uncompensated\noston: New England Medical\nThere are identifiable groups of safety-net\ncare costs became somewhat more concen-\nAllen Jr., \"A Core Set of Survey\nproviders-such as public hospitals, some\ntrated among fewer hospitals. The percentage\nlan Consumer Satisfaction: A\nprivate nonprofit hospitals, and community\nto the National Committee for\nhealth centers-that serve a disproportion-\nof hospitals that provided high levels of un-\nin Annual Member Health Care\ncompensated care (10 percent or more of total\nion 1.0 (Washington: National\nately high number of the medically indigent.\ncosts) decreased between 1990 and 1994, al-\nality Assurance, 1995); and J.\nHowever, focusing only on these providers is\nski, and S. Keller, How to Score\nd Mental Health Summary Scales\nPeter Cunningham is a health researcher at the Center for Studying Health System Change in Washington,\nh Institute, 1995).\nD.C. Ha Tu is an analyst at the center.\nm b C 4\nHEALTH AFFAIRS July/August 1997\n1997 The People-to-People Health Foundation, Inc.\nHEALTH TRACKING: TRENDS\nthough this smaller group incurred a larger\ncontract, compared with about 55 percent in\nproportion of total hospital uncompensated\n1990.4 To compound the problem, govern-\ncare costs (from 36.5 percent in 1990 to 38.2\nment subsidies now cover a smaller share of\npercent in 1994). Public hospitals also in-\nhospital uncompensated care costs, down\ncurred a larger burden of uncompensated care\nfrom 27.8 percent in 1980 to 19.3 percent in\nin 1994 than in 1990, as indicated by increases\n1994 (Exhibit 1).\nin both the percentage of total costs devoted\nIf private hospitals become increasingly\nto uncompensated care (from 11.8 percent in\nlimited in their ability and willingness to pro-\n1990 to 12.8 percent in 1994) and the propor-\nvide uncompensated care, it will be difficult\ntion of all hospital uncompensated care costs\nfor public hospitals to make up the difference.\nincurred by public hospitals (from 33.4 per-\nMedicaid expansions and sizable increases in\ncent in 1990 to 36.8 percent in 1994).\nMedicaid disproportionate-share funding\nWhile these changes are fairly modest in\nmay have helped to fuel the small increase in\nmagnitude, they represent a reversal of the\nuncompensated care provided by public hos-\ntrends of the previous decade. Between 1980\npitals during the 1990s, but continued reli-\nand 1990, the relative costs of uncompensated\nance on this source of revenue is not likely\ncare for all community hospitals increased by\nbecause of smaller increases in the number of\nabout 20 percent, with much of the increase\nMedicaid beneficiaries and new limitations\noccurring in the first half of the 1980s. More-\non disproportionate-share payments.\nover, uncompensated care costs became less\nIn addition, if private hospitals simultane-\nconcentrated during the 1980s. Substantial in-\nously reduce their uncompensated care bur-\ncreases in the number and percentage of hos-\ndens and attract more Medicaid managed care\npitals that provided a large amount of uncom-\npatients, then public hospitals will be com-\n168\npensated care were accompanied by equally\npelled to absorb even larger numbers of unin-\nsubstantial decreases in the number of hospi-\nsured patients while at the same time losing a\ntals that provided relatively little uncompen-\nvital source of revenue. This could seriously\nsated care. Also in contrast to the 1990s, pri-\nharm public hospitals' role as providers of last\nvate hospitals incurred the largest increase in\nresort. Not only do public hospitals provide\nuncompensated care costs during the 1980s,\nmore than twice as much uncompensated\nin terms of both the relative amount provided\ncare as private hospitals provide on average,\nand the proportion of total uncompensated\nbut research has shown that the total amount\ncare costs.\nof uncompensated care provided in a commu-\nWhy have the trends in the amount and\nnity tends to be greater in communities with a\ndistribution of uncompensated care that oc-\npublic hospital than in those without.⁵\ncurred in the first half of the 1980s started to\nPRIVATE PHYSICIANS. Although most\nreverse in the 1990s, despite the fact that the\nphysicians receive no explicit subsidy to pro-\n\"demand\" for uncompensated care has contin-\nvide services to the uninsured, data from the\nued to increase? Perhaps the most plausible\nAmerican Medical Association's (AMA's)\nexplanation is that market pressures faced by\nSocioeconomic Monitoring System (SMS)\nhospitals have constrained their ability and\nsurvey indicate that the amount of charity\nwillingness to provide uncompensated care in\nand uncompensated care provided by physi-\nrecent years. The ability to shift costs has\ncians is on the the rise.⁶ Nearly 68 percent of\nbeen more limited in the 1990s compared with\nphysicians in 1994 provided some charity care\nthe 1980s, partly because of increasing compe-\n(that is, care that was provided free of charge\ntition and involvement with managed care\nor at a reduced fee because of the financial\nplans, which many consider to be more vigor-\nneed of the patient) (Exhibit 2).⁷ This repre-\nous in controlling costs.³ More than two-\nsents an increase from 1990 (63.8 percent) and\nthirds of private hospitals now have at least\n1988 (62.0 percent). For physicians who pro-\none health maintenance organization (HMO)\nvided charity care, the average amount of time\nHEALTH AFFAIRS Volume 16, Number 4\nHEALTH TRACKING: TRENDS\nEXHIBIT 1\nTrends In Hospital Provision Of Uncompensated Care, 1980-1994\nYear\nTrends\n1980\n1985\n1990\n1994\nTotal costs of uncompensated care (billions of\ndollars)ᵃ\nAll hospitals\n$13.8\n$15.7\n$16.7\n$16.8\nPublic hospitals\n6.0\n5.5\n5.6\n6.2\nPrivate hospitals\n7.8\n10.2\n10.9\n10.6\nUncompensated care costs as percent of total\ncosts\nAll hospitals\n5.1%\n5.7%\n5.9%\n6.1%\nPublic hospitals\n11.2\n11.4\n11.8\n12.8\nPrivate hospitals\n3.6\n4.5\n4.7\n4.7\nPercent distribution of hospitals, by level of\nuncompensated care costs incurredᵇ\n0-2% of total costs\n31.1\n19.1\n20.6\n21.6\n3-5%\n48.8\n48.4\n45.3\n45.2\n6-9%\n14.8\n23.4\n24.4\n24.7\n10% or more\n5.3\n9.0\n9.6\n8.5\nPercent distribution of hospitals with \"high\"\nuncompensated care costs, by ownershipᶜ\nPublic\n68.4\n59.7\n54.9\n56.1\nPrivate\n31.6\n40.3\n45.1\n43.9\n169\nPercent distribution of total uncompensated care\ncosts, by level of uncompensated care costs\nincurredᵇ\n0-2% of total costs\n9.3\n5.4\n5.0\n5.4\n3-5%\n34.1\n35.4\n31.9\n31.3\n6-9%\n19.3\n22.9\n26.5\n25.1\n10% or more\n37.2\n36.4\n36.5\n38.2\nPercent of total uncompensated care costs\nincurred by public hospitals\n43.1\n35.1\n33.4\n36.8\nPercent of uncompensated care costs covered\nby government subsidies\nAll hospitals\n27.8\n19.8\n21.0\n19.3\nSOURCE: American Hospital Association annual surveys.\nNOTES: Estimates are for all U.S. registered community hospitals.\na Results for 1980-1990 were adjusted for inflation to reflect 1994 dollars, based on the Consumer Price Index for hospital\nservices.\nb Uncompensated care costs were measured as a percentage of total costs for each hospital.\nC A hospital is classified as incurring high uncompensated costs if these costs equal or exceed 10 percent of its total costs.\nthey spent providing that care also increased\nincreases in the number of uninsured may\nbetween 1990 and 1994. Total uncompensated\nhave contributed to the greater amount of\ncare costs (including bad debt) in 1994 were\ncharity care provided, increases in bad-debt\nestimated at $21 billion, an increase of about\ncosts may have resulted from greater numbers\n65 percent from 1990 after inflation.\nof underinsured patients.⁸ More vigorous ef-\nThe increase in physician uncompensated\nforts to control uncompensated care costs on\ncare was a combination of increases in both\nthe part of hospitals as well as the general shift\ncharity care provision and bad debt. While\nof health care resources from hospital-based\nHEALTH AFFAIRS July/August 1997\nHEALTH TRACKING: TRENDS\nEXHIBIT 2\nUncompensated Care Provided By Physicians, 1988-1994\nYear\n1988\n1990\n1994\nPercent of physicians providing charity care\n62.0%\n63.8%\n67.7%\nTime spent providing free and reduced-fee careᵃ\nHours per week that physicians provide charity\ncare\n6.6\n6.5\n7.2\nCharity care hours as percent of total hours\nworked\n11.0%\n10.6%\n12.4%\nUncompensated care costs (billions of dollars)b\n_C\n$12.76\n$21.14\nSOURCE: D.W. Emmons, \"Uncompensated Physician Care,\" in Socioeconomic Characteristics of Medical Practice, 1995\n(Chicago: American Medical Association, Center for Health Policy Research, 1995).\na Based on physicians who provided charity care.\nb Results for 1990 were adjusted by the authors for inflation to reflect 1994 dollars based on the Consumer Price Index for\nphysician services.\nC Not available.\ninpatient care to office-based outpatient care\nnue expanded during the 1990s, in large part\nmay also be fueling the increased demand for\nbecause of a substantial increase in revenue\nphysician uncompensated care.\nfrom Medicaid. Medicaid accounted for one-\nNevertheless, the substantial increase in\nthird of all community health center revenue\nphysician uncompensated care is somewhat\nin 1995, up from 20.3 percent in 1990. The\n170\nunexpected, since physicians are subject to\ninflux of Medicaid revenue is the result of\nthe same financial pressures that are making\nboth expanded eligibility and legislation that\nit difficult for hospitals to provide charity\nmandated cost reimbursement for Medicaid\ncare. It is possible that the potential effects of\npatients at federally qualified health centers.\nsome of these pressures either have been ex-\nExpansion of community health center ca-\naggerated or have yet to have a meaningful\npacity has permitted an increase in the\nimpact on physicians. For example, the pro-\nnumber of uninsured patients treated at com-\nportion of physician practice revenue derived\nmunity health centers-from about 2.2 mil-\nfrom managed care contracts increased only\nlion in 1990 to almost three million in 1995.\nfrom 28 percent in 1990 to 34 percent in 1994,\nThe estimated cost of providing services to\nwhich suggests that the level of managed care\nthe uninsured was more than $1 billion in\ninvolvement among physicians was not yet\n1995, a 55 percent increase from 1990 even\nhigh enough during that time to have a signifi-\nafter taking inflation into account. 10 However,\ncant impact on charity care provision.⁹\nit is important to note that even though unin-\nCOMMUNITY HEALTH CENTERS.\nsured patients comprise 40 to 45 percent of\nCommunity health centers receive grants\ncommunity health centers' patients and total\nfrom the federal government and other\ncosts, community health centers actually\nsources to provide comprehensive primary\nserve only a small percentage of the total unin-\nhealth care services to persons living in medi-\nsured population (6.5 percent). The aggregate\ncally underserved areas. According to data\ncosts for community health centers of provid-\nfrom the Bureau of Primary Health Care, the\ning services to the uninsured are much smaller\nnumber of community health center sites, pa-\nthan the aggregate costs of services provided\ntients served at those sites, and primary care\nto the uninsured by physicians and hospitals.\nphysicians practicing at those sites increased\nThe relatively small number of uninsured per-\nsubstantially during the 1990s (Exhibit 3).\nsons served by community health centers is\nCommunity health center capacity and reve-\nprobably attributable to the fact that such\nHEALTH AFFAIRS Volume 16, Number 4\nHEALTH TRACKING : TRENDS\ncenters are not located in all U.S. communi-\nrates could result in greater competition from\nties. In addition, many uninsured persons,\nprivate health plans and other providers for\nmost of whom have incomes above the pov-\nthis vital source of revenue. In contrast to\nerty line, do not live in the medically under-\npublic hospitals, however, community health\nserved areas targeted by community health\ncenters may be in a better position to survive\ncenters.\nby entering into various types of arrange-\nAlthough increased Medicaid revenue has\nments with health plans and providers, which\nprobably benefited community health cen-\nare attracted to community health centers be-\nters' Medicaid and uninsured patients, in-\ncause of their primary care emphasis and ex-\ncreasing reliance on Medicaid as a major\npertise and large Medicaid patient base.\"\nsource of revenue carries some risks, espe-\nDISTRIBUTION OF UNCOMPENSATED\ncially given the shift to managed care. Low\nCARE COSTS. The trends in Exhibits 1, 2, and\ncapitation rates for Medicaid managed care\n3 imply that there has been a shift of uncom-\npatients may result in greater financial strains\npensated care resources from hospital to non-\nfor community health centers, whereas high\nhospital sources, although direct comparisons\nEXHIBIT 3\nTrends In Community Health Centers, 1990-1995\nYear\nTrends\n1990\n1993\n1995\nNumber of community health center sites\n1,396\n1,574\n1,647\nNumber of patients served (thousands)\n5,887\n6,866\n7,787\n171\nPrimary care physicians\nTotal number of primary care physicians\nserving at community health centers\n2,450\n3,000\n3,450\nAverage number of primary care physicians\nper 10,000 medical care users\n4.80\n5.03\n5.14\nCommunity health center revenues\nTotal revenues (millions of dollars)\n$1,631\n$2,158\n$2,396\nPercent of revenue by source\nFederal grants\n40.3%\n31.7%\n28.4%\nState, local, and other sources\n15.9\n16.7\n16.7\nMedicaid\n20.3\n31.9\n33.4\nMedicare\n6.0\n5.2\n6,5\nOther third party\n8.8\n7.7\n8.3\nPatient fees\n8.8\n6.7\n6.7\nUninsured patients at community health centers\nNumber of uninsured patients (thousands)\n2,240\n2,770\n2,985\nPercent of all community health center patients\n38.0%\n39.0%\n39.0%\nPercent of uninsured persons who use community\nhealth centers\n5.4\n5.6\n6.5\nCost of care for uninsured persons\nFor all services (millions of dollars)ᵃ\n$680\n$943\n$1,054\nPercent of total costs\n41.8%\n44,5%\n44.0%\nSOURCE: U.S. Department of Health and Human Services, Public Health Service, Bureau of Primary Health Care, unpublished\ndata.\nNOTE: Includes community and migrant health centers covered under Sections 329 and 330 of the Public Health Service Act.\na Estimated as the sum of costs of medical services to uninsured persons, plus the costs of enabling services, minus patient\nfees. Results for 1990 and 1993 were adjusted for inflation to reflect 1995 dollars based on the Consumer Price Index for\nphysician services.\nHEALTH AFFAIRS July/August 1997\nHEALTH TRACKING : TRENDS\nin the relative magnitude of hospital, physi-\ncare measures do not reflect the relatively high\ncian, and community health center uncom-\npercentage of uninsured persons who report\npensated care costs are problematic because\nunmet health needs, nor do they indicate ap-\nof different data collection and computational\npropriateness of care setting and whether care\nmethods. Charity care and bad-debt expenses\nwas received in a timely and cost-effective\nin the American Hospital Association (AHA)\nmanner.¹³ Direct measures of access to care are\nannual survey are obtained directly from hos-\nrequired to draw more firm conclusions about\npital financial records.¹² Physician uncompen-\ntrends in access to care for the uninsured.\nsated care is calculated by the AMA by multi-\nWhile physician uncompensated care in-\nplying hours of charity care by the value of\ncreased substantially between 1990 and 1994,\nphysicians' time (that is, mean gross earnings)\nphysician use by the uninsured remained es-\nand adding bad-debt expenses. All three com-\nsentially unchanged during this period (about\nponents of the physician uncompensated care\nthree visits a year on average per person), and\nmeasure are based on respondents' estimates,\nthe gap in physician use between uninsured\nwhich are subject to recall error. Estimates of\nand privately insured persons widened\nthe amount of care that community health\nslightly (Exhibit 4). Furthermore, the per-\ncenters provide to uninsured patients in-\ncentage of uninsured persons who lacked a\nvolved multiplying the number of uninsured\nusual source of care, a traditional measure of\npatients by the average cost per patient plus\naccess that facilitates entry into the system,\nthe costs of enabling services.\nincreased from 27.7 percent in 1977 to 35.9 per-\nThese methodological differences make it\ncent in 1993 (with 7 percent of the increase\ndifficult to conclude that physicians actually\ntaking place between 1977 and 1987), which\nprovided more uncompensated care than hos-\nindicates that nonemergency access to pri-\n172\npitals provided in 1994 ($21 billion for physi-\nmary care among uninsured persons is dete-\ncians versus $16.8 billion for hospitals). How-\nriorating. The percentage of privately insured\never, community health centers clearly have a\npersons without a usual source of care de-\nmuch smaller share of the uncompensated\ncreased between 1987 and 1993, thereby\ncare market than either physicians or hospi-\ngreatly increasing the disparity between un-\ntals have. Nevertheless, because these meth-\ninsured and privately insured persons, which\nods are at least consistent over time, it is plau-\nmay reflect increases in the percentage of pri-\nsible that the almost zero growth in hospital\nvately insured persons who were in HMOs\nuncompensated care costs during the first\nand other managed care plans that require en-\npart of the 1990s and the large increases in\nrollees to have a primary care provider.¹²\nphysician uncompensated care costs and\nInferences about access based on hospital\ncommunity health center costs strongly sug-\nuncompensated care measures are even more\ngest that there has been a shift in uncompen-\nproblematic. First, hospital utilization for all\nsated care costs from hospitals to nonhospital\ngroups has declined substantially over the\nsources.\npast fifteen years as a result of cost control\nIMPLICATIONS FOR ACCESS. While\nmeasures by payers as well as innovations in\nthe amount of uncompensated care provided\ncare delivery that allow more procedures to be\nhas important implications for access to care\ndone in ambulatory settings. Thus, changes in\nfor the uninsured, it is a relatively poor meas-\nthe consumption of hospital services, as indi-\nure with which to assess changes in access.\ncated by hospital uncompensated care, rates\nFirst, it is unclear whether trends in uncom-\nof discharges, or lengths-of-stay, tell us very\npensated care reflect changes in demand (for\nlittle about changes in access to hospital care.\nexample, the number of uninsured persons)\nSecond, hospitalizations are increasingly\nor changes in supply (for example, as a result\nviewed as reflective of poor outcomes of care\nof competition and providers' financial cir-\nfor certain conditions that are considered\ncumstances). In addition, uncompensated\ntreatable in ambulatory care settings. Pre-\nHEALTH AFFAIRS - Volume 16, Number 4\nHEALTH TRACKING: TRENDS\nEXHIBIT 4\nEstimates Of Access To Care And Service Use By The Uninsured And Privately Insured,\nBy Measure And Year\nEstimate for\nEstimate for privately\nMeasure and year\nuninsured persons\nInsured persons\nAverage number of annual physician visits\nper personᵃ\n1986\n3.1\n4.2\n1990\n3.2\n4.4\n1994\n3.0\n4.6\nPercent of persons without a usual source\nof careᵇ\n1977\n27.7%\n15.3%\n1987\n34.6\n18.5\n1993\n35.9\n11.3\nHospital discharges per 1,000 personsᵃ\n1986\n63.3\n75.5\n1990\n58.3\n64.1\n1994\n47.7\n56.6\nPercent of hospital discharges for ACS\nconditionsᶜ\n1985\n6.8%\n5.3%\n1990\n10.1\n6.6\n173\n1994\n12.7\n7.6\nSOURCES: National Health Interview Survey; National Medical Expenditure Survey; and National Hospital Discharge Survey.\nNOTES: All estimates are for persons under age sixty-five. All estimates are age-adjusted. using as the base the age distribution\nof the privately insured population in the most recent year (1993 for the usual source of care measure; 1994 for all other\nmeasures). ACS is ambulatory care-sensitive.\na Based on the National Health Interview Survey.\nb Based on the National Medical Expenditure Survey (1977, 1987) and the National Health Interview Survey (1993).\nC Based on the National Hospital Discharge Survey. ACS conditions were defined similarly to those used in J.S. Weissman, C.\nGatsonis, and A.M. Epstein, \"Rates of Avoidable Hospitalization by Insurance Status in Massachusetts and Maryland,\" Journal of\nthe American Medical Association (4 November 1992): 2388-2394. Conditions include ruptured appendix, asthma, cellulitis,\ncongestive heart failure, diabetes, gangrene, immunizable conditions, malignant hypertension, pneumonia, pyelonephritis, and\nchronic obstructive pulmonary disease. Uninsured is defined as discharges that were classified as self-pay or no charge.\nvious research has found that rates of hospi-\nproviding timely treatment in the least costly\ntalizations for ambulatory care-sensitive\nsettings, it will be difficult to determine\n(ACS) conditions are higher among the unin-\nwhether increases in hospital uncompensated\nsured than the privately insured, which re-\ncare costs reflect increased access to hospital\nsearchers interpret as indicating less access to\nservices or decreased access to office-based\nprimary care for uninsured persons.¹⁵ Further-\nprimary care for uninsured persons.\nmore, this disparity has grown worse over the\npast ten years. The percentage of hospital dis-\nCONCLUSION\ncharges for ACS conditions among uninsured\nPolicymakers are concerned about potential\npersons almost doubled between 1985 and\nreductions in the provision of uncompensated\n1994, and discharges for these conditions\ncare because these reductions could result in a\nwere 1.67 times higher for uninsured persons\nserious deterioriation in access to care for the\nthan for privately insured persons in 1994,\nuninsured. While much of this concern is fo-\ncompared with 1.29 times higher in 1985 (Ex-\ncused on the traditional safety-net provid-\nhibit 4).¹⁶ With the increasing emphasis on\ners-public hospitals and community health\nHEALTH AFFAIRS July/August 1997\nHEALTH TRACKING: TRENDS\ncenters-there is perhaps an even greater\nthese more complex phenomena.\nthreat from the pressures on mainstream\nproviders-private hospitals and physi-\nThe Center for Studying Health System Change is sup-\ncians-to limit their uncompensated care\nported in full by The Robert Wood Johnson Founda-\nservices. While mainstream providers cur-\ntion. The authors gratefully acknowledge the support\nrently provide most of the uncompensated\nand assistance of the individuals and organizations\ncare, they receive few government subsidies\nthat provided data for this study: Joe Martin and\nto finance this care, and, unlike public hospi-\nLloyd Wackerling, American Hospital Association;\ntals and community health centers, many do\nBonnie Lefkowitz Bureau of Primary Health Care;\nnot have an explicit mission to serve the unin-\nJoel Cohen, Agency for Health Care Policy and Re-\nsured. Given the substantial challenges that\nsearch; Ann Hardy, Peggy Barker, and Susan Jack, Na-\ntraditional safety-net providers face just to\ntional Center for Health Statistics (NCHS) (National\nsurvive in the near future, they will have diffi-\nHealth Interview Survey); and Robert Pokras, Maria\nculty absorbing an even greater burden of un-\nOwings, and Elaine Wood, NCHS (National Hospital\ncompensated care.\nDischarge Survey). The authors also acknowledge\nData required to understand these changes\nLinda Kohn of the Center for Studying Health System\nand their effect on access to care for the unin-\nChange (CSHSC) for valuable contributions to the\nsured are limited. Previous research has fo-\nconceptualization of this study; Mary Harrington of\ncused on hospital uncompensated care, in\nMathematica Policy Research, Inc., Jack Ashby of the\npart because these data are more extensive\nProspective Payment Assessment Commission, and\nthan are data on uncompensated care by other\nDavid Emmons of the American Medical Association\nproviders. However, this focus ignores the siz-\nfor helpful comments on some specific findings; Paul\nable amount of charity care provided by physi-\nGinsburg, Joy Grossman, Mark Legnini, and Jim\n174\ncians and other nonhospital providers. In addi-\nReschovsky from CSHSC and Jane Stein from The\ntion, changing delivery patterns are probably\nStein Group for their review of earlier drafts and\nresulting in a shift of uncompensated care re-\nmany helpful comments; Jeremy Pickreign from\nsources from inpatient to outpatient services.\nCSHSC and Ase Sewall and Beny Wu from Social and\nThe shift from high-cost hospital-based serv-\nScientific Systems for programming support; and Ruth\nices to lower-cost outpatient facilities in-\nGriffin from CSHSC for assistance with the prepara-\ncreases the importance of examining changes\ntion of tables.\nin the distribution of uncompensated care\nNOTES\ncosts across different providers. This will be\ndifficult, if not impossible, for most communi-\n1. Local public health departments and clinics that\ndo not receive federal grant funds (but are consid-\nties, however, because of the lack of data on\nered federally qualified health centers for Medic-\nuncompensated care from private physicians\naid reimbursement) also provide a disproportion-\nand other providers and problems of compa-\nately large amount of indigent care. However,\nrability of data sources when they exist.\nthey are excluded from this analysis primarily be-\ncause of lack of national data. Although conclu-\nFinally, more direct measurement of access\nsions about trends in the capacity of the safety net\nthrough population surveys and other meth-\nmust consider these omissions, it is likely that\nods is necessary to fully understand the im-\nthey treat only a small portion of the care pro-\npact that system changes are having on care\nvided to uninsured patients. Federally assisted\nfor the uninsured. Access is not just a ques-\nhealth centers account for more than 80 percent\nof all patients of federally qualified health centers\ntion of adequate resources but also includes\n(estimated by comparing data on all federally\nthe process of getting care (for example, the\nqualified health centers, compiled by the National\nproximity of providers, the convenience of of-\nAssociation of Community Health Centers, with\nfice hours, and getting the necessary care and\ndata on federally assisted community health cen-\ntreatment once inside the system) and the ap-\nters from the Bureau of Primary Health Care).\nAlso, approximately 70 percent of local health de-\npropriate level and type of care. Measures of\npartments do not provide primary care services.\nuncompensated care tell us very little about\nNational Association of County and City Health\nHEALTH AFFAIRS Volume 16, Number 4\nHEALTH TRACKING: TRENDS\nOfficials and Centers for Disease Control and\nputed by AHA researchers.\nPrevention, 1992-1993 National Profile of Local Health\n13. M.L. Berk, C.L. Schur, and J.C. Cantor, \"Ability\nDepartments (Washington: NACCHO, 1995).\nto Obtain Health Care: Recent Estimates from\n2. Hospital uncompensated care costs are defined\nthe Robert Wood Johnson Foundation National\nas the sum of charity care and bad debt. Uncom-\nAccess to Care Survey,\" Health Affairs (Fall 1995):\npensated care net of government subsidies is a\n139-146; and K. Donelan et al., \"Whatever Hap-\nmore appropriate measure when assessing the\npened to the Health Insurance Crisis in the\nfinancial burden of uncompensated care on hos-\nUnited States?\" Journal of the American Medical Asso-\npitals.\nciation 276, no. 16 (1996): 1346-1350.\n3. J. Mann et al., \"Uncompensated Care: Hospitals'\n14. In this analysis, persons who identified a hospi-\nResponses to Fiscal Pressures,\" Health Affairs\ntal emergency room as their usual source of care\n(Spring 1995): 263-270.\nwere classified as not having a usual source of\n4. Prospective Payment Assessment Commission,\ncare. The differences in usual source of care men-\nMedicare and the American Health Care System: Report\ntioned in the text were statistically significant at\nto the Congress (Washington: ProPAC, 1996), 34.\nthe .05 level. Some caution should be used with\n5. K. Thorpe and C. Brecher, \"Improved Access to\nthese findings, since the estimates for different\nCare for the Uninsured Poor in Large Cities: Do\nyears are based on different surveys. The Na-\nPublic Hospitals Make a Difference?\" Journal of\ntional Health Interview Survey (NHIS) was used\nHealth Politics, Policy and Law (Summer 1987):\nfor the 1993 estimates, while the National Medi-\n313-324.\ncal Expenditure Survey (NMES) was used for\n6. It is important to note that the method for deter-\n1977 and 1987. Usual source of care was not avail-\nmining charity and uncompensated care costs\nable in earlier versions of the NHIS, and the most\nfor physicians is considerably different than it is\nrecent NMES is not yet available. Although there\nfor hospitals. The former relies largely on physi-\nare some differences, the surveys use similar word-\ncians' recall of the number of hours they spent in\ning for the usual-source-of-care question, mode of\nthe previous week providing care for free or at a\ninterview (in-person), use of proxy responses,\nreduced charge and is subject to reporting error.\nlength of interview (and placement of the question\nThe extent of this error is unknown. By contrast,\nin a special supplement toward the end of the in-\n175\nhospital uncompensated care costs are derived\nterview), and design of population weights. There\ndirectly from hospital financial records.\nwere some differences in the way that insurance\n7. The estimates of physician charity and uncom-\nstatus was ascertained, although attempts were\npensated care in Exhibit 2 were computed by\nmade to minimize these differences.\nresearchers at the American Medical Associa-\n15. J.S. Weissman, C. Gatsonis, and A.M. Epstein,\ntion. See D.W. Emmons, \"Uncompensated Physi-\n\"Rates of Avoidable Hospitalization by Insur-\ncian Care,\" in Socioeconomic Characteristics of Medi-\nance Status in Massachusetts and Maryland,\"\ncal Practice, 1995 (Chicago: AMA, Center for\nJournal of the American Medical Association (4 No-\nHealth Policy Research, 1995).\nvember 1992): 2388-2394.\n8. P.F. Short and J.S. Banthin, \"New Estimates of\n16. The differences mentioned in the text are statis-\nthe Underinsured Younger than 65 Years,\" Jour-\ntically significant at the .05 level. It is important\nnal of the American Medical Association 274, no. 16\nto note that the percentage of hospital dis-\n(1995): 1302-1306.\ncharges for ACS conditions increased for all per-\n9. D.W. Emmons and C.J. Simon, \"Managed Care:\nsons during this period, regardless of insurance\nParticipation, Revenues, and Risk,\" in Socioeco-\nstatus. Thus, caution should be used in inter-\nnomic Characteristics of Medical Practice, 1995.\npreting these findings, since it is likely that fac-\n10. Since community health centers receive federal\ntors unrelated to access are affecting changes in\ngrants, few services are truly uncompensated.\nhospital use patterns. Nevertheless, the fact that\nThe estimates cited serve as a proxy for uncom-\nthe rate of increase for uninsured persons was\npensated care costs and refer to services that are\nmore than double that of privately insured per-\nnot generally reimbursable through third parties\nsons suggests that the use of inpatient services\nbut are subsidized through federal grants and\nfor ACS conditions relative to more general\nother funding sources.\nchanges in hospital utilization patterns is in-\n11. D.J. Lipson and N. Naierman, \"Effects of Health\ncreasing.\nSystem Change on Safety-Net Providers,\" Health\nAffairs (Summer 1996): 33-48.\n12. Financial data are missing for some hospitals in\nthe AHA survey. For example, uncompensated\ncare data were missing for about 18 percent of\nhospitals in the 1994 AHA survey and were im-\nHEALTH AFFAIRS July/August 1997\nHEALTH TRACKING : FROM THE FIELD\nLessons From Arizona's\nMedicaid Managed Care\nProgram\nWith thirteen years of experience, Arizona's Medicaid managed care\nprogram offers valuable insight into the potential and pitfalls of this\nform of safety-net system.\nBY NELDA MCCALL\nTH INCREASED PRESSURE on\nices and nursing home services to Medicaid\nW\npublic programs to decrease costs\neligibles who are at risk of institutionalization.\nwhile maintaining a safety net of\nThe AHCCCS program provides managed\nservice use for beneficiaries, many state\ncare services through acute care plans and\nMedicaid programs are looking to managed\nlong-term care contractors capitated by the\ncare as an important component of an im-\nstate. Covered services for Medicare benefici-\nproved and more cost-effective delivery sys-\naries are paid for on a fee-for-service basis by\n194\ntem. This paper examines the lessons learned\nMedicare. Providers are required to bill Medi-\nfrom evaluations of the first statewide man-\ncare first for such services.² The state provides\naged Medicaid program, the Arizona Health\nthe overall direction for the program with re-\nCare Cost Containment System (AHCCCS).\nsponsibilities for eligibility and enrollment;\nThe paper focuses on two questions: (1) Does\nselecting, paying, and regulating capitated\nthe experience in Arizona justify expansion of\nproviders; monitoring the quality and appro-\nMedicaid managed care to other states? (2)\npriateness of care; and maintaining an infor-\nWhat lessons learned in Arizona should other\nmation system to support program opera-\nstates consider as they move into Medicaid\ntions. In addition, Arizona provides\nmanaged care?\nreinsurance for inpatient services and cover-\nTHE AHCCCS PROGRAM\nage for catastrophic services. In the past,\nAHCCCS has also acted as the plan or con-\nThe AHCCCS program began in October 1982\ntractor in counties for which it was unable to\nas an alternative to traditional Medicaid. Be-\nfind a qualified organization at an acceptable\nfore AHCCCS, Arizona was the only state not\ncapitation rate.\nparticipating in Medicaid. The original\nPlans and contractors engage in a broad\nAHCCCS program did not cover nursing home\nrange of service delivery, internal monitoring,\ncare. In 1989 the Arizona Long-Term Care Sys-\nand data-sharing activities. Besides providing\ntem (ALTCS) was incorporated into\ncase-managed covered services, they must\nAHCCCS. ALTCS provides a full range of\nmanage a provider network, distribute a\nacute home-based and community-based serv-\nmember handbook, and collect third-party\nNelda McCall is president of Laguna Research Associates, in San Francisco. She has directed numerous\nmajor government and private foundation evaluations, including two six-year Health Care Financing Ad-\nministration evaluations of the Arizona Health Care Cost Containment System. She is also the project\ndirector of The Robert Wood Johnson Foundation's evaluation of its Partnership for Long-Term Care.\nHEALTH AFFAIRS Volume 16, Number 4\n© 1997 The People-to-People Health Foundation, Inc.\nHEALTH TRACKING: FROM THE FIELD\nand patient liabilities. They must also main-\nArizona program: cost of the program, use of\ntain systems for quality management, finan-\nhealth care services by program beneficiaries,\ncial management, grievance and appeals, and\nand access to and quality of services received\ndata management. The data management sys-\nunder AHCCCS.⁵\ntem must support timely submission of data\nCOST. The cost analyses for the acute\nto AHCCCS.\ncare AHCCCS program compared the actual\nAHCCCS covered about 463,000 benefici-\nper capita costs by eligibility group with an\naries in the acute part of the program and\nestimate of the per capita costs of a traditional\n23,000 beneficiaries in ALTCS in March 1997.\nMedicaid program in Arizona.⁶ Comparison\nMost acute care beneficiaries were served\nstates were those with reliable and complete\nthrough one of the fourteen health plans se-\ndata that were similar to Arizona in their\nlected through a competitive bidding process.\nMedicaid requirements. For fiscal year (FY)\nThe state releases a request for proposals that\n1983 through FY 1993, savings of $197 million\nlists participation requirements and evalu-\nwere estimated, an average savings per year of\nation criteria. Acute care beneficiaries choose\napproximately II percent of medical service\nfrom among the health plans available in their\ncosts and 7 percent of total costs (medical\ncounty. All of the counties have more than one\nservice costs plus administrative costs). The\nhealth plan.\naverage annual increase in cost for the\nThe long-term care program serves those\nAHCCCS acute care program (9.1 percent)\nwho are financially eligible (up to 300 percent\nwas also smaller than the annual increase for\nof Supplemental Security Income eligibility)\nthe traditional program (10.3 percent).\nand determined by state assessors, using a\nThe ALTCS cost analyses compared the\npreadmission screening instrument, to be at\nnumber of users, cost per month, and total\nrisk of institutionalization. Beneficiaries in-\ncost for ALTCS with estimates of what a tra-\n195\nclude the elderly and physically disabled and\nditional Medicaid program in Arizona would\nthe mentally retarded/developmentally dis-\nhave cost for the first five years of the pro-\nabled. Five counties and two private entities\ngram, 1989 through 1993. Estimates for the\nserve as contractors for the elderly and physi-\ntraditional program were made for compari-\ncally disabled. There is only one contractor\nson states using the Medicaid Statistical In-\nper county. The Arizona Department of Eco-\nformation System. ALTCS costs, including\nnomic Security is the contractor for the men-\nmedical and administrative costs, were on av-\ntally retarded/developmentally disabled.³\nerage 16 percent per year lower than the costs\nContractors provide case management and\nof a traditional Medicaid program in Arizona.\nmake placement decisions.⁴ They are capi-\nWhen considering medical services alone, the\ntated at a negotiated rate using a methodology\nALTCS savings were 18 percent per year. To-\nthat provides economic incentives to care for\ntal cost savings were almost $290 million, and\nbeneficiaries at home rather than in nursing\nthese savings increased substantially over\nhomes.\ntime (0.2 percent in 1989, 8 percent in 1990, 15\npercent in 1991, and 21 percent in 1992 and\nDOES ARIZONA'S EXPERIENCE\n1993). The average annual cost increases for\nJUSTIFY EXPANSION OF MEDICAID\nALTCS were lower than the increases for a\nMANAGED CARE?\ntraditional program, 4.0 percent per year ver-\nsus an estimated 9.6 percent per year.\nThe successes and failures of the longest-\nUSE OF SERVICES. Examination of the\nrunning statewide Medicaid managed care\nlong-term care and acute care programs found\nprogram serving all eligibility groups should\nthat beneficiaries in these programs used less\nbe considered by other states in their deci-\ninstitutional care than Medicaid beneficiaries\nsions about whether and how to implement\nin New Mexico used and more or about the\nMedicaid managed care. This section summa-\nsame amount of ambulatory care.\nrizes the evaluation of several aspects of the\nHEALTH AFFAIRS July/Augusi 1997\nHEALTH TRACKING: FROM THE FIELD\nThe analysis for the acute care program\nrevealed that Arizona's elderly and disabled\nwas based on Medicaid claims and encounters\nbeneficiaries were more likely than New\nfor a 5 percent randomly selected sample of\nMexico's elderly and disabled to have pres-\nAHCCCS and New Mexico Medicaid benefi-\nsure sores, fever, or a catheter inserted and\nciaries for FYs 1991 and 1992. For the long-\nless likely to be offered an influenza vaccine.\nterm care program, ALTCS claims data and\nThe incidence of falls and fractures and the\nencounters data for all eligible beneficiaries\nuse of psychotropic drugs were similar in the\nreceiving chronic long-term care services in\ntwo states. In response to these findings,\nNew Mexico were linked with Medicare\nALTCS initiated steps to include monitoring\nclaims to give a full picture of all medical care\nof the problem areas in their ongoing quality\nservice use for January 1991 through Septem-\nassurance activities.\nber 1992. Services in both states were catego-\nAnalyses of new admissions of elderly and\nrized using the same methods.⁷\nphysically disabled beneficiaries into the\nBeneficiaries of Aid to Families with De-\nlong-term care program found a much more\npendent Children (AFDC) in the acute care\ncoordinated system of care in Arizona, which\nprograms in Arizona and New Mexico had\npermitted transitions from nursing home to\nabout five evaluation and management visits\nhome care and vice versa. Evaluation of the\n(that is, office, home, nursing home, specialty,\ncost-effectiveness of home care provided un-\nconsultations, and therapies) per person-year,\nder ALTCS indicated that attempts to limit\nbut the number of hospital days per thousand\nspending on long-term care by diverting cli-\nperson-years was 40 percent lower in Arizona\nents to home care settings have been success-\n(590 days versus 976 days). Evaluation and\nful. This is in stark contrast to previous evalu-\nmanagement visits for ALTCS elderly and\nations of home and community-based care\n196\nphysically disabled eligibles resulted in more\nprograms that have found home care to be a\nvisits (thirteen versus nine), but the number\ncomplement to institutional care, not a sub-\nof hospital days per thousand person-years\nstitute.⁸\nwas 22 percent lower than in the traditional\nThe analyses of the two programs' out-\nprogram in New Mexico (3,692 versus 4,731).\ncomes indicated that the program was provid-\nQUALITY AND ACCESS. Several quality\ning better access to good-quality care at a\nand access studies of the AHCCCS and\nlower cost. However, some of the analyses\nALTCS programs, which used New Mexico\nshowed the program to be doing worse than a\nMedicaid beneficiaries as the control group,\ntraditional Medicaid program, suggesting the\nrevealed mixed results. An access and satis-\nneed for close monitoring of quality-of-care\nfaction survey conducted in 1985 indicated\nand access issues. Studies performed over the\nbetter access to routine and urgent care in\ncourse of the two evaluations highlighted ar-\nAHCCCS than in the New Mexico program\neas in which AHCCCS procedures and meth-\nand high satisfaction scores for services. Pri-\nods could be improved and demonstrated the\nmary prevention, preventive care use, and use\nimportance of having an ongoing analysis ca-\nof medical care for particular symptoms were\npacity within the administration of a man-\nsimilar in the two states. A review of Medic-\naged care program.\naid records of AFDC children and pregnant\nwomen in 1985-1987 indicated earlier, more\nLESSONS FROM ARIZONA?\nfrequent, and more complete health care for\nchildren in Arizona. Maternity care and preg-\nBesides looking at program outcomes, the\nnancy outcomes were similar in the two\nanalyses also examined the implementation\nstates, but Arizona had later initiation of\nand operation of the major features of the\npregnancy care and fewer prenatal visits.\nAHCCCS program. States implementing\nWith respect to the ALTCS program, a re-\nmanaged care may want to consider the fol-\nview of nursing home records in 1991 and 1992\nlowing five recommendations in designing\nand implementing their programs.\nHEALTH AFFAIRS\nVolume 16. Number 4\nHEALTH TRACKING: FROM THE FIELD\nSELECT LEADERS WHO HAVE THE\nprogram is the relationship between the state\nREQUIRED SKILLS. When designing and im-\nand the health plans and contractors. Respon-\nplementing new managed care programs,\nsibilities of the contracting entities in\nleadership is needed in three key positions\nAHCCCS include a specifically defined broad\nthat do not always exist in traditional Medic-\nrange of service delivery, internal monitoring,\naid programs: a chief information officer who\nthird-party liability identification and collec-\nknows how to collect and analyze data and\ntion, and data-sharing activities. The state\nuse it for policy development; a chief financial\nalso has defined responsibilities that are con-\nofficer who has experience in the financial\nsistent with its primary responsibility for\nmanagement of managed care organizations\noverall program direction. Clear definition of\nand who can lead the monitoring of plan op-\nauthorities and responsibilities is critical to\nerations; and a chief medical officer who has\nprogram success.\nexperience leading activities in assuring and\nRules for participation and selection. Participa-\nmonitoring the quality of care in the delivery\ntion and selection criteria also need to be\nof Medicaid services. AHCCCS did not have\nclearly defined. Participation in AHCCCS is\npersons with these kinds of skills in place at\nrestricted to organizations that meet specific\nthe beginning of the program, and many of its\nnetwork requirements; have internal controls\nimplementation problems can be traced to\nfor quality, financial, and grievance monitor-\nthis lack of leadership. In the three areas out-\ning; and have systems in place to produce the\nlined, the leaders must be in senior positions\ndata that are required under the program. Se-\nin the organization and have sufficient stature\nlection of winning bidders follows a defined\nto conduct discussions and negotiations with\nstructure that emphasizes the provider net-\nother state organizations and with health\nwork in selection but also considers the capi-\nplan chief executive officers. In addition,\ntation rate bid, the ability of the contractor to\n197\nthese leaders must have the vision to formu-\nmeet the program's requirements, and the\nlate a specific plan of operation and select and\nqualifications of the organization. Having\nmanage appropriate staff.\nbidders who do not win contracts is impor-\nCONSIDER DESIGN OPTIONS FOR THE\ntant to keeping the marketplace competitive.\nPROGRAM'S OPERATIONS. The design of\nMethod of setting capitation payment. The way\nthe following operational areas is particularly\ncapitation rates have been set has evolved\nimportant to the AHCCCS program and\nover time in the AHCCCS and ALTCS pro-\nshould be carefully considered in any new\ngrams. Specific marketplace considerations\nprogram.\nshould be taken into account in developing an\nEligibility determination and enrollment.\nappropriate strategy for each solicitation. In\nAHCCCS acute care implementation analyses\nthe beginning of a new program, the active\nsuggest that carefully thinking through the\ninvolvement of the state may be necessary to\nprocess of enrollment and how it interacts\nhelp safety-net providers and other potential\nwith the process of eligibility determination\nbidders get organized to participate. In addi-\ncan help to avoid unnecessary fee-for-service\ntion, the state may need to assume some of the\nliabilities. If beneficiaries select from multiple\nrisk of the cost of delivering services by retro-\nplans, the plans' marketing materials, media\nactively adjusting certain components of the\nadvertising, and solicitation methods should\ncapitation payment.\nbe reviewed and approved. In Arizona's\nMeeting the information needs of the program.\nALTCS program, the state performs the func-\nPerhaps most important in managed care is\ntional/medical assessment for eligibility using\nhaving a management information system\na preadmission screening instrument and\nthat emphasizes not only the operation of\ntherefore controls entry into the system.\nhardware and software but also the quality of\nResponsibilities of the capitated entities and the\nthe input data. Both the state and the partici-\nstate. Central to the workings of the AHCCCS\npating plans in AHCCCS have demonstrated\nHEALTH AFFAIRS July/August 1997\nHEALTH TRACKING: FROM THE FIELD\nthat credible data on eligibility, enrollment,\nUtilization and access monitoring. Analysis of\nnetworks, cost, use of services, and, for long-\nthe use of medical services is a critical compo-\nterm care beneficiaries, medical and func-\nnent in understanding how a medical care\ntional assessments can be captured and play\nprogram is performing. In a capitated pro-\nan important role in managing the program.\ngram, it is of special importance to ensure that\nOrganizations that cannot provide data on\nbeneficiaries are receiving appropriate treat-\nplan operations and on the use and cost of\nment by analyzing which services are pro-\nservices should be excluded from participa-\nvided and by monitoring the adequacy of ac-\ntion because without such data they cannot\ncess to ambulatory care.\nbe a cost-effective partner for the state.\nPlanning. Future planning requires systems\nADDRESS THE STATE'S MANAGE-\nthat can estimate costs of program modifica-\nMENT RESPONSIBILITIES.\ntions and future program\nMany of the early problems in\ncosts, identify areas that pro-\nimplementing the AHCCCS\n\"Integral to the\nmote long-run cost contain-\nprogram were related to not\nment, and research areas for\nhaving effective systems in\nsuccess of a state's\ngeneral study.\nplace to deal with the state's\nmanaged care\nCoordination with other states.\nmanagement responsibilities.\nStates should participate in\nNot having financial manage-\nprogram is the\ndeveloping structures and\nment structures in place re-\ninvolvement of the\nprocesses that support na-\nsulted in Arizona's not being\ntional standardization of data\nable to detect impending plan\ngovernor, key legis-\ncollection, sharing of technical\nbankruptcies until they were\nlators, and their\nknowledge, evaluating what\n198\nimminent. In the early pro-\nstaffs.\"\nworks and what does not, and\ngram years, quality assurance\nsharing ideas through collabo-\nactivities, utilization and ac-\nrative forums.\ncess monitoring, and program planning func-\nFOSTER WORKING RELATIONSHIPS\ntions had no leadership and few operating\nWITH THE GOVERNOR, KEY LEGISLA-\nmechanisms. Early attention to performance\nTORS, AND THE MEDIA. Integral to the suc-\nin these areas is critical to a well-functioning\ncess of a state's managed care program is the\nprogram, which can avert quality and access\ninvolvement of the governor, key legislators,\nproblems.\nand their staffs. In Arizona both the legisla-\nFinancial management. Contracting entities\nture and the governor were supportive of the\nmust be monitored for financial solvency. If\nprogram during its early years. This support\nplan failures occur, the stability of the pro-\nwas unwavering despite substantial problems\ngram is threatened. Monitoring the plans can\nin the program's administration during its\nhelp to identify problems before they become\nfirst eighteen months. Fostering strong rela-\ncritical or provide early warnings to help a\ntionships with the legislature and governor is\nstate make arrangements for a plan's orderly\neven more critical today. In an era of concern.\nphaseout. The state's ability to secure com-\nabout state spending, strong marketing will\npetitive rates also depends on the availability\nbe necessary to convince policymakers to allo-\nof accurate data on costs.\ncate sufficient resources for infrastructure de-\nQuality assurance. Quality assurance activities\nvelopment. Without this development, a state\nrequire early and concerted energy. Important\nruns a serious risk of problems with access,\nareas include activities to detect underuse of\nquality of care, and plan viability.\nservices, review of treatment patterns by diag-\nProgram staff should include a press\nnosis, monitoring of selected procedures, detec-\nspokesperson who briefs the media regularly\ntion of fraud and abuse, and profiling of plans\nabout what the state is trying to accomplish\nand physicians for quality and appropriateness.\nand who serves as a contact person for spe-\nHEALTH AFFAIRS Volume 16, Number 4\nHEALTH TRACKING: FROM THE FIELD\ncific problems or concerns. AHCCCS had\nAHCCCS program at the conclusion of the evalu-\ndaily negative press coverage in its first year,\nations; and present and former AHCCCS staff mem-\nwhich was exacerbated by the lack of staff to\nbers. The author also acknowledges Diane Rowland\nrepresent AHCCCS's positions to the press.\nand Alina Salganicoff from the Kaiser Family Foun-\nIn the early implementation stages of any new\ndation, Stephen Somers from the Center for Health\nprogram, there are likely to be problems.\nCare Strategies, Inc., and Jodi Korb of Laguna Re-\nThese problems have less potential to explode\nsearch Associates for their comments on an earlier\nif members of the media are familiar with the\nversion of this paper.\nprogram and its aims and have a specific per-\nNOTES\nson to contact when they have questions.\n1. This paper is based on results from two six-year\nSUPPORT THE DEVELOPMENT OF\nHealth Care Financing Administration (HCFA)\nNATIONAL REPORTING STANDARDS. Re-\nevaluations of AHCCCS. Detail on the method-\nporting formats should include detailed defi-\nologies, findings, and policy conclusions are\nnitions for each element to be included in en-\navailable in HCFA evaluation reports, discussion\ncounter data, financial reports, quality\npapers, articles submitted for publication, and\npublished articles. A complete list of these docu-\nassessment instruments, satisfaction surveys,\nments is available from the author on request, at\nand grievance reports. These standards for\nLaguna Research Associates, Suite 1190, 455\ndata collection will not only help the state to\nMarket Street, San Francisco, California 94105.\nattract qualified contracting entities and\n2. Arizona has attempted to get a waiver from\nthereby support market competitiveness, but\nHCFA to integrate Medicare and Medicaid fi-\nnancing but has been unsuccessful.\nthey also have wide societal uses. Data sets\n3. Two of the counties and the Arizona Depart-\nthat are integrated across several states can be\nment of Economic Security are mandated con-\nused to determine disease incidence, docu-\ntractors. Three counties became contractors\nment treatment trends, develop knowledge\nthrough legislatively mandated county right of\n199\non the success of treatments, and provide a\nfirst refusal in effect until July 1995. The two\nprivate entities were selected through a com-\ndatabase for general research inquiry. Such a\npetitive bidding process.\ndatabase will make it possible to improve\n4. The ALTCS program as a whole has a HCFA-\nhealth care for all Americans, not just those\nimposed cap on the percentage of elderly and\neligible for Medicaid.\nphysically disabled beneficiaries that can be\nserved in home care. Originally set at 10 percent\nThis paper was prepared in part with funds from The\nof beneficiaries, it was slowly raised as the pro-\ngram's use of home care was demonstrated to be\nHenry J. Kaiser Family Foundation. It is based on\ncost-effective. The rate was 40 percent in fiscal\nwork funded under Health Care Financing Admini-\nyear 1995.\nstration (HCFA) Contract nos. 500-83-0027 and\n5. The analyses described were limited by the avail-\n500-89-0067. The analysis and conclusions are solely\nability of data and especially by the lack of base-\nthose of the author and do not express any official\nline information on Arizona's experiences before\nthe program was implemented.\nopinion or endorsement by the Kaiser Family Founda-\n6. The per capita costs of the traditional Medicaid\ntion or HCFA. The author thanks the many persons\nprogram in Arizona were calculated from HCFA\nwho contributed to the two HCFA evaluations of the\nReports 2082s and 64s.\nArizona Health Care Cost Containment System\n7. The methods used were based on The Urban In-\nDemonstration: her coauthors of the AHCCCS\nstitute Type of Service Classification System,\nadapted by Laguna Research Associates for\nevaluation reports: Donald Balaban, Ellen Jones\nMedicaid beneficiaries.\nBauer, Michael Crane, Suzanne Pollack Driver, Susan\n8. The risk of institutionalization was developed\nHaber, Jodi Korb, Stanley Moore, Lynn Paringer,\nwith data from the Institutional Population\nGordon Trapnell, Pamela Turner, Alice Wade,\nComponent of the 1987 National Medical Expen-\nWilliam Weissert, John Wilkin, and C. William\nditure Survey.\nWrightson; Sidney Trieger, Paul Lichtenstein, William\nEngland, Ronald Lambert, James Hadley, Joan Peterson,\nand Henry Tyson of HCFA; Mabel Chen and Linda\nRedman, the director and deputy director of the\nHEALTH AFFAIRS July/August 1997\nPERSPECTIVE\nin 1996, to maintain this\na constant basis, 29,000\nreflecting the fact that\nand on Medicaid. Fewer\nBeyond The Safety Net In Dallas\npatients disenrolled from\na competing HMO.\nStructural solutions are needed for the structural problems of the safety net.\npartners. Although main-\nby Ron J. Anderson and Paul J. Boumbullan\npopulation through par-\nedicaid HMO was critical,\nP\nARKLAND MEMORIAL HOSPITAL is one\ncenters, five school-based clinics, and two\nserving other populations\nof this nation's largest, most active institu-\nmobile clinics. Parkland has learned over the\nRecently, another new\ntions making up the health care safety net. It\npast ten years that to truly improve health, it\nthat includes a number\nhas served the residents of Dallas County\nmust address the social and economic deter-\nor government hospi-\n(Texas) for more than a hundred years.\nminants of disease such as lifestyle, education,\nColorado. They share the de-\nJust ten years ago, Parkland was a large,\nand employment opportunities. This means\nin governance while wish-\ncentrally located hospital, which, while recog-\nworking with school districts, housing\nenefits of a provider network.\nnized for its volume of service, excellence in\nauthorities, police and fire departments,\njoined this entity. How-\npatient care, and high-quality postgraduate\nchurches and synagogues, and employers.\ncannot yet be assessed.\nmedical training programs, was not known for\nThese efforts have begun to bear fruit. In 1995\nexcellence and new markets.\nits amenities or for having a patient-centered\na study was completed that compared COPC\ncompetitive environment, it is\nenvironment. Over the course of the past dec-\npatients to non-COPC area residents. COPC\nDenver Health obtain new pa-\nade, Parkland has transformed itself. It is now\npatients were two times less likely to be ad-\nthat fit within its safety-net\nan integrated health care system. It has cre-\nmitted through the emergency department,\nalso provide additional reve-\nated centers of excellence that make available\nhad shorter hospital stays (3.4 days versus 5.3\n27\nthe provision of care for the\nto the entire community the latest knowl-\ndays, on average), and had 50 percent lower\nsuch efforts are the aggres-\nedge in such areas as arrhythmia manage-\ncharges.¹\nof centers of excellence in\nment, cerebral vascular disease, epilepsy, and\nin prisoner care. The former\ngastrointestinal treatment. It also has devel-\nSolving Structural Problems\nsafety-net mission of caring for\noped its own health maintenance organiza-\nTo pay for the decentralized system of com-\nof the entire population; the\ntion (HMO). In addition, it has gone through\nmunity care and its educational and research\nmission of caring for the needs\nextensive reengineering. These efforts have\nmissions, Parkland has had to diversify its\nulations.\nproduced a decline in the average length-of-\nrevenue streams and cross-subsidize. Fifteen\nefforts to provide a strong\nstay from seven days in 1986 to fewer than five\nyears ago, 66 percent of Parkland's funding\nsafety-net system appear to\ndays in 1995; a reduction in emergency depart-\ncame from local property taxes. Today only 33\nand are approaches that should\nment visits from more than 160,000 visits in\npercent of its funds come from that source.\nby other similar systems.\n1987 to 122,495 visits in 1995; and a reduction\nThis shift was accomplished primarily by\nin average \"dwell\" time in the outpatient clinic,\nmeans of an increased Medicaid entitlement\npaper was presented at The Robert\nincluding all diagnostic studies and pharmacy\nFoundation/Health Affairs/Alpha the Safety\nin Texas and through Medicare and commer-\nservices, from 7.5 hours to 2.5 hours.\ncial payments for services provided through\n\"What\nIs\nHappening\nto\nParkland decentralized by developing a na-\nParkland's centers of excellence. As Parkland\n1997, Washington, D.C.\ntionally recognized primary care delivery sys-\nhas worked toward providing relief to local\ntem based on a community-oriented primary\ncommunication with Larry Wall, direct\ncare (COPC) model, which aims to improve\ntaxpayers while expanding community serv-\ncare for both individuals and communities.\nices and maintaining its missions, it has be-\nHospitals Reference Guide to Financial Hos\nHospital Association.\nThe COPC network now includes eight com-\ncome more vulnerable in several ways.\nmunity health centers, nine youth and family\nMaintaining market share. First, even\nData, 1991 through 1996). 1995 (Colorado\nthough Parkland prevents many unnecessary\nssociation,\nJuly\nAnderson, a physician, is president and chief executive officer of Parkland Memorial Hospital in Dallas,\nPaul Boumbulian is senior vice-president of Parkland.\n4\nAFFAIRS - July/August 1997\nto People Health Foundation, Inc.\nPERSPECTIVE\nhospital admissions, it does not \"save money,\"\nsons who lack coverage.\nbecause beds that would be empty in a closed\nBusiness versus safety-net ethic.\nHMO system are filled at Parkland with un-\nWhile Parkland has attempted to improve the\nmanaged, uninsured, high-cost patients who\nhealth status of its patients and the commu-\nfell out of the market basket.\nnity, others have been busy trying to refine the\nSecond, as traditional commercial insur-\n\"sick care\" system-learning how to trans-\nance has dwindled and managed care has re-\nform clinical services into commodities and\nduced the margins on other patients, the\npatients to \"covered lives.\" Risk selection has\nMedicaid market (especially uncomplicated\nbecome far more profitable than true innova-\nobstetrical cases) has become more appealing\ntion. We are finding that what is a good com-\nto other hospitals with excess capacity. Com-\nmodity may not be what is good for the com-\npetition for these patients has resulted in a\nmunity's health. It is difficult for the public to\nloss of approximately 2,000 deliveries per year\nunderstand what is to be gained in preserving\nat Parkland.\nsafety-net \"utility\" services such as trauma,\nThrough the efforts of the medical school\nburn, and neonatal care and specialized drugs\nfaculty, Parkland has a low cesarean-section\nfor patients with human immunodeficiency\nrate (18 percent), whereas some of the recent\nvirus (HIV)/acquired immunodeficiency syn-\nentrants into the Medicaid market have rates\ndrome (AIDS), as few citizens have direct\nmore than twice as high. Parkland and the\ncontact with these services. Most healthy\nmedical school also have achieved an infant\npeople never consider their vulnerability to\nmortality rate for African Americans that is\ninjury or catastrophic illness, much less that\nessentially equal to the rate for whites (na-\nof their community to natural and man-made\n28\ntionally it is two and one-half times higher).\ndisasters.\nHowever, this breakthrough has not been ac-\nBuilding social capital. The means for\nknowledged by the local market, where other\naddressing the determinants of health reside\ninstitutions trade on amenities rather than\nnot in the efficiencies of the market but in\noutcomes. Women and their newborns who\nwhat Robert Putnam termed social capital, that\nare served by other institutions are routinely\nis, the bonding that characterizes a strong\ntransferred to the Parkland system after\ncivil society.² Communities with rich net-\nMedicaid is discontinued or if complications\nworks of nonmarket relationships provide a\narise.\nsense of trust, sharing, and cooperation that\nThird, Parkland always will be subject to\nincreases the efficiency of human relation-\nadverse selection by virtue of its mandate:\nships. Yet most of the dollars available to affect\nAny patient can enter the system by being\nhealth status remain in the business of curing.\nsick or injured any day of the year.\nSome of the public systems that make up\nThe Medicaid situation is poised to be-\nthe safety net, a few of the not-for-profit insti-\ncome even more distressing. Parkland now\ntutions, and a very limited number of the for-\nhas 60 percent of the Medicaid market share\nprofit systems have recognized this situation.\nin Dallas County. Under the proposed state\nThese foresighted organizations are begin-\nMedicaid managed care program, Medicaid\nning to establish partnerships with other\npatients will be distributed among three or\npublic or nonprofit institutions to improve\nfour managed care vendors, and although\nhealth. Current market incentives may\nParkland will inevitably be one of these\nlethally harm these institutions and leave in-\nproviders, it would be foolish to assume that\nefficient, ineffective, and bureaucratic institu-\nit can maintain its current market share. The\ntions unharmed. Policymakers should not\nloss of this revenue will jeopardize the sys-\npreserve publicly sponsored health care insti-\ntem's ability to provide educational, research,\ntutions for the sake of publicly sponsored\nand mandated services, particularly to per-\nhealth care. What should be preserved is a\nHEALTH AFFAIRS Volume 16, Number 4\nPERSPECTIVE: SAFETY NET\nsystem of choice by patients. Monolithic pal-\nwhich creates a windfall for the managed care\naces of patronage and institutions that treat\norganization.\npatients as captives in a disease-oriented sick\nLegislative support also will be required to\ncare system should be reengineered, redi-\nsupport safety-net pricing, especially with re-\nrected, and held to high standards of account-\ngard to pharmaceuticals for special popula-\nability for clinical outcomes, patient satisfac-\ntions. We are especially concerned about\ntion, and community health status. A civil\npharmaceutical support for HIV/AIDS popu-\nservant mentality must give way to servant\nlations because their care is often provided by\nleadership and community partnership.\nonly a few hospitals in a given area. Pharma-\nceuticals are part of the safety net but are not\nSummary And Recommendations\nalways provided at present.\nWithout these structural changes, we will\nThe current structural problem, wherein\ncontinue to see the results of failing federal\nsafety-net institutions must operate not only\nand state policies into the next millennium.\nin the marketplace but also under state man-\nTheir failure will ultimately come down hard\ndates to provide utility services, must be rec-\non local governments as unfunded mandates.\nonciled through structural change. That\nCommunities strapped for revenues to meet\nmeans that an all-payer system is required to\nchallenges of devolution ultimately will be re-\nsupport medical education and provide for the\nquired to reduce services. Unless some funda-\n\"stand-ready\" costs of services such as trauma,\nmental change is made, they will focus only on\nburn, and neonatal intensive care. Reimburse-\nthe \"resurrection medicine\" of the emergency\nment strategies must acknowledge not only se-\nroom and pull back from real innovation and\nverity of illness but also social severity. The pro-\nreal reform capable of improving the health\n29\nvision of care to large numbers of low-income\nand productivity of all of our nation's resi-\npersons carries a real cost for \"wraparound\"\nservices that lower barriers to care.\ndents. We do have a structural problem that\nneeds a structural solution, but we also need a\nSafety-net institutions should also provide\nchange in our hearts to better understand that\nincentives to maintain or create a community\nwe cannot leave needed reform of the health\nhealth care infrastructure that enhances the\ncare delivery system to the invisible hand of\nhealth status of the community, particularly\nthe marketplace.\nin underserved inner-city and rural areas. This\ncould include the development and mainte-\nA version of this paper was presented at The Robert\nnance of comprehensive primary care net-\nWood Johnson Foundation/Health Affairs/Alpha\nworks in underserved communities. Reim-\nCenter conference, \"What Is Happening to the Safety\nbursement methodologies and grant awards\nNet?,\" '9-10 January 1997, in Washington, D.C.\nshould foster collaborative models and not\ncreate disruptive competitive behaviors that\nNOTES\nundermine community partnerships.\n1. Y.M. Coyle and P.J. Boumbulian, Parkland's COPC\nFunds need to be made available for\nEvaluation System (Unpublished quality manage-\nment report, 1995).\noutcomes-oriented research. Both safety-net\n2. R.D. Putnam, \"The Prosperous Community: So-\nand other providers must be able to show that\ncial Capital and Public Affairs,\" The American Pros-\ninvestments of new dollars or redirected dol-\npect (Spring 1993): 2; and R.G. Evans and G.L.\nlars lead to desired results. Safety-net institu-\nStoddard, \"Producing Health, Consuming\ntions should directly receive direct and indi-\nHealth Care,\" Social Science and Medicine, 31, no. 12\nrect payments for medical education and\n(1990): 1347-1363.\ndisproportionate-share adjustments for car-\ning for a large volume of low-income persons\ninstead of leaving these payments in the ad-\njusted average per capita cost (AAPCC),\nHEALTH AFFAIRS July/August 1997\nMEDICAID MANAGED CARE\nnd. I've been to NIH, and\ngoing on there, which is\nhave NIH and its advisory\nMedicaid Managed Care\nsions on research funding.\nfiscal 1998, do you favor\nAnd Community\nthorization?\nfunded. The more of these\nProviders: New\nrs delve into-such as the\nhe potential for improving\nH has reasonable funding to\nPartnerships\nCan the \"arranged marriage\" of managed care plans with\ntraditional community-based providers improve care for the poor?\nring most of your working\nday? Is it as satisfying as it\nby Debra J. Lipson\nappreciated by the public\nust of government a healthy\nPROLOGUE: Partnerships between organizations are often\nor do you view it otherwise?\ncompared to marriages-some are based on necessity, some on\nem now of being presumed a\nconvenience, some on synergy, and others on arrangements made\nheartening. But I get so much\nby third parties. Harvard professor Rosabeth Moss Kanter\nrough to correct things. It is\ndescribes five stages in the development of organizational\nprogress, to feel good about\nalliances: engagement, selection or courtship, setting up\nSAFETY NET\n91\nThat is so rewarding that I\nhousekeeping, learning to collaborate, and changing within. The\nrious about your long friend-\nability of organizations to progress through these stages depends\nrity leader, who's obviously a\non the dynamics between them and around them. But ultimately,\nare. How would you charac-\nthe endurance of an organizational partnership, just as some\nerybody knows that you sing\nwould say of marriage, depends on the continued mutual benefit\nor is it just\nof the alliance to each of the participants.\ny-two years now, first in the\nThis paper by Debra Lipson explores the \"arranged marriage\"\nm, and I think he respects me.\nof community-based health care providers and managed care\nhe feels, and we have main-\nplans. Although these two entities make strange bedfellows,\nknows that I can get things\nthey have courted each other for the opportunity to draw on\nthings done.\nelped you repel some of the\neach other's strengths in an increasingly competitive\ning of the chairmanship?\nmarketplace. As they come together to negotiate contracts for\nhat mainly his help was based\nserving Medicaid patients, will they learn to collaborate or find\nit things done. He has a sincere\na way to fulfill their missions and financial objectives\nparty but on a bipartisan basis.\nindependently? Here Lipson looks at what is at stake and at the\none. We don't always agree on\nprospects for their success.\nght to be done, but we share a\nUntil June of this year, Lipson was associate director of the\nAlpha Center, a nonprofit health policy research organization in\nWashington, D.C. In honor of her own commitment to\nmarriage, she has since moved with her husband to Geneva,\nSwitzerland, where she will complete studies in progress and\nseek new alliances with international health organizations.\nr\n+\nHEALTH AFFAIRS July/August 1997\n1997 The People-to-People Health Foundation. Inc.\nSAFETY NET\nABSTRACT: Growing enrollment in managed care plans among Medicaid recipi-\nents represents a new market for these plans but presents challenges to those\nproviders that traditionally have served this population. To continue serving\nMedicaid patients, community-based providers must develop contracts or other\ntypes of partnerships with Medicaid-contracting health plans. This paper reviews\nthe challenges to such collaboration and discusses the practical issues that\nplans and community-based providers must resolve to develop productive work-\ning relationships. Keys to successful collaboration are identified. Ways in which\nfederal and state governments can help the collaborative process are suggested.\nY JUNE 1996, 13.3 MILLION Medicaid recipients were en-\nB\nrolled in some form of managed care, 40 percent of all Medic-\naid eligibles and nearly twice the proportion two years earlier.¹\nThe growth curve will surge upward again in 1997 as large states\nsuch as California, Illinois, and New York attempt to enroll large\nnumbers of Medicaid recipients into managed care plans or expand\nMedicaid managed care programs into additional counties. States\nbegan their enrollment in managed care with women and children\nwho qualify for Medicaid, and they are turning their attention to\nelderly and disabled recipients, who have been largely exempt from\nmandatory managed care enrollment policies.\n92\nMEDICAID\nThe growth of Medicaid managed care has spawned new interest\non the part of private managed care plans in serving the Medicaid\npopulation. In some cases, plans that have contracted with Medic-\naid for many years are expanding their capacity to serve additional\npatients. However, as more states jump on the Medicaid managed\ncare bandwagon, plans without prior experience in serving Medic-\naid patients and brand-new plans are diving into the Medicaid mar-\nket as never before. Thirty-eight percent of health maintenance or-\nganizations (HMOs) and preferred provider organizations (PPOs)\nsurveyed by the Group Health Association of America (now the\nAmerican Association of Health Plans) in 1995 participated in\nMedicaid at the end of 1994; another 11 percent began serving\nMedicaid patients in 1995, and another 12 percent intended to do so\nin 1996.² This trend is good news for Medicaid officials, who had\ntrouble recruiting enough plans and providers to serve Medicaid\npatients.\nHowever, there are a number of concerns about the rapid entry of\nso many new plans into Medicaid managed care. For example, plans\nnew to the Medicaid market sometimes have underestimated the\namount of enabling services, such as case management, language\ntranslation, and outreach, needed to serve the Medicaid population.\nThe entry of such plans into the Medicaid market also raises con-\ncerns about how community-based providers that previously had\nlarge Medicaid caseloads will fare under these new arrangements.\nHEALTH AFFAIRS Volume 16, Number 4\nMEDICAID MANAGED CARE\nTraditional community-based providers are the organizations\nthat historically have provided much of the primary and preventive\ncare delivered to Medicaid recipients and the uninsured.³ These\nproviders include approximately 635 community and migrant\nhealth centers with 1,647 sites; more than 3,000 rural health clinics;\nnearly 3,000 city and county health departments; and a multitude of\nmaternal and child health clinics, specialty clinics for children with\nspecial health care needs, school-based clinics, family planning clin-\nics, and acquired immunodeficiency syndrome (AIDS) care provid-\ners (operating under the Ryan White Comprehensive AIDS Re-\nsearch and Education [CARE] Act). Although each community has\na different mix of such providers, some or all of them form the\nprimary and preventive \"safety net\" for the poor and uninsured.\nApproximately 43 percent of patients in federally qualified commu-\nnity and migrant health centers are covered by Medicaid, Medicare,\nor other public insurance, and an equal proportion are uninsured.\nTo survive, the centers rely heavily on federal, state, or local govern-\nment grants and to a lesser extent on private donations.\nGiven their dependence on Medicaid, it is imperative that\ncommunity-based providers participate in evolving Medicaid man-\naged care plans and networks. If they do not, they may lose the\nSAFETY NET\n93\nability to serve Medicaid patients, which in turn could threaten\ntheir ability to serve uninsured patients.\nCommunity-based providers have two major strategies for par-\nticipating in Medicaid managed care. They can form a managed care\nplan by themselves or in partnership with others, or they can seek\ncontracts with managed care plans. This paper focuses on the sec-\nond strategy. First, I explain the issues that have arisen as Medicaid\nmanaged care plans seek to develop provider networks that can\nserve the diverse Medicaid population and negotiate contracts with\ntraditional community-based providers. I then highlight some of the\nstrategies and elements that have proven effective in overcoming\nbarriers to mutually beneficial contracts. I conclude by discussing\nthe policy implications for federal and state governments in their\nrole as purchasers of care for the Medicaid population.\nChallenges And Opportunities For Collaboration\nWhy would private managed care plans want to contract with\ncommunity-based providers of primary and preventive care to\nMedicaid patients? Primarily because they share the same mis-\nsion-providing high-quality, easily accessible primary and preven-\ntive care services to Medicaid patients. But they differ in how they\napproach this mission, based on different organizational cultures\nand strengths.⁵ Just as state health departments and Medicaid agen-\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\ncies used to engage in what was once called the \"Cold War between\nthe health zealots and the Medicaid infidels,\" community-based\nproviders and managed care plans today often view each other with\nsuspicion.6\nManaged care organizations (MCOs) typically have experience\nin claims processing, provider relations, utilization review, and fi-\nnancial risk management. As insurers, they are focused on purchas-\ners and enrollees, most of whom have been commercially insured. If\nnonprofit, they are accountable for financial performance to boards\nof directors; if for-profit, they are accountable to stockholders. In\ncontrast, community-based providers have experience in providing\npersonal health care services to \"vulnerable\" persons-those with\nlow incomes, special health care needs, or cultural or language barri-\ners to care-regardless of ability to pay. It is their mission to provide\ncare to the uninsured; in some cases they are obligated to do so. They\nare accountable to consumer-dominated boards of directors or to\ngovernmental bodies if they are publicly sponsored or funded. These\ndifferent cultures and strengths present both challenges and oppor-\ntunities for collaboration.\nChallenges. The two organizations come to the negotiating\n94\nMEDICAID\ntable speaking different languages, using the same words to mean\ndifferent things. For example, to MCOs, \"case management\" is the\nresponsibility of primary care physicians to coordinate and author-\nize all of the care needed by patients. But to community-based\nproviders, case management is often the responsibility of social\nworkers, nurses, or community health aides to assess patients'\nneeds, refer patients to all appropriate services, and advocate on\npatients' behalf to assure access to care.\nThere are also many stereotypes that must be broken down be-\nfore real collaboration can occur. MCOs may assume that all\ncommunity-based providers serve high-risk, sicker populations,\nwhich would increase the risk of adverse selection for MCOs that\nhave community-based providers in their provider network. There\nmay also be the misconception that most community-based provid-\ners are unable to comply with HMOs' utilization controls and qual-\nity assurance processes. Some community-based providers believe\nthat MCOs' only concern is the bottom line and that MCOs are not\ncommitted to providing the full range of services needed to effec-\ntively serve vulnerable populations.\nOpportunities. Despite these differences, private managed care\nplans and community-based providers are drawn to each other for a\nnumber of reasons, even if it sometimes seems more like an arranged\nmarriage than love at first sight. MCOs find community-based\nproviders attractive because they have experience serving the\nHEALTH AFFAIRS - Volume 16, Number 4\nMEDICAID MANAGED CARE\nMedicaid population, a powerful advantage for any MCO that has\nnot worked with this group of patients before. \"Community provid-\ners have leverage, if they're good,\" said David O'Brien of the Gateway\nHealth Plan in Pittsburgh. \"If a community-based provider func-\ntions well and is well-thought-of in a particular community, a plan\nwill need to include that provider in its network if it wants to\nattract members from that community.\" MCOs readily acknowl-\nedge that they do not know how to deliver care to a less-educated\npopulation, how to help patients get to appointments, and how to\ncommunicate the importance of prevention and early treatment.\nThey find that community-based providers are particularly strong in\nproviding enabling services. \"The most I've learned from anyone\nabout the Medicaid population is from the community-based\nproviders,\" said John Monahan of Blue Cross of California. In addi-\ntion, MCOs find that community-based providers may be among the\nfew, or perhaps the only, providers of primary care services to\nMedicaid patients in some neighborhoods. One plan turned to\ncommunity-based providers after it tried to sign up physicians in its\ncommercial plan to serve Medicaid patients but was unable to con-\nvince enough physicians to serve the number of such patients ex-\npected to enroll. In other communities, community-based providers\nSAFETY NET\n95\nmay be the only source of particular services required by the state\nMedicaid agency, such as maternity care coordination or human im-\nmunodeficiency virus (HIV) counseling. Finally, MCOs may perceive\nefficiencies in having low-income residents, who frequently go on\nand off Medicaid, see the same provider regardless of their eligibility.\nCommunity-based providers find that they need to court MCOs\nfor many reasons, as well. As mentioned earlier, contracts with\nMCOs are essential to continue provider/patient relationships with\nMedicaid beneficiaries and to retain Medicaid revenues. This is par-\nticularly true if community providers cannot or choose not to form\ntheir own MCOs. Community-based providers find that being part\nof an MCO provider network gives them access to information sys-\ntems, management expertise, and specialty clinical services that\nmight otherwise be unavailable or unaffordable. Finally, the experi-\nence of dealing with Medicaid managed care can give community-\nbased providers the tools they need to be more competitive.\nMajor Contracting Issues\nAs plans and community-based providers dance closer together, they\nare just starting to come to grips with the details and issues that go\ninto contracting. Before contracts can be signed, community-based\nproviders and MCOs must decide on (1) the role and authority of the\nprovider in relation to other parts of the delivery system; (2) which\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\nquality and basic operating standards must be met and how compli-\nance will be monitored; and (3) the basis on which payments will be\nmade. Finding common ground on these issues is rarely easy.\nProviders' scope of responsibilities. Depending on what is\nincluded in the contract between a state Medicaid agency and man-\naged care plans, Medicaid MCOs need to arrange for a wide array of\nclinical and supportive services. It is administratively simpler for\nMCOs to contract with providers that can deliver a comprehensive\nset of services. But other than community health centers, most\ncommunity-based providers can only deliver pieces of the package.\nMCOs that contract with multiple community-based providers\nmust decide whether preauthorization from a primary care physi-\ncian will be needed for every visit, which services require prior\nauthorization, whether there are certain conditions for which pa-\ntients need not get authorization from their primary care provider,\nand whether they will guarantee a minimum number of referrals to\nthe community-based provider. The two parties must also decide\nhow case management will work.\nQuality assurance and reporting requirements. Defining\n\"high-quality\" care can be difficult when the two parties are ac-\n96\nMEDICAID\ncountable to different accreditation bodies that emphasize different\naspects of quality, have different standards, and require reporting of\ndifferent indicators and measures. Health plans are oriented to Na-\ntional Committee for Quality Assurance (NCQA) standards and\nHealth Plan Employer Data and Information Set (HEDIS) measures,\nwhereas community-based providers are accountable to a variety of\ngovernment and private organizations and report various data de-\npending on funding sources. For example, community and migrant\nhealth centers follow federal rules, local health departments use the\nAssessment Protocol for Excellence in Public Health system to as-\nsess and improve their organizational performance, and Ryan White\ncenters have their own federal requirements.\nIndividual providers also complain about the conflicting require-\nments imposed by health plans. Zoila Torres-Feldman, a community\nhealth center director in Massachusetts, said, \"Each plan has differ-\nent formularies, different labs, separate utilization review staff, dif-\nferent benefit packages, different hospitalization contracts, differ-\nent networks of specialists, and different expectations of the levels\nof provider control.\"\nMCOs sometimes criticize the capacity of community-based\nproviders to meet basic operational expectations, such as submit-\nting utilization and quality data in formats that are consistent with\nplans' information systems, answering telephone lines in a timely\nfashion, and making appointments available within required time-\nHEALTH AFFAIRS Volume 16, Number 4\nMEDICAID MANAGED CARE\n\"Payment between plans and community-based providers is one of\nthe most important and difficult issues in contract negotiations.\"\nlines. According to Herb Wheeler of HealthKeepers, Inc., a subsidi-\nary of Trigon Blue Cross Blue Shield, in Richmond, Virginia, \"In\ntrying to gain NCQA accreditation, we feel like we're being pulled\ndown if we deal with community health centers.\" He and other\nhealth plan officials say that not all community-based providers'\nrecord keeping and quality of care are yet up to NCQA's credential-\ning standards.\nPayment arrangements. Payment between plans and\ncommunity-based providers is one of the most important and diffi-\ncult issues in contract negotiations. Community-based providers\ntypically want fees to cover the costs of providing all care to patients,\nwhereas MCOs want to pay the same amount they pay other provid-\ners for the same services. Some MCOs complain that community-\nbased providers do not have adequate information systems to deter-\nmine unit costs, and some community-based providers say that\nMCOs do not want to pay adequately for the additional services\nthey provide.\nSAFETY NET\n97\nMost plans and community-based providers believe that few\nsuch providers are equipped to accept financial risk for the entire\nspectrum of services patients need. Their data systems are ill-\nequipped to manage utilization, their financial systems are not so-\nphisticated enough, and sometimes they are legally unable to suffer\neven short-term financial losses. \"Community health centers could\nsurvive if they can share in the savings they generate for the health\nplans,\" said Torres-Feldman. But if community-based providers are\nonly capitated for primary care, they could lose money.\nStrategies For Contracting Collaboration\nThe growing number of contracts between MCOs and community-\nbased providers indicates that many have overcome the obstacles in\norder to gain the potential benefits. The examples presented here\ndemonstrate that the secret to success lies in translating contract\nrequirements into practical working relationships.\nProviders' scope of responsibility. In some cases MCOs con-\ntract with community-based providers for a limited set of services\ninitially and increase the scope of services included in the contract\nas the providers' expertise expands. For example, Blue Cross of\nCalifornia's HMO initially contracted with a Planned Parenthood\nclinic for just family planning and obstetrical services but expanded\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\nthe contract to include primary care services when the clinic devel-\noped the capacity to provide them. When community-based provid-\ners can deliver only a specific component of the service package,\nsome plans have authorized the providers to deliver all necessary\ncare without requiring referrals for each visit. In a study of seven\nschool-based clinics, for example, nineteen of twenty-six health\nplan contracts authorized the clinics \"to provide a specific set of\nservices-typically well-child visits and acute, episodic sick\ncare-to the plan's beneficiaries without preauthorization by the\nplan.\" The Health Start school-based clinics in St. Paul, Minnesota,\ndeveloped three different models for relating to primary care provid-\ners in health plans, depending on each plan's particular needs.8\nQuality assurance and reporting requirements. NCQA ac-\ncreditation standards and HEDIS 3.0 measures, which recently in-\ncorporated Medicaid HEDIS, are becoming the industry standard.\nAs more Medicaid agencies use them to hold contracting health\nplans accountable, they have become de rigueur for plans and their\nproviders to follow. But how plans meet such standards may differ\nfrom one plan to the next. When working with community-based\nproviders, plans often find that they must work closely through\n98\nMEDICAID\njoint quality assurance committees, provide data that show provid-\ners how their practice patterns compare with those of others, and\nagree on priorities for improvement. For example, one publicly\nsponsored health plan (CareOregon) holds monthly quality im-\nprovement and utilization committee meetings with community\nhealth centers in its network to educate both plan and clinic repre-\nsentatives about meeting quality assurance requirements.\nOn the other hand, some community-based providers may be able\nto show other providers in a plan's network how to meet certain\nstandards. As John Bartkowski, a community health center director\nin Wisconsin, pointed out, \"The health centers bring to the HMOs\nthe ability to meet Early and Periodic Screening, Detection, and\nTreatment [EPSDT] requirements. We also were able to demon-\nstrate that health centers are much better on certain outcomes [for\nexample, emergency room visits, low-birthweight rates] than other\ncontracted physicians.\"\nStill, plans sometimes find that community-based providers re-\nquire technical assistance and funds to meet quality, utilization, or\nother reporting expectations. One health plan that was created by\ncommunity health centers allocates half of its annual earnings to\ncommunity providers to invest in infrastructure; in recent years\nmanagement information systems have been the priority for these\nfunds. A commercial health plan has provided computer hardware\nand software to several community-based providers to enable them\nHEALTH AFFAIRS Volume 16, Number 4\nMEDICAID MANAGED CARE\nto bill for services and comply with prior-authorization require-\nments. In addition, MCOs may provide community-based providers\nwith information about their patients' overall utilization patterns to\nhelp manage care more effectively and EPSDT check-up lists to help\nwith follow-up. Trigon provided funds to a group of minority physi-\ncians to help them organize into a group, hire administrative staff,\nand collect data.\nPatience and willingness to educate community-based providers\nabout administrative practices are also necessary. One plan official\nexpressed frustration with county health departments that have inade-\nquate or nonexistent appointment systems and other community-\nbased providers that cannot meet reporting or performance stan-\ndards. But the plan is giving these providers time to get up to speed\nbecause they are regarded as valuable partners for ensuring access\nand service capacity. Community-based providers are clearly will-\ning and able to change. One study found that contracts with man-\naged care plans had prompted management improvements at com-\nmunity health centers, such as strengthening twenty-four-hour\non-call service, more timely scheduling of appointments, and more\nactive monitoring of utilization and specialty referrals.¹⁰\nPayment arrangements. Flexibility is the key to setting suit-\nSAFETY NET\n99\nable payment arrangements and rates. Many contracts between\nplans and community-based providers still include fee-for-service\npayments for each contracted service, based on previous Medicaid\nfee schedules, fixed global visit rates, or a service-specific rate nego-\ntiated by the provider and the plan. However, some plans have de-\nveloped successful risk contracts with community-based providers.\nPlan and provider representatives caution that to work well, such\narrangements require (1) experience on the part of providers, in\nmanaging risk, and (2) substantial numbers of enrollees seeing the\nprovider. As Robert Gomez, head of a community health center in\nTucson, Arizona, said, \"If you can talk capitation, you've broken a lot\nof the cultural barriers between us and the plans.\"\nEven when providers do not have prior experience in risk as-\nsumption, some form of risk sharing can still be developed. For\nexample, in the first year of St. Paul Health Start's contracts with\nMCOs, the plans held back a small portion of the rates paid to other\ncapitated clinics to fund school clinics while both parties collected\nutilization data to set future rates. Another plan has several shared\nrisk contracts that \"carve out capitation for services that CBPs\n[community-based providers] perform and pay the rest on a fee-for-\nservice basis which we manage ourselves,\" according to Dan Bailey\nof Compcare Health Services, an HMO owned by United Wisconsin\nServices, Inc., the Blue Cross Blue Shield affiliate in Wisconsin.\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\nSome unusual payment arrangements are found in contracts be-\ntween plans and community-based providers, which again reflects\nthe special needs of some providers and the willingness of plans to\nbe flexible. For example, Blue Cross of California pays for some\nservices provided by its community-based providers up front to help\nwith cash flow. The cost of intensive outreach such as home visits\nfor preventive services is not usually included in capitation rates, so\nhealth plans and community providers sometimes cooperate in\nseeking additional funding through public or private grants. The\nBlue Cross Blue Shield plan in Wisconsin developed a risk-sharing\nagreement with providers on EPSDT to give providers more incen-\ntive to help the plan meet its screening goals.\nWhile plans' payment arrangements with federally qualified\nhealth centers are governed to a great extent by federal and state\nregulations, which I discuss later, health plans and health centers\nstill have some options that form the basis for rate negotiations. For\nexample, the Blue Cross plan in California convinced one center to\nwaive cost-based reimbursement and accept capitation in exchange\nfor the promise of a larger volume of patients. It is not clear whether\nthe non-cost-based payment rates will be adequate in the long run.\"\n100\nMEDICAID\nKeys To Successful Contracting\nThis brief review highlights several factors that are critical to suc-\ncessful Medicaid contracting.¹²\n(1) The parties to a contract must learn as much as they can about\neach other-their language, values, and goals. The differences that\nexist among MCOs and community-based providers are important\nwhen considering the pros and cons of a particular contract. Local\nmarket factors also influence the dynamics of contracting.\n(2) To meet the needs of Medicaid patients, both parties must be\nwilling to cast aside \"business as usual.\" Health plans that believe\nthat community-based providers are valuable partners must be will-\ning to trust what has been learned and invest in the services that are\nneeded to make a difference (for example, offering services in multi-\nple languages, respecting the influence of culture on health-seeking\nbehavior, and developing special exceptions for confidential serv-\nices). Plans may need to stretch usual primary care physician\nauthorization requirements, while community-based providers may\nneed to hire new staff to perform the tasks required by MCOs.\n(3) Plans and community-based providers must be willing to\nmeet regularly to work out the details of collaboration. Even when\nmanagers agree to referral and authorization procedures, there will\nalways be special cases that require accommodation and compro-\nmise. It may be useful to start with something small before embark-\nHEALTH AFFAIRS Volume 16, Number 4\nMEDICAID MANAGED CARE\ning on a long-term contract.\n(4) Plans and community-based providers must be willing to\nexperiment with different payment arrangements. Because not all\ncommunity-based providers are alike (some can handle capitated\nrates, others cannot), and because each state's Medicaid program is\ndifferent (capitation rates in some states may be meager, some have\nmore sophisticated risk-adjustment mechanisms than others have),\nMCOs and community-based providers must work together to re-\nsolve the array of payment issues that will arise.\n(5) Community-based providers have shortcomings in informa-\ntion systems and facilities, but these can be improved. These provid-\ners have limited capital budgets to invest in computer systems or\nfacility improvements. Plans must be willing either to invest in com-\nmunity-based providers' infrastructure or help them to find alterna-\ntive sources of capital. Community-based providers must put a high\npriority on finding the resources and technical staff to bring their\nfacilities and systems up to speed.\n(6) Continuous quality improvement requires partnerships.\nCommunity-based providers that deliver high-quality services and\nhave a good reputation in their community should have little diffi-\nculty contracting with Medicaid MCOs. Those that do not continu-\nSAFETY NET\n101\nously improve quality may not belong in Medicaid managed care. To\nimprove quality, plans, providers, consumers, and state agencies\nmust join together to collect and analyze data, conduct patient sat-\nisfaction surveys, stress the importance of EPSDT screenings, and\nimprove risk-adjustment methods.\nPolicy Context\nContracts between Medicaid managed care plans and community-\nbased providers do not occur in a vacuum. They are greatly affected\nby federal and state Medicaid laws and policies and by the specific\nlanguage in the contracts between state Medicaid agencies and\nmanaged care plans. Thus, the relationships that evolve between\nMedicaid managed care plans and community-based providers will\nin part reflect the federal and state policies that govern them.\nFederal policy. Most of the formal policies related to community-\nbased providers' relationships with MCOs emanate from the Health\nCare Financing Administration (HCFA) through its administration\nof state Medicaid waivers. Several other Department of Health and\nHuman Services (HHS) agencies are involved in providing technical\nassistance.\nState waivers of federal Medicaid regulations. Federal Medicaid law\n(Sections 1115 and 1915(b) of the Social Security Act) grants HCFA -\nthe authority to grant waivers to states that wish to be exempt from\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\ncertain rules.¹³ States with 1915(b) waivers must assure cost-based\nreimbursement to federally qualified health centers participating in\nmanaged care when they affirmatively assert their right to it. But\nbecause states requesting Section 1115 waivers can ask to waive\nhealth center payment rules, many centers have been forced to ac-\ncept less than they formerly received. Even so, many of the states\nwith Section 1115 waivers still make some provisions that help feder-\nally qualified health centers to participate in or get adequate pay-\nment from Medicaid managed care plans.\nIn approving a state's Medicaid waiver program, HCFA may im-\npose special terms and conditions with which the state must com-\nply. Most often, these terms and conditions relate to rules that\nMCOs must follow in reimbursing community-based providers and\nto assuring access to services provided by community-based provid-\ners. For example, HCFA's special terms and conditions often include\na requirement that MCOs contract with community-based provid-\ners to assure that MCOs have adequate capacity to serve Medicaid\npatients. If MCOs have adequate capacity to serve vulnerable popu-\nlations without contracting with federally qualified health centers,\nMCOs are relieved of this requirement. If health centers develop\n102\nMEDICAID\ntheir own Medicaid managed care plan, other MCOs in the same\nservice area are also exempt from this condition. In most states\nHCFA also requires states to adjust rates for patient case-mix.\nTechnical assistance and training for community-based providers and MCOs.\nHHS agencies-including the Bureau of Primary Health Care, the\nMaternal and Child Health Bureau, the Centers for Disease Control\nand Prevention, and the Agency for Health Care Policy and Re-\nsearch-conduct comprehensive training and technical assistance\nprograms to help community-based providers to participate in or\nrelate to Medicaid managed care programs. These agencies also pro-\nvide information and training to MCOs on developing delivery sys-\ntems for Medicaid patients.\nState policies. Within the context of federal policies and\nguidelines, states have adopted a range of approaches to ensure a\nrole for community-based providers in Medicaid managed care\nplans. Some states have designed systems to assure that community-\nbased providers are given the chance to develop their own plans. For\nexample, California formulated a managed care model \"designed in\nlarge part to protect safety-net institutions.\"¹⁴ Medicaid recipients\nin twelve densely populated counties will choose between two dif-\nferent plans-a commercial plan and a Local Initiative plan that is\ndeveloped by county governments and safety-net and traditional\nMedi-Cal (California Medicaid) providers, along with community\nrepresentatives. \"Unlike the commercial plans, Local Initiatives are\nHEALTH AFFAIRS - Volume 16, Number 4\nMEDICAID MANAGED CARE\nrequired to contract with safety-net and traditional providers.\"⁵\nRates paid to the Local Initiative plans are higher to reflect cost-\nbased payments made to federally qualified health centers.\nHowever, most states have not gone this far. Instead, they are\nusing one or more of the following more limited strategies to encour-\nage partnerships between community-based providers and MCOs.\n(1) Contract requirements. Some states explicitly require that Medic-\naid managed care plans contract with particular community-based\nproviders. For example, Oregon mandates that Medicaid managed\ncare plans contract with county health departments and other pub-\nlicly funded programs to provide immunizations, as well as screen-\ning for sexually transmitted and other communicable diseases. Ex-\nceptions are allowed when community providers are not available.¹⁶\nMinnesota's Medicaid program requires that at least one MCO in a\ngiven service area include a community clinic in its network.\n(2) Preference points for collaboration. Arizona, California, and Massa-\nchusetts, among other states, add points to plans' bids if they in-\nclude community providers in their networks.\" For example, Mas-\nsachusetts specifies in its request for proposals that plans must\nattempt to establish linkages, though not necessarily contracts,\nwith school-based clinics and encourages coordination with public\nSAFETY NET\n103\nhealth clinics to serve persons with HIV infection.\n(3) Preferential assignment to plans that include community providers. In\nsome states Medicaid MCOs that include community-based provid-\ners are assigned patients who do not voluntarily choose a plan. For\ninstance, in California, the Orange County CalOPTIMA (Orange\nPrevention and Treatment Integrated Medical Assistance) program\nhas made the plan organized by Children's Hospital of Orange\nCounty and the University of California-Irvine the default for per-\nsons who fail to select a plan.¹⁸\n(4) Setting quality or access standards to encourage collaboration. Certain\nperformance standards in a state's contracts with health plans can\nincrease the likelihood that plans will contract with community-\nbased providers. For example, Minnesota requires health plans to\nsubmit annual \"action plans\" that specify how they will deliver\nservices, provider capacity, quality improvement plans, and policies\nand procedures for serving high-risk or special-needs populations.\nThis serves as an incentive for plans to work with community-based\nproviders that can provide care to special-needs groups.\n(5) \"Transitional\" payments and other enhanced reimbursement to\ncommunity-based providers. To help federally qualified health centers or\nother essential community providers to develop the systems and\neconomies of scale needed to assure cost-effectiveness while con-\ntinuing to deliver health care to all persons regardless of their ability\nHEALTH AFFAIRS - July/August 1997\nSAFETY NET\nto pay, many states have developed transitional payments to feder-\nally qualified health centers. For example, Rhode Island pays an\nextra $10 per member per month to the center's HMO for each\nMedicaid recipient enrolled. And in Hawaii and Vermont, federally\nqualified health centers in Medicaid managed care plans are entitled\nto a \"wraparound\" payment to settle the difference between capita-\ntion and cost. In addition, some states provide enhanced capitation\nrates to plans that contract with federally qualified health centers to\ncompensate for a more costly patient case-mix. Minnesota adjusts\nfor a more costly case-mix in some safety net-sponsored plans.19\nNext Steps\nFederal and state policymakers often find themselves caught be-\ntween the plans, which argue that strict governmental requirements\nto contract with community-based providers or pay them certain\nrates hurt the plans' ability to provide cost-effective care, and\ncommunity-based providers, which assert that the government\nmust protect them from competition that could threaten their finan-\ncial viability. In an effort to find middle ground, HCFA and most\nstate Medicaid agencies have tried to ensure, at a minimum, that\n104\nMEDICAID\ncommunity-based providers are given a chance to participate in\nMedicaid managed care plans, even if they cannot guarantee their\nultimate success. After all, federal and state governments have in-\nvested in community-based providers through grants, loans, and\nother support for several decades. If community-based providers are\nforced to close as a result of competitive forces, and private managed\ncare plans decide to abandon the Medicaid market because it is not\nsufficiently profitable (as some already have done), governments\nwill have to rebuild the entire care infrastructure for the poor. But if\ncommunity-based providers, which have always formed the safety\nnet, are given a greater chance of survival, they might constitute an\nimportant \"fallback.\"\nPolicies to improve provider competitiveness. If this re-\nmains the preferred policy choice, there are avenues in addition to\nthose already mentioned that might make community-based\nproviders more competitive. For example, subsidies for care of the\nuninsured and funds for capital investments (data systems and facil-\nity improvements) might make it easier for community-based\nproviders to accept discounted fees from managed care plans. Risk-\nadjusted rates to plans might make it more likely that they would\ncontract with community-based providers that serve sicker groups\nof people. For these to occur, MCOs might have to join with provid-\ners in the political process.\nThe question remains whether such policies are needed temporarily\nHEALTH AFFAIRS - Volume 16, Number 4\nMEDICAID MANAGED CARE\nor on a permanent basis. For example, some argue that it goes\nagainst the economic interests of MCOs to contract with commu-\nnity-based organizations that are not directly controlled by the\nMCO. Robert Reischauer of The Brookings Institution stated, \"A\nmanaged care entity that wants to operate efficiently [would want]\nto exercise complete control over the provision of services either\ndirectly or by subcontracting that control through a risk-based con-\ntract to some other provider.\" Based on this possibility, as MCOs\ngain more experience in serving the Medicaid population, they may\nfind it more efficient to operate a van service to bring Medicaid\nmembers to risk-assuming providers, rather than contract with\nneighborhood-based providers that are seeking higher payments to\nsupport their social mission.\nCommunity-based providers may find that as they become better\nat managing care and financial risk under capitation, they will be\nbetter able to form their own managed care plans. Hospitals and\nphysician groups are now forming integrated delivery systems and\nprovider-sponsored networks that are trying to compete directly\nwith MCOs; some community-based providers are emulating this\nstrategy. If they are successful at managing risk and winning con-\ntracts with Medicaid or other purchasers, they might be able to\nSAFETY NET\n105\nchannel the savings or profits into services for the uninsured.\nThere are those who believe that contracts between Medicaid\nMCOs and community-based providers will last because they bene-\nfit both parties and result in higher-quality service and better out-\ncomes for Medicaid recipients. Others caution that no matter what\nworks best economically or from a health care delivery perspective,\npolitics will remain influential. \"On the state level, which plans get\nMedicaid contracts has to do with politics. In communities, which\nproviders are chosen to get plan contracts has to do with politics,\"\nsaid JoAnne Fischer of the Maternity Care Coalition in Philadelphia.\nThe value of collaboration. Before plan/community provider\ncontracts can develop into long-term relationships, more \"dating\"\nmust begin. One way to encourage such pairing might be to provide\nevidence of the value of such collaborations. There are few, if any,\nstudies showing that collaborations have improved care to vulner-\nable populations or resulted in more cost-effective care. What have\ncommunity-based providers done that has helped MCOs to meet\nthe special needs of Medicaid clients? How have the resources and\nexpertise of MCOs improved the care of Medicaid enrollees?\nSolid, objective studies that answer these questions would do\nmuch to move the collaboration agenda. Indeed, evidence of the\ncost-effectiveness of individual community-based providers or par-\nticular outreach or educational methods could be the strongest sell-\nHEALTH AFFAIRS July/August 1997\nSAFETY NET\ning point of all. MCOs might not need any prodding to contract\nwith community-based providers if such proof were readily avail-\nable. On the other hand, evidence of Medicaid managed care plans'\nability to improve access to care, which is also in short supply, might\nconvince community-based providers of the value of collaboration\nwith private plans.²⁰\nOpportunities for discussion. As important as research and\npublic policy are to encouraging collaboration, it may be just as\nimportant to provide opportunities for representatives of private\nMedicaid managed care plans and community-based providers to\ndiscuss openly the differences that divide the groups and to sort out\nmutually acceptable roles, responsibilities, and mechanisms to en-\nsure accountability for the care of vulnerable populations. Further\nmeetings, held at the community level, may go a long way in moving\nthese \"arranged marriages\" toward true compatibility.\nVEN WITH SUCH STEPS, Medicaid managed care plans might\nE\nstill come to the conclusion that they can serve the Medicaid\npopulation without the help of community-based providers. If\nthat is the case, and federal or state policymakers still believe that\n106\nMEDICAID\npartnerships between the two sets of organizations are important,\nfinancial incentives or regulatory requirements may be needed to\nensure that such arrangements are established.\nThis paper was adapted from an issue brief prepared for a meeting sponsored by\nthe Health Resources and Services Administration (HRSA) of the U.S. Depart-\nment of Health and Human Services and the National Institute for Health Care\nManagement (NIHCM). The meeting, held 15 November 1996 in Washington,\nD.C., brought together representatives of private Medicaid managed care plans\nand community-based providers to discuss the issues that promote or hinder con-\ntracts between them and to explore how collaboration can improve care for vul-\nnerable populations. Any quotes in this paper that are not footnoted are drawn\nfrom the transcript of that meeting. The author appreciates the support of the\nNIHCM and HRSA for this paper and extends particular thanks to Kathleen\nEyre, Jeanne Ireland, Michael Sparer, Ed Neuschler, and Julia Tillman for their\nreview and comments.\nNOTES\n1. Health Care Financing Administration, \"Medicaid Managed Care Enrollment\nReport\" (Washington: HCFA, Office of Managed Care, 30 June 1996).\n2. American Association of Health Plans, 1995 HMO-PPO Trends Report (Washing-\nton: AAHP, 1995).\n3. Other safety-net providers include public and private hospitals, including\nacademic health centers, that serve disproportionate numbers of Medicaid\nand uninsured patients. This paper focuses on providers that provide only\noutpatient care.\nHEALTH AFFAIRS Volume 16, Number 4\nMEDICAID MANAGED CARE\n4. National Association of Community Health Centers, Access to Community Health\nCare: A National and State Data Book (Washington: NACHC, 1996).\n5. C. Orbovich, \"Collaborative Strategies for Success in the Changing Medicaid\nMarket: The Perspectives of Community-Based Providers and Managed Care\nOrganizations\" (Prepared for the Health Resources and Services Administra-\ntion, U.S. Department of Health and Human Services, April 1996).\n6. D. Lewis-Idema and M. Falik, Health Departments and Medicaid Agencies: Is the Cold\nWar Really Over? (Portland, Maine: National Academy for State Health Policy,\nAugust 1990).\n7. J. Schlitt et al., \"School-Based Health Centers and Managed Care: Seven School-\nBased Health Center Programs Negotiate a Difficult Fit\" (Washington: Making\nthe Grade Program, The George Washington University, February 1996).\n8. D.J. Zimmerman and C.J. Reif, \"School-Based Health Centers and Managed\nCare Health Plans: Partners in Care,\" Journal of Public Health Management Practice\n(Winter 1995): 33-39.\n9. Orbovich, \"Collaborative Strategies,\" 21.\n10. Lewin-VHI and MDS Associates, Community Health Centers' Performance under\nManaged Care (Rockville, Md.: Bureau of Primary Health Care, Health Re-\nsources and Services Administration, December 1994).\n11. Bureau of Primary Health Care, HRSA, \"Impact of Changes in the Nation's\nHealth Care System on Safety Net Providers: The Case of Health Centers\"\n(Memorandum from Marilyn Gaston, associate administrator, Bureau of Pri-\nmary Health Care, HRSA, July 1996). According to the bureau's review of 150\ncontracts between health plans and federally qualified health centers, an esti-\nmated 50 percent do not offer reasonable reimbursement rates (page 18).\n12. Some of these lessons appear in the paper by Orbovich, \"Collaborative Strate-\nSAFETY NET\n107\ngies,\" 15, and were raised by participants in the meeting sponsored by HRSA\nand AAHP, \"Collaborative Strategies,\" 1-2 April 1996, in Washington, D.C.\n13. Under Section 1915(b), states can ask to waive provider \"freedom-of-choice\"\nrules to mandate enrollment of Medicaid beneficiaries into managed care ar-\nrangements. Under Section 1115, states can ask to waive many rules to conduct\nresearch and demonstration projects that advance the objectives of the Medic-\naid program. Typically, Section 1115 waivers have been granted to allow states\nto implement mandatory managed care plans on a statewide basis and use the\nsavings to extend eligibility to additional groups of low-income persons.\n14. M.S. Sparer, \"Medicaid Managed Care and the Health Reform Debate: Lessons\nfrom New York and California,\" Journal of Health Politics, Policy and Law (Fall\n1996): 433-460.\n15. \"California Local Initiatives: Managing Healthcare in the Community,\" Access\n(Newsletter of the California Association of Public Hospitals and Health Sys-\ntems) (Summer 1996): 7.\n16. T. Mark et al., Medicaid Managed Care Program Access Requirements, Final Report\n(Report prepared for the Prospective Payment Assessment Commission, Pro-\nject HOPE, Center for Health Affairs, Bethesda, Maryland, March 1995).\n17. Ibid., 3-2.\n18. D.J. Lipson and N. Naierman, \"Effects of Health System Changes on Safety-Net\nProviders,\" Health Affairs (Summer 1996): 33-48.\n19. M. Gold, M. Sparer, and K. Chu, \"Medicaid Managed Care: Lessons from Five\nStates,\" Health Affairs (Fall 1996): 153-166.\n20. Ibid., 164.\nHEALTH AFFAIRS July/August 1997"
}